India's Digital Leap: the Multi-Trillion Dollar

M BLUEPAPER September 26, 2017 08:04 PM GMT

The Next India India's Digital Leap - The Multi- Trillion Dollar Opportunity ndia's digitization drive has raised our confidence in our long-term growth estimates. We forecast GDP to reach US$6 trillion, equity market cap to rise Ito US$6.1 trillion and the market value of the financials and consumer sectors to hit US$1.8 trillion and US$2 trillion by 2027.

Morgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of Morgan Stanley Research. Investors should consider Morgan Stanley Research as only a single factor in making their investment decision. For analyst certification and other important disclosures, refer to the Disclosure Section, located at the end of this report. + = Analysts employed by non-U.S. affiliates are not registered with FINRA, may not be associated persons of the member and may not be subject to NASD/NYSE restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account. M BLUEPAPER Contributors

MORGAN STANLEY ASIA LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ Anil Agarwal Ridham Desai Equity Analyst Equity Strategist +852 2848-5842 +91 22 6118-2222 [email protected] [email protected]

MORGAN STANLEY ASIA LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ Derrick Y Kam Parag Gupta Economist Equity Analyst +852 2239-7826 +91 22 6118-2230 [email protected] [email protected]

MORGAN STANLEY ASIA LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ Chetan Ahya Subramanian Iyer Sumeet Kariwala Economist Equity Analyst Equity Analyst +852 2239-7812 +91 22 6118-2234 +91 22 6118-2235 [email protected] [email protected] [email protected]

MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ Sheela Rathi Gaurav Rateria Rahul Gupta Equity Strategist Equity Analyst Research Associate +91 22 6118-2224 +91 22 6118-2237 +91 22 6118-2233 [email protected] [email protected] [email protected]

MORGAN STANLEY INDIA COMPANY PRIVATE LIMITED+ MORGAN STANLEY & CO. INTERNATIONAL PLC+ MORGAN STANLEY & CO. LLC Aayushi Kukreja Andrea Ferraz, CFA Brian Nowak, CFA Economist Equity Analyst Equity Analyst +91 22 6118-2245 +44 20 7425-7242 +1 212 761-3365 [email protected] [email protected] [email protected]

MORGAN STANLEY ASIA LIMITED+ MORGAN STANLEY & CO. LLC MORGAN STANLEY ASIA (SINGAPORE) PTE.+ Grace Chen James E Faucette Nick Lord Equity Analyst Equity Analyst Equity Analyst +852 2848-5835 +1 212 296-5771 +65 6834-6746 [email protected] [email protected] [email protected]

MORGAN STANLEY MUFG SECURITIES CO., LTD.+ MORGAN STANLEY & CO. LLC MORGAN STANLEY & CO. LLC Tetsuro Tsusaka, CFA Toni Kaplan Vasundhara Govil Equity Analyst Equity Analyst Equity Analyst +81 3 6836-8412 +1 212 761-3620 +1 212 761-3609 [email protected] [email protected] [email protected] M BLUEPAPER Contents

4 Preface

5 Quantifying the Digital Lift

6 India's Digital Economy - Two Pillars of Change

10 Investment Implications

14 Key Risks

15 Digital Payments and GST - The Catalysts of Change

28 Implication 1: GST Plus Digital Payments = Significant Big Data Opportunity

41 Implication 2: Digitization = Economic Growth Booster

48 Implication 3: A Massive eCommerce Opportunity

53 Implication 4: Digital to Drive Equity Bull Market

61 Implication 5: Financials – US$1.8 Trillion Market Cap in 10 Years

78 Risks to our Story

79 Individual Stock Section

112 Appendix 1 - JAM: Catalyzing Digitization

118 Appendix 2 - Paytm - The Disruptor That Is Becoming a Big Player

MORGAN STANLEY RESEARCH 3 M BLUEPAPER Preface

'Take up one idea. Make that one idea your life - think of it, dream of it, live on that idea. Let the brain, muscles, nerves, every part of your body, be full of that idea, and just leave every other idea alone. This is the way to success'. Swami Vivek- ananda

Digitization is that idea in India, right now. The government and the Central Bank are on a mission to rapidly formalize and financialize the Indian economy. India has introduced a universal biometric identification system (Aadhaar), initiated measures to boost financial inclusion (Jan Dhan), moved to a new fully online value-added goods and services tax system and implemented real-time payment systems (Unified Payments Inter- face and Bharat QR). Coupled with rising smartphone penetration, likely doubling from 300 million to nearly 700 million by 2020, these changes are driving India's digitization. We expect a step change in India's per capita income, banking system and stock market performance over the coming years. The channels of change include more financial penetration, greater tax compliance and increased credit to micro enterprises and consumers.

The result could be a multi-trillion dollar investment opportunity. Aside from the near-term teething issues involved in execution of such big changes and other cyclical problems faced by the economy, there is scope for visible shifts in economic activity starting in 2018 eventually leading to India being a) the third-largest economy in the world with a GDP of US$6 trillion, b) among the top five equity markets in the world with a market capitalization of US$6.1 trillion and c) the country with the third-largest listed financial services sector in the world with a market cap of US$1.8 trillion by 2027. We also expect India's consumer sectors to add about US$1.5 trillion to their current market cap of US$500 billion over this period.

There are implications beyond India. The concomitant increase in e-commerce, consumption basket, financial products and investments will make India a significant market for global corporations. Most importantly, if India succeeds, it will become the template for other emerging nations. While increasing financial inclusion has been the policy objective across emerging nations, India can provide leadership with its unique model. Hence, it is very important for corporates, investors and policymakers across the globe to observe and understand these developments in India. Indeed, there may be lessons for developed countries too.

William Greene Director of Asia Research Hong Kong, September 2017

4 M BLUEPAPER Quantifying the Digital Lift

Exhibit 1: India in F2027: The Power of Compounding Is Likely to Be Seen Across the Economy and Markets

Source: RBI (Reserve Bank of India), NHB (National Housing Board), IRDA (Insurance Regulatory and Development Authority), CEIC, AMFI, NPCI, NSE, BSE, Bloomberg, Morgan Stanley Research (E) estimates

Exhibit 2: The Stock Winners

Market Cap 3M traded value Price (US$ bn) (US$ mn) Indian Financials Bajaj Finance Rs 1,796 16 34 Edelweiss Financial Services Rs 255 3 12 HDFC Bank / HDFC Bank - ADR Rs 1,800 / US$ 94 72 41 ICICI Prudential Life Insurance Rs 389 9 9 Kotak Mahindra Bank Rs 1,000 29 31 LIC Housing Finance Rs 622 5 19 Mahindra & Mahindra Financial Services Rs 404 4 15 Indian Consumption Asian Paints Rs 1,197 18 16 Eicher Motors Rs 30,980 13 23 ITC Rs 262 49 75 MakeMyTrip US$ 29 3 17 Maruti Suzuki Rs 7,965 37 51 Ultratech Cement Rs 3,902 16 15 Global stocks with Indian optionality Alibaba US$ 170 434 2,855 Amazon US$ 940 451 3,431 DBS SG$ 21 39 65 Naspers ZAc 296,200 95 261 Softbank JPY 9,160 90 406 TransUnion US$ 45 8 74 Visa Inc US$ 103 236 721 MasterCard Inc US$ 139 148 414

Source: Bloomberg, Morgan Stanley Research. Note: Data as of September 25, 2017

MORGAN STANLEY RESEARCH 5 M BLUEPAPER India's Digital Economy - Two Pillars of Change

Exhibit 3: India in the Context of the World Current Position F2027E Position

GDP Ranks 7th on the basis of nominal GDP 3rd largest based on nominal GDP Ranks among the top 10 market caps Top 5 market caps, market cap to GDP @ Market Cap (Overall) globally, market cap to GDP @ 86% 101% Market Cap Market cap to GDP at 15% vs. 22% for G-7 At US$1.8 trillion, market cap to GDP at (Financials) countries 31% Among bottom quartile of the top 15 Within top quartile of the current list of top eCommerce eCommerce markets globally 15 eCommerce markets

Source: Bloomberg, RBI, IMF, Morgan Stanley Research (E) estimates

JAM and GST - the Forces Driving India's 1. JAM - The Holy Trinity of Jan Dhan, Technological Leap Aadhaar and Mobile Sets the Stage for India to Leapfrog into the Digital Age Over the past seven years India has undertaken two major reforms. The first involved a combination of schemes to biometrically identify all Indian citizens (Aadhaar) and to promote broad financial inclusion Jan Dhan: This initiative by the government launched in 2014 has (Jan Dhan). The second was the implementation of the indirect tax essentially ensured that almost every household in India has a bank reform in July this year, which moved India from its former, archaic account and can access their account anywhere. Some 285 million and complicated, tax system to a unified and completely digital accounts have been opened over the past three years. goods and services tax (GST) regime. Aadhaar: The beginning of India's digital revolution was the launch These two reforms have "digitized" India and brought the country to in 2010 of Aadhaar, which in Hindi means "foundation". The project an inflexion point in terms of growth, with a concomitant impact on was an ambitious one involving biometric identification of all India's stock returns, financial sector dynamics, consumption growth and people (1.3 billion of them). It has taken seven years but now the e-commerce activity. project is almost complete, with 1.2 billion Indians identifiable with either fingerprints or a retina scan - the scale and scope of this project is probably unmatched in world history for any government.

APIs (application programming interfaces) have been developed using Aadhaar to launch payment systems that allow real time C2C and C2M (customer-to-customer and customer-to-machine) transactions without needing any physical infrastructure - except a mobile phone. It also enables the completion of an electronic KYC (Know Your Customer) and the download of digital signatures.

6 M BLUEPAPER The benefits of Aadhaar can also be extracted via the private initiative transfer (DBT) program, which is a social welfare scheme under of Indiastack.org. IndiaStack is a set of opensource APIs that allows which subsidy benefits are directly transferred to the beneficiary’s governments, businesses, startups and developers to utilize India's bank account rather than via the issuance of checks, cash payments unique digital infrastructure to solve problems by facilitating pres- or rebates. Cumulatively, the total direct benefit transfer so far has ence-less, paperless, and cashless service delivery. Most recently, UPI been approximately Rs2 trillion and savings have been ~Rs500bn. - the unified payments interface - was added to this stack; it enables instant funds transfer over a smartphone using a virtual address. The GST has the potential to lift medium-term profit growth: From government has supplemented the digital infrastructure with laws the corporate sector's perspective, warehousing costs, freight costs, restricting values of cash transaction and incentivizing digital pay- and inventory levels could all fall. Together with a more transparent ments. and efficient input tax credit system and the removal of interstate barriers, these should lead to improved profitability. Over the next Mobile: We estimate that India has around 800 million unique two to three years, large companies could also gain share from micro, mobile users, and about 430 million have Internet access - a third of small and medium enterprises (MSMEs), whose effective tax subsidy India's population. We believe Internet access will double in the next (if they do not pay taxes) is likely to vanish under GST. That said, it is 10 years and we estimate that 915 million Indians will be on the not all bad news for MSMEs. Their entry into the formal economy, in Internet by 2026. addition to the transparency of the GST system, will enable flow- based lending to these entities, eventually lifting their growth as well as India's ratio of aggregate credit to GDP. 2. GST – Arguably India's Most Important Reform Since the Early 1990s Exhibit 4: How Digitisation Will Likely Impact the Economy and Markets

GST implementation and digitization should help government finances: India's brand new Goods and Services Tax (GST), with its ability to simplify India's complicated indirect taxation system and lift government revenues, has the potential to boost growth. GST completely alters the way government finances are managed in India. Hitherto, tax collection in India was decentralized, while expenditure was centralized through the Planning Commission, which set expenditure priorities for the states. GST centralizes India's taxation in the hands of the Central Government, while the abolition of the Planning Commission in 2014 had already set the stage for decentralization of expenditure to the state governments. This shift, in our view, will provide a major boost to Indian government finances.

India’s ratio of tax revenues to GDP is lower than the average for emerging markets and this has been a key reason why the fiscal deficit has been relatively higher. The implementation of GST will likely increase tax compliance on both indirect and direct taxes even while the tax rates are in and of themselves revenue-neutral. We estimate that if the primary fiscal deficit stays at 1.2% of GDP (as compared with an average of 2.1% over the past five years and 1.6% in F2017 on our estimates), the ratio of public debt to GDP will decline to below 60% by F2027 from 69% currently.

Digitization will help bring down government expenditure too, by helping the government conduct welfare spending with smaller leakages. The government has already implemented a direct benefit Source: Morgan Stanley Research estimates

MORGAN STANLEY RESEARCH 7 M BLUEPAPER Implications of India's Digital b) Financial Sector Revolution Transformation Historically the banking system has primarily catered to large corporates and well-heeled individuals. The costs of taking banking to the a) Greater confidence in future growth masses was fairly elevated, driving banks to focus on areas where revenue generation was high. However, this is changing for four main India is likely to be the world's fastest-growing large reasons: economy in the next ten years... 1.Regulations will force banks to lend to segments other than large India's economy already had strong growth prospects for the next corporates. Unlike in the past, future funding to large corporates will ten years. The trend line in India's annual GDP growth has been accel- be shared between banks and bond markets. From F2020 any com- erating from 5.8% in the 1990s to 6.9% in the 2000s. We think this pany with debt of greater than US$1.5 bn will be defined as a large trend will likely continue for the next decade given the following corporate. These large companies will have to meet 50% of incre- structural factors: mental funding through the bond markets. This will force the banks to look for new avenues of growth. Favorable demographics: Over the next 10 years, 122 million individuals are likely to enter the work force, which is equivalent to about 2.Technology is likely to bring down the cost of providing financial ser- 20% of India's current work force. vices significantly and to help grow consumer credit by around a 17% CAGR over the next decade. Even after such growth, total consumer Globalization: This provides the enabling factors of external demand loans will be low at 25% of GDP. There are two ways in which tech- and financing that can be used to boost growth. nology is facilitating change:

Reforms: The government is continuing the reforms that India started a) By reducing the cost of opening accounts and hence expanding the in the early 1990s, which relate to the ease of doing business, FDI, reach of banks. Aadhaar and the associated eKYC has cut the cost of government finances, taxation, infrastructure and greater autonomy opening a deposit account by 90%. A recent example of this is Kotak's for states. 811 initiative based on digital infrastructure, where the bank opened over a million accounts in one quarter on top of an existing customer …and digitization reinforces our views. We lift our annual long-term base of eight million. growth forecast by 50 bps. b) By enabling faster loan delivery. An example would be HDFC Digitization is integral to two changes: a) policy initiatives that are Bank's "loan in ten seconds" for unsecured personal loans. The bank boosting financial inclusion and b) technology changes that are uses its own data base on individuals' transactions along with credit reducing the cost of delivering financial services to the masses and scores from bureaus to offer quick loans to individuals. We expect small enterprises. These, along with the government's focus on this system to increase and to be implemented, on the same scale, by employment for all, will make growth more inclusive, which in turn other banks too. The other driver of loans to individuals would be makes us more confident about India's growth outlook. We estimate government schemes around affordable housing. that digitization will provide a boost of 50-75bps to GDP growth and forecast that India will grow to a US$6.0trn economy and 3. Cash flow data from the GST network and digital payment systems achieve upper-middle income status by F2027. We expect India's will boost credit to MSME's: These enterprises account for 80% of real and nominal GDP growth to compound annually by 7.1% and employment generation in India. Hitherto, due to a lack of reliable 11.2% (10.2% in US dollar terms) respectively over the coming financial data, MSMEs were largely excluded from the formal decade. We project gross FDI inflows amounting to US$120bn by banking system for credit and thus accounted for around 14% of F2027, almost double the current 12-month trailing run rate of credit from banks, we estimate. The online infrastructure of the GST US$64bn. network enables any taxpayer to securely share tax payment records (and therefore cash flow data) with anyone else. For the first time, Indian banks will have reliable data on their cash flows. Combined with a rise in usage of digital payments, a credit provider can now make a much more informed decision on the creditworthiness of

8 M BLUEPAPER such small borrowers. We expect lenders to reduce spreads on d) An eCommerce Boom MSME lending by up to 300 bp as they get access to reliable and better quality data. We think this will drive MSME loan growth at a high teen CAGR for the next decade. This in turn could be a catalyst With over 900mn Internet users in 2026 and over half of them for overall growth, likely adding ~30 bps to real GDP growth on an shopping online, we believe India's eCommerce market could annualized basis over the next decade, India's digital transformation grow to US$200bn by F2027, making it one of the most and employment generation. attractive opportunities globally. This growth is being driven by a combination of rising internet penetration, a drop in data access 4. Gains for the private sector will continue at the expense of the state- costs, a shift to smartphones and a flow of credit to consumer and owned banks: The starting point is that there is financial under-pene- micro enterprises. tration in India, with loans as a percentage of GDP at 67%, consumer credit at 16% of GDP and formal system lending to MSMEs at ~10% 1. Strength of MSMEs should help augment the supply chain: of GDP. India's rapid digitization will impact financials negatively by Currently, the number of merchants selling on online platforms such reducing revenue yields across key lines and resulting in lower CASA as Flipkart is around 150,000. Amazon India currently has 160 million (current and savings account) and lower distribution fees. That said, products listed on its platform, versus 400+ million for Amazon in for tech-savvy banks, digitization should likely positively alter the the US. We expect the numbers in India to increase meaningfully as growth trajectory in two ways: more merchants enter the formal economy and have better access to credit. i.Volumes - We expect tech-savvy banks to take market share from the data laggards, especially the SOE banks. 2. Faster growth of online shoppers and per capita spending: India had about 60 million online shoppers in 2016, which was ~14% ii. Costs- The cost base of India lenders has historically been fairly high of the total Internet population versus 64% in China. An AlphaWise compared with that of other banks regionally due to the fragmented survey (link here) conducted in December 2014 showed that Internet banking system and weak economies of scale. However, as tech- users with less than two years of activity on the Internet were less nology takes root, we expect costs to drop significantly, compen- prone to transacting. However, users with over five years of experi- sating for lower revenue yields. ence on the Internet were more likely to transact online.

From F2020, more than half of India's Internet population will have c) Strong Household Consumption matured, with five years' or more usage, and we think this will likely mark an inflexion point in online shopping. We expect 475mn online There are two major trends driving consumption in India. First is rising shoppers in F2027 and online spending of ~10% of per capita income, consumer credit penetration aided by digitization and the likely cul- underpinning a boom in eCommerce. ture shift happening as a younger generation becomes less debt averse. Rising consumer credit, which we think will compound at 17% 3. Evolving payment methods: Cash on delivery is a prominent pay- annually over the next decade, could add 20 bps to annual real GDP ment method, accounting for about 55-60% of total eCommerce growth. Second is the shift away from food consumption to non-food sales. One of the key reasons for the high dependence on cash is the consumption. Based on the CPI basket weights, food items account trust deficit that exists between customers and online merchants. for 47% of Indian consumers' expenditure, which is one of the highest However, with the government's push to increase digitization of pay- percentages globally. This likely reflects India's current level of low ments, we believe cash will lose market share to credit/debit cards economic development. As per capita incomes rise, the share of and payment wallets. This would improve the online shopping expe- expenditure is likely to shift towards non-food consumption, which rience. will therefore grow faster than the economy as a whole. Such growth will be further augmented by the growth in housing, which forms the key platform for non-food consumption growth. Further, digitization will likely lead to an increase in the share of the organized sector in the consumption space. Versus an overall annual GDP growth of 11%, it is quite possible that organized non-food consumption registers mid-teen growth in the coming decade.

MORGAN STANLEY RESEARCH 9 M BLUEPAPER Investment Implications

Exhibit 5: Summary of Our Forecasts F2027E US$ Bn Current Bull Base Bear Real GDP, YoY % (Avg)* 7.1% 8.4% 7.1% 5.9% Nominal GDP, F2017 2,279 7,095 6,040 5,368 Nominal GDP, YoY % (Avg)* 11.3% 13.0% 11.2% 9.5% GNI Per Capita (US$), F2017 1,702 4,809 4,135 3,546 Total Loans, F2017 1,530 5,784 4,698 3,305 loans as % of GDP 67.1% 81.5% 77.8% 61.6% Mutual Funds AUM, F2017 273 3,415 1,935 879 Domestic Equity Mutual Funds AUM 111 1,530 1,000 481 10-year Domestic Equity Flows 60 706 420 211 Life Insurance Premiums, F2017 62 286 185 93 General Insurance Premiums, F2017 18 80 59 26 Total Internet users (Mn), F2017 432 1,010 915 841 E-commerce, F2017 15 251 200 105 Market Cap (Overall) 2,100 8,500 6,100 3,700 Market Cap (Consumers) 525 2,975 2,000 962 Market Cap (Financials) 460 2,904 1,844 917 BSE Sensex (Points) 31,627 1,30,000 1,00,000 65,000 Aggregate Trading Volumes 21,289 57,375 40,951 23,618 Cash Trading Volumes 1,131 6,375 4,575 2,220

Source: RBI, CEIC, AMFI, Capitline, NSE, BSE, Bloomberg, Morgan Stanley Research (E) estimates. Note * GDP growth forecast is 10-year average. a) Ultimately, Stronger GDP Growth = Profit share in GDP is a function of the investment cycle, the current account, and the saving rate of the government and households. We Stronger Equity Market Performance see the investment cycle turning in the coming months. The government’s saving rate is rising, which is a likely drag on earnings – but the The concomitant impact of higher GDP growth is that corporate changing mix of household saving from physical to financial is a big earnings growth is also likely to accelerate. India's history shows a boost for profits. The current account appears to be contained, which good correlation between nominal GDP growth and earnings. We is again good for the share of profits in GDP. expect this relationship to persist in the future. Ultimately, earnings growth is what drives share prices – but we also expect Indian multi- 2. We also expect equity multiples to expand: Two factors are ples to expand. We think India's stock market could be among the likely to drive this expansion. world's best performers in the next ten years, leading to India's market cap rising from ~US$2 trillion to ~US$6 trillion. Strong domestic participation: We think India is in the midst of a domestic liquidity supercycle. The starting point of under penetra- 1. Stronger economic growth should drive stronger corporate tion combined with digitization almost guarantees growth in finan- earnings growth: Strong nominal GDP growth plus a rising ratio of cial assets and liabilities on the household balance sheet driving up credit to GDP should be good for corporate earnings, and a low the size of the domestic mutual fund industry and household owner- starting point of corporate profits / GDP, which sits well below trend, ship of equities. We project equity saving of US$420bn–US$525bn augurs well for corporate profit growth in the coming years. over the next ten years, versus the respective US$60bn and

10 M BLUEPAPER US$120bn that households and foreign portfolios invested over the Prior to the ongoing technological revolution, our view was that previous ten years. The three main sources of these flows are mutual Indian banks would struggle to grow credit higher than GDP growth funds, pension and provident funds, and insurance companies. ETF given the expansion in the base. However, we have altered our view AUM could see 30x growth in the next decade, driven by provident due to the advent of fintech, which aids financial penetration, as well fund flows. as the likely transfer of market share from the capital starved SOE banks that currently control almost 55% of the market - now we The effect of the declining ratio of public debt to GDP: As we think banks will grow faster than nominal GDP in the coming decade. argue, robustness in government revenues due to the execution of We expect private sector banks, non-banking finance companies, cap- GST will likely to lead to lower public debt to GDP. This leads to a ital market intermediaries, private sector insurance companies and crowding in of private investments with a positive impact on interest select fintech companies to be the beneficiaries, accentuating the rates and inflation. This will also lead to higher multiples, which is trend that has already been in place for the past five years. evident from the historically inverted relationship between change in public debt and PE multiples. We see the BSE Sensex crossing the Consumption sector - We forecast market capitalization of the 100,000 mark, albeit the bulk of the returns are likely to be consumption sector to grow 4x from current levels, to US$2 tril- front ended in the coming five years. lion in the next ten years: Future consumption is being advanced through leveraging, and non-food consumption itself is likely to grow faster than overall consumption, setting this sector, especially dis- b) Financials and Consumption Sectors to cretionary consumption, up for strong performance in the coming Gain Share in the Market at the Expense ten years. Consumption-related sectors including consumer discretionary and consumer staples are just over 25% of the market. We of Global Sectors estimate this ratio could rise to 33%.

Together the financials and consumer sectors could account for 63% of India's market by 2027, up from 47% currently, largely at c) The Stock Winners the expense of global sectors such as software services, pharmaceuticals, materials and energy. Exhibit 6 outlines the potential winners from the digitization story of India. We cannot claim this is an exhaustive list but it is good We forecast market capitalization of the financials sector to starting point for an investor who wishes to participate. The charac- grow 4x from current levels, to US$1.8 trillion in the next ten teristic of this list is that it contains companies from across the world, years: The past two decades have been very strong for Indian finan- because the benefits of India's digital initiatives will be felt globally. cials. Listed market cap has increased from less than 2% of GDP in 1998 to 16% now – an >80x increase in nominal terms. This has been For global companies on this list, we have focused on their exposure driven by various factors with the setting up of efficient businesses to India as a share of their overall business combined with the foot- (private banks and well-run non-banking finance companies) as a key print plans. Indeed, the drivers of growth, I.e., financial, consumption anchor. These businesses took away share away from the inefficient and e-commerce for a key filter criteria for us. state-owned banks that dominated the banking scene until the mid- 1990s. We expect the financials sector's share in India's market cap to For Indian companies, we look for companies with a strong track rise to 30% from 22%, currently. record of return on capital as well as our estimated return on capital combined with reasonable stock valuations. The macro exposure to consumption and the financial sector is an over riding feature for our selection. We have left out smaller stocks from this list.

MORGAN STANLEY RESEARCH 11 M BLUEPAPER

Exhibit 6: Key Stocks: Our Comments on Why Indian Growth Matters

Source: Morgan Stanley Research estimates

12 M BLUEPAPER

AlphaWise Survey Results: MSME Survey - Rising Awareness While awareness of UPI / BHIM is quite healthy, its adoption is Towards Digital India unlikely to pick up in a big way like cards or wallets since they don’t see much demand from customers for this. There has to be more We carried out our survey in July and August 2017 with 1,522 micro steps taken to increase awareness and hence usage of these new pay- & small businesses and merchant establishments across 28 cities in ment options. India using face-to-face interviews. The survey sought to understand the payment and financing options used and likely to be adopted by Digital payments to drive access to credit; cost of funding to be these businesses, as well as the drivers and barriers to adoption. The key: According to the respondents, currently only a third of working survey represents the micro & small businesses in the large metro capital is being funded through formal channels (funding is mainly to and smaller cities in India. Some of the key takeaways from the survey larger and B2B business). The major share is coming from own funds, were: followed by personal loans from banks / NBFCs (among those who don’t have bank credit). However, as businesses adopt digital pay- Adoption of digital payments to be driven by eco-system support ments and see businesses that have adopted digital payments get and convenience: There has been a meaningful increase in awareness better access to formal credit, they are more inclined to explore avail- and usage of digital payment options – even for newer options like ability of formal credit – though cost of funds would be the key UPI (Unified Payment Interface) / BHIM (Bharat Interface for Money). deciding factor. Businesses do believe GST will help them in getting Businesses have seen a significant drop in cash transactions in the access to formal credit. past 6-12 months. The share of new channels is still low and the bulk of demand is still driven by cards. Data suggests that customer We have incorporated the results of this survey in this note. demand, support by the eco-system and convenience are the key drivers of adoption, and to that extent the cost of transaction is secondary.

MORGAN STANLEY RESEARCH 13 M BLUEPAPER Key Risks

Investors should be cognizant of the things that can go wrong with What If Digitization Fails? our thesis:

l Politics and policy: Political instability or shifts in policy If India's initiatives face road blocks in any of the aforementioned away from digitization would present risks to the core ways, our forecasts will be challenged. For our bear case to unravel, thesis. other things also need to go wrong. But failure of the digitization l Cyber security: Since a large part of this transformation is effort will likely cause our estimates for the economy, markets and about the Internet, cyber security presents an active and financial services to end up somewhere between our base and bear important risk to the story. case. This means that: l Privacy debate around Aadhaar: While the Supreme Court in India has made privacy a fundamental right in a 1. GDP will likely settle at US$ 5.5 trillion, implying a nominal growth recent landmark judgement, private parties will likely con- rate of 9% over the next ten years. tinue to question whether Aadhaar violates privacy rights. Any adverse judgement in the courts could derail one the 2. India's market capitalization could grow more modestly at 9% to main anchors of our framework. US$ 5 trillion by 2027. l Smooth transition to GST: The GST rollout has begun well but it is still work in progress. In the short run there 3. Loan growth by banks will likely be closer to 10% CAGR for the could disruption to smaller businesses causing job losses coming decade as MSMEs struggle to get loans and corporates move and a general slowdown in economic growth. to bond markets compared to the current base case of ~13% CAGR. l India’s often discussed macro risks: India's demographic dividend could become a problem if the country is unable 4. Accordingly the financial and consumer sector returns may also be to execute on infrastructure, generate enough jobs, and tempered and the respective market capitalizations will likely settle educate the people and keep them healthy. at lower levels somewhere between our bear and base case. l Deep global recession: While India's story is idiosyncratic given the size of the economy, it needs a stable global economy to deliver growth. A global crisis like we saw in 2008 could damage India's growth prospects.

14 M BLUEPAPER Digital Payments and GST - The Catalysts of Change

Key Takeaways

1. India is shifting from being a cash-dominated economy to one that is digital enabled, by Aadhaar, large-scale bank account openings, increased mobile penetration and a sharp decline in data charges (JAM). The technology infrastructure includes a real- time mobile payment system (with most banks as participants) that allows C2C and C2B payments, an inter-operable Quick Response or QR code system and a real-time bill payment system. Bank-backed systems are likely to benefit considerably, but larger mobile payment companies should also do very well given significant customer and merchant acquisitions. 2. In our base case, we assume that digital payments can increase to ~20% of GDP over the next ten years, implying a CAGR of ~30%. Currently Visa and Mastercard are the biggest players in digital payments, with ~80% share. We expect new avenues (such as UPI, Mobile Wallets and Rupay Cards) to quickly account for two-thirds of the digital payments market by 2027. 3. GST, the other leg of India’s digital lift, is a sweeping change to India’s archaic tax laws that will boost government revenues and corporate profits. It will also likely lead to a major change in credit availability for micro firms, in turn formalizing the Indian economy.

Why Have Digital Payments Lagged in Exhibit 7: India has Lagged Materially in Digital Transactions (Per Capita Non India? Cash Transactions) Non Cash Payments Per Annum Per Capita 728 Cash is the predominant form of payments in India. As per Niti Aayog 750 (an Indian government policy think tank), the number of non-cash 600 transactions per capita is one of the lowest among big economies at

~11 per year. 450 355

This has been driven by both aversion to non-cash transactions and 300 a lack of payment infrastructure. To start with, there was reluctance 142 150 among consumers to use means other than cash. This was accentu- 70 26 32 ated by the lack of payment options. This was true in both urban and 11 0 rural areas. In urban areas, bank branches have been omnipresent - India China Mexico South Brazil UK Singapore Africa but acceptance of non-cash form of payments has been limited. For Source: Niti Aayog instance, if we look at POS terminals in India, as of 2015 the POS density was ~1,000 terminals per million people. The numbers in other EMs have been much higher - for instance, Brazil is almost 25x India.

MORGAN STANLEY RESEARCH 15 M BLUEPAPER This has been driven by various factors: Exhibit 8: The Number of Point of Sale (POS) Terminals per Million People Is Very 1. Consumer awareness towards digital payments has been fairly lim- Low in India ited in India. Digital payment as a mode was not available to most of POS Terminals per Million population (2015) them - and even when it was available, they preferred cash given 42000 habits and the ease of use. The government drive since late 2016 to 35000 increase digital payments has increased awareness and usage, but it 28000 will take time for them to proliferate. 21000 14000

2. The dominance of cash in the economy was due to the fact that 7000 both merchants and customers wanted to pay in cash, as that would 0 UK India Brazil Russia Mexico

keep transaction records away from formal reporting. Canada Australia Singapore South Africa South 3. The cost of setting up physical infrastructure has been quite high, Source: Niti Aayog Report especially for smaller businesses and merchants. As per Visa, the cost of a POS machine is Rs 8,000-12,000. On top of this, the annual operating cost per machine is Rs 3,000-Rs 4,000. This makes it costly for Exhibit 9: a small merchant to set the machine and start accepting card pay- Why Is Usage of Digital Options Lagging? ments. Habit / Ease 88%

4. While the interchange rate (MDR - Merchant Discount Rate) in percentage terms is not very high in India (0.5-2.0%), merchants are Supplier driven / to manage purchases / cash flow 61% averse to paying this MDR given the relatively low margins of their operations. A number of small merchants would add the MDR to the Due to customer preference 57% bill when a customer paid by card - thereby incentivising cash payments. No transaction / equipment costs 31%

5. The low value of transactions at smaller merchants coupled with Source: AlphaWise, Morgan Stanley Research the capping of MDR by RBI (Reserve Bank of India) meant that banks were averse to expanding POS terminals to smaller merchants.

AlphaWise: Our survey of ~1,500 businesses showed that the share of cash transactions in turnover (for this set) is 59%. About 30% have more than a 75% share of cash, about 25% have a share of cash between 51-75%, and for 45% the share of cash transactions is 50% or lower. Businesses with a lower share of cash are those with a higher turnover and B2B businesses, while the larger segment i.e. B2C businesses have a much higher share of cash transactions, especially in Tier III cities. Habit/ease of use and help in managing payments / cash flow better were mentioned as the key drivers of cash usage.

16 M BLUEPAPER Government and Regulatory Drive In our report, Global Financials and Payments: Fintech – A Gauntlet to Riches (17 May 2017), the Morgan Stanley research team high- Towards Digitisation lighted the key drivers for a successful fintech framework. In our view, supportive regulations and availability of data to everybody are among the key factors required for a successful framework - India The situation with digital payments is changing now. Both the gov- clearly embodies this. ernment and regulators realize the importance of cashless transactions and the potential impact on sustainable economic growth. Exhibit 10: Government Targets a 6x Increase in the Number of Annual Retail Dig- From a regulatory perspective, there has been a significant push to ital Transactions create a digital payment infrastructure to increase adoption. From 30000 Digital Transactions (Volume, Mn) the government side also, there has been a continued push towards 25000 adoption of digital payments. This will improve the visibility of cash 20000 flows in the economy and thereby help the economy grow on a more 15000 sustained basis. 10000

5000 The pace of pickup in India has been fairly good (off a low base), with about a 50% CAGR between F2013 and F2017. However, the govern- 0 F13 F14 F15 F16 F17 ment has laid out a very ambitious target of increasing the number

of digital transactions to about 25 billion in F18 - around a 6x increase Target) (Govt F18 in one year. And it is moving aggressively towards ensuring it reaches Source: RBI, Morgan Stanley Research the target.

Exhibit 11: Morgan Stanley Framework for Fintech Success - India Embodies Most of These Factors Success in Fintech

Financial Infrastructure Does this product truly add value? Underdeveloped? Enough to drive change among inertial consumer financial behavior? If not, stop here. No Yes

Well-Defined & Accommodating Government Actively Driving Regulation? Change?

No Yes No

Amid Inflection in Industry Highly Yes Consumer Behavior, e.g. Yes Fragmented? Digitization will help improve data Technology? availability materially

Yes

No Financial Data Accessible? No

No Yes

Disruption less likely; Better off partnering with incumbents; Disruption possible, but Rapid Disruption/ Or slow disruption carried out by incumbents slower to play out TAM Creation

Source: Morgan Stanley Research

MORGAN STANLEY RESEARCH 17 M BLUEPAPER We believe that it will be tough for India to achieve 25 billion transac- Exhibit 12: tions in F2018. But given the ambitious target and policy makers' Digital Payment Options: Awareness of Older Options Has Increased drive to achieve this, we would still expect a 3-4x increase in transac- Meaningfully - Though This Remains Low As Share of Transactions 80% tions in one year. All the parts of the infrastructure required to drive Awareness Usage Share in transactions a strong pickup are falling into place in India. 65%

52% As mentioned in the AlphaWise survey above, the main reason for the 50% high usage of cash in the economy has been awareness, habit and 30% ease of use of cash compared to various digital payment options. But 28% 21% the government drive to increase awareness and the strides made by 16% mobile wallet companies during the Currency Replacement Pro- 4% 5% 3% 0% gramme (CRP) at the end of 2016 have increased awareness about Debit / Credit Card Direct transfer Mobile Wallet BHIM/UPI digital payment options materially. Source: AlphaWise, Morgan Stanley Research BHIM = Bharat Interface for Money

India Has Set Up a State of the Art Digital Payment the country today being smartphones. Since 2012, the share of Infrastructure smartphones in total shipments has increased from 7% then to 50% now, and we estimate it could reach 100% by 2020. Three big trends have emerged in India over the last six to seven years. These have brought about a universal identification / authenti- The other factor that is acting as a big driver is the sharp decline in cation mechanism for the entire population (Aadhaar), ensured that data charges. Data charges were already coming off meaningfully every household in India has access to a bank account (Jan Dhan), and and the launch of Reliance Jio has caused this to fall even more (the led to a surge in mobile penetration and a rapid pickup in smartphone blended data charge is down from around Rs. 400 per GB in 2012 to usage. Government and policy makers realized that combining these around Rs 60 per GB currently). This means that everybody with a trends would be quite powerful, and this has given rise to the JAM smartphone also has a bank account and digital identity, and that (Jan Dhan, Aadhaar, and Mobile) initiative. Appendix 1 explains in person can now be constantly connected to the internet. detail the various payment networks that have been set up in the country. India Stack - Bringing This All Together Some key indicators about this JAM mega-trend are as follows. The above initiatives are major advances but somebody needs to take 1. In August 2014, India launched a plan to ensure everybody has a the lead and combine them into product offerings. This is being bank account. Since then, ~285mn bank accounts have been opened driven by India Stack. This is a set of APIs (Application Programming under this scheme. Prior to the launch of this policy, ~35% of house- Interfaces) that leverages this infrastructure to move towards a pres- holds in India did not have a bank account. But now we estimate that ence-less, paperless and cashless service delivery - see Appendix 1 for most Indian households have access to a bank account. details.

2. The universal identity programme was launched in 2010. The A number of APIs have been launched, with full regulatory backing, intent was to ensure that every Indian has a form of identification. that utilise this infrastructure. Some of the key ones are e-KYC (which Since then, the government has created a biometric digital database allows electronic KYC, making customer acquisition quick and very (with fingerprints and retina scans) of 1.2 billion people, which is 92% low cost); UPI (which allows mobile based payments on a real-time of the country's population. Prior to this, nearly half of the country basis between individuals and businesses); the Bharat QR code (an had no form of identification. interoperable QR code system with all the key payment gateways and banks on board); the Bharat Bill Pay System (moving all bill pay- 3. There has also been a surge in usage of mobile phones in India, with ments in India - telecoms, utilities, government and so on - to one the number of mobile connections at ~90% of the population. This platform); and the Aadhaar Enabled Pay System (which enables was enabled by fairly cheap telecom tariffs in the country and the people to carry out transactions using an Aadhaar Card and finger- launch of no-frills low-priced phones. There has also been a sharp prints for biometric authentication). increase in smartphone penetration, with ~25% of mobile phones in

18 M BLUEPAPER This may feel like a long list of initiatives. But they essentially cover the entire gamut of payments in the country. All one needs to make these payments is a mobile phone and a bank account. These are real time and also, given that these systems are regulator backed, they are interoperable, which should help digital payments to grow quickly in the country. A detailed explanation of these initiatives is in Appendix 1.

Exhibit 13: India: Key Initiatives Using India Stack and JAM

Source: Morgan Stanley Research

MORGAN STANLEY RESEARCH 19 M BLUEPAPER Fintech companies have been using the increased mobile pene- Exhibit 14: tration to grow digital payments quickly M-Wallets Transactions Versus POS

1000 10.0% M-Wallets Transaction Values (Rs Bn) As % of POS Transactions, RS

As mobile phone penetration has increased in India over the last few 800 8.0% years, various fintech start-ups became significant players in the pay- 600 6.0% ments eco system. This is similar to the situation in other big markets such as the US and China where a majority of the large fintech compa- 400 793 4.0%

532 nies originated in the payments space given the large market oppor- 200 2.0% tunity and the need for a ubiquitous experience for all in the value 82 206 29 0 10 0.0% chain - especially consumers and merchants. F2013 F2014 F2015 F2016 F2017

Mobile wallets have been the pioneers of digital payments in India. A F1Q18 (Annualized) F1Q18 number of these were set up over the last few years (PayTM, Mob- Source: RBI, Morgan Stanley Research iKwik and Freecharge among others). These wallets started off focusing on fairly niche payment streams - like mobile phone recharging - but have evolved into systems straddling the payments Exhibit 15: space. Transaction values on mobile wallets have also picked up Breakdown of Fintech Funding into Various Sub-categories: Payments materially - from a slow start in F2012, the transacted value reached Dominate the Total Funds Raised 100% 1% 1% 1% Rs532bn in F2017 (~8% of POS transactions) and if we annualize 3% 1% 5% 9% 5% 9% 2% 3% 9% F1Q18 data, it will likely be 1.5x the F2017 figure in F2018. 4% 3% 80% 37%

Funding into fintech has been very strong - A number of products 60% offered by fintech payment providers have to be seeded by a strong 84% 86% capital base - given fairly large losses in the early years of customer 40% 82% acquisition. This has been helped by strong fund flows into the Indian 55% 20% fintech space from venture capitalists, private equity and strategic partners since 2015. Total funding to fintech companies has been 0% US$3.1bn since 2014. 2015 2016 2017 YTD 2017 As these wallets have evolved, their usage have moved to various Payments Mobile POS Lending Wealth Management InsuranceCumulative types of payments. For instance, telecom recharging was a large part Source: Economic Times, VC Circle, Morgan Stanley Research of wallet usage until two or three years back. This is now around a quarter of usage and various other uses have picked up, driving the sharp growth seen in the last one or two years. This is similar to China, where usage has evolved quickly over the last decade.

20 M BLUEPAPER

Exhibit 16: Digital Wallets' Segment-wise GMV Share: Uses for Digital Wallets Have Expanded Over Time On-demand online 16% 18-23% platforms and Offline (not included in China data)

16% 47% P2P, utilities and others 20-22% Online gaming 11% 1-2% 9% 2% 3% Telecom recharging 3% 4% 25-30% 8% B2B eCommerce

14% 15-20% Flight ticket/travel 45% Online shopping 22% 20-25%

0% Fund purchasing China (2005) China (2016) India (2016)

Source: iResearch 1Q-CY2016 press release for China data, Company data and industry sources for India data, Morgan Stanley Research Note: China data in the chart above is for desktop only; segment-wise data (especially for on-demand online platforms and offline transactions) on mobile is not available publicly and hence not included

Roll-Out of Digital Infrastructure Should Exhibit 17: Cross Country Comparison: Digital Payments as a Percentage of GDP 50% Cause Greater Acceptance and a Surge in 44% Digital Payments 40% 32% 32% 30% 30% 27% 25%

19% 19% 20% Digital payments in India grew at a good pace from F2005-F2014, at 15% 16%

~26% a year. However, this was against the backdrop of fairly good 9% 10% 8% 8% 7% nominal GDP growth and a strong pickup in private banks' share in 5% India, implying greater card usage. If we look at digital payments as 0% UK USA ITALY INDIA* BRAZIL KOREA RUSSIA MEXICO TURKEY FRANCE a percentage of GDP, the move was fairly gradual from ~1% of GDP CANADA GERMANY AUSTRALIA

in F2006 to ~2% in F2014. Since then we started seeing a pickup to NETHERLANDS AFRICA SOUTH 3% of GDP in F2016 and ~5% in F2017 (with the sharp uptick driven Source: BIS, Morgan Stanley research; India data is for F17. For others data as of 2015. by the government's various initiatives). Exhibit 18: When we compare India with other markets, digital payments in India Cross Country Comparison: Digital Payments as a Percentage of PCE (even after the F17 pickup) have lagged those in most markets. We 120% look at payments as a percentage of GDP and also payments as a per- 91.7% centage of national personal consumption expenditure - on both 80% counts, India has lagged. 55.9% 51.4% 42.6% 43.8% 45.1% 40% 34.5% 35.0% 26.9% 28.8%

14.8% 15.9% 11.6% 12.3% 8.3%

0% UK USA ITALY INDIA* KOREA BRAZIL RUSSIA MEXICO TURKEY FRANCE CANADA GERMANY AUSTRALIA SOUTH AFRICA SOUTH NETHERLANDS

Source: BIS, Morgan Stanley research; India data is for F17. For others data as of 2015. MORGAN STANLEY RESEARCH 21 M BLUEPAPER We expect this to increase to around 20% of GDP by F2027, implying Exhibit 19: a CAGR of ~30% for the next ten years. What will drive this? India: Millennials as a Percentage of Work Force 80% 69% 67% 70% 64% 1. Continued increase in younger population in the work force - 62% 59% 60% 56% India has among the highest numbers of young people entering the 54% 51% 48% work force over the next two decades. As per UN estimates, in 2027, 50% 45% 42% 39% millennials in the work force will be ~680 million people, an increase 40% 36% of ~350 million compared with the 2016 level - that’s ~70% of the 30% work force. These individuals are not attached to old banking habits 20% and they will have access to mobile phones / smartphones. We would 10% expect them to adopt digital banking much more quickly than past 0% generations. 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

Source: UN Population database, Morgan Stanley Research 2. The payment infrastructure that is now available in India is amongst the best in class and completely inter operable (with almost the entire banking system and various payment networks). Hence, Exhibit 20: the issue of availability has been taken care of. Currency in Circulation as a Percentage of GDP Currency in circulation, % of GDP 13% 3. Less currency in the system - Between F2000 and F2016, the 12% amount of currency in circulation went up sharply (~8x), growing at 11% a CAGR of 14%. Also, as a percentage of the formal banking system, 10% the proportion rose materially to 20% from 12% in F2000. This flow 9% of currency enabled cash transactions to flourish. 8%

7% However, the currency replacement programme has disrupted this 6% trend. Currency in circulation as a percentage of GDP had declined to ~8.5% from 12% in F2016. Although this number has started rising F1999 F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 Sep-2017 (now ~9.7%), we expect it to stabilise at well below pre-November Source: RBI, Morgan Stanley Research 2016 levels. This reduction in cash availability will raise the levels of acceptance amongst both consumers and merchants of digital payments. 5. Lower Merchant Discount Rates (MDR) - A deterrent to acceptance of non-cash payments historically has also been the relatively 4. More aggressive tax code will reduce the ability to evade taxes high MDR charges, especially on debit cards where the issuing bank - India also launched the Goods and Services Tax (GST) regime in July is not taking any credit risk. RBI has lowered the rates since 2012 on 2017. Under GST, a buyer of goods or services will be able to take debit card transactions - from similar to credit card rates to rates credit for the tax paid only when the supplier files taxes and gives the based on the value of the transaction (currently this stands at 0.25% data to the GST. The GST Network (GSTN) will maintain the data on for transactions up to Rs. 1,000 in value, 0.5% for Rs 1,000-2,000 tax paid by the entire value chain - and hence, if any part of that chain and 1% for transactions above Rs 1,000). evades taxes, the cross check will capture this discrepancy. This should dis-incentivise tax evasion. Although the pressure to reduce MDR is reasonably high, RBI has to ensure it is economical for merchant acquirers to conduct this busi- As tax evasion becomes tougher, one of the deterrents to digital ness. Otherwise these banks will stop acquiring merchants. Now RBI acceptance - that cash transactions are not recorded - will become has issued a draft paper suggesting MDR on debit cards should be less important. This should make merchants - big and small - more linked to the size of the merchant. The idea is to ensure that payment amenable to the acceptance of digital payments. infrastructure can be adopted by small merchants also.

22 M BLUEPAPER

Exhibit 21: RBI: New MDR Proposal Korea could be a case study for India MDR as % of Transaction Value For Physical POS For Digital POS Just as India has launched the government initiatives Small / Special category merchants Not > 0.4% Not > 0.3% All other merchants Not > 0.95% Not > 0.85% described above, Korea took steps in the late 1990s to Flat fee of Rs5 for up to Rs1000. Fees of Rs10 for reduce the shadow economy. This drove a sharp increase Government transactions Rs1000-2000. MDR not exceeding 0.5% for greater than Rs2000 with a cap of Rs250. in digital payments. In the 1990s the Korean government

Source: RBI, Morgan Stanley Research wanted to increase card transactions (away from cash) to increase tax compliance. This drove up tax compliance among businesses in Korea. Some of the steps taken by A likely effective move by RBI has been to differentiate between the government were as follows: physical POS and digital POS (UPI, QR and so on) based payments. For the non-physical payments, banks do not have to invest in the 1. The Korean government introduced tax incentive POS infrastructure and so on and hence MDRs should be much lower schemes to increase the use of credit cards. In 1999, it than with physical POS. While RBI has suggested a 10 bps gap introduced a new scheme which offered tax rebates if the between physical and non-physical, we would expect the downward total spend through credit cards, as a percentage of pressure on non-physical to intensify over the next few years. annual income, was above a certain threshold (25%). But this was a big driver for the surge in credit card usage. This will be driven by the fact that the cost of these transactions is 2. It also introduced a lottery scheme where the game relatively low. And also there has been a sharp increase in competi- was based on the invoice number of each credit card tion. PayTM has been aggressive in acquiring merchants and is transaction. There were weekly winners with both the charging no MDR. Since November 2016, it has acquired almost 5 mil- consumer and the merchant (at whose business this lion merchants (this compares with 2.8 million merchants who transaction was conducted) winning a prize. accepted card-based payments as of July 2017). Although these are 3. The government also made it mandatory for businesses for transactions on its wallet - not interoperable - the surge in mer- to accept card payments if the consumer wanted to pay chant acquisition will likely cause incumbent players to revisit the with a card. MDR charged on digital networks or stand to lose share to fintech Exhibit 22: alternatives. Korean Experience: Credit, Debit and Prepaid Card 6. Access to credit - As we discuss in detail in the next section, the Transactions as a Percentage of GDP Rose Sharply data trail from increased usage of digital technology should have a Following Introduction of TIETP in 1999 Incentivizing lot of upside for both merchants and consumers. Lenders will be able Wage and Salary Earners to Use Credit Cards to use this data to map their cash flows better. This will enable 45% Credit Card Debit Card Prepaid Card increased lending to the smaller borrowers at reasonable rates 36% (given improved underwriting). This should also help increase 27% acceptance of digital payments by a significant amount.

18%

0% 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: World Bank Group Report, Bank of Korea

MORGAN STANLEY RESEARCH 23 M BLUEPAPER There were other factors too - like banks' willingness to Exhibit 23: lend aggressively on credit cards after the Asian crisis. Digital Payments (Percentage of GDP) Credit card transactions grew ~9x in absolute terms and Digital Transaction Values / GDP 25% from 4.9% to 34.3% as a proportion of GDP between 20% 1999 and 2002. Korea extended the tax scheme to 20% transactions on debit cards also, which saw a surge in 15% debit card based transactions. 10% There was a credit card crisis in Korea after 2002, given 6% 5% 5% 3% 3% the sharp loan growth. But for India, the government can 2% 2% 2% 1% 1% 1% 1% 1% 1% 1% 1% offer tax incentives targeted towards digital transactions 0% linked to existing savings accounts (debit cards, UPI, pre F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F1Q18 paid wallets and so on). The introduction of tax incentives F2027E Source: RBI, NCPI, Morgan Stanley Research (E) estimates drove up the amount of debit card transactions. Given the plethora of digital payment options available in India Exhibit 24: now, such incentive schemes can act as a catalyst to Digital Payments (Percentage of Personal Consumption Expenditure) quicker digital adoption. Digital Transaction Values, as % of Private Consumption 40% 36%

We expect the value of digital transactions to increase over 10 32% years by ~12x in our base case and ~18x in our bull case 24%

Total digital payments as a percentage of GDP was ~5% in F2017, up 16%

10% from ~3% the prior year. As a proportion of total personal consump- 8% 8% 5% tion (PCE) this stood at ~8%, well below those of other EM countries 4% 4% 2% 3% 2% 3% 3% 3% 1% 2% 2% 2% such as Brazil (30%), South Africa (27%) and Russia (16%). Given all 0% the changes afoot, in our base case we expect this to increase to F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F1Q18 ~20% of GDP by F2027 - which would imply a CAGR of 30% for the F2027E Source: RBI, NCPI, CEIC, Morgan Stanley Research (E) estimates next 10 years. This would take digital transactions as a percentage of PCE to ~36%. Exhibit 25: Is this a big stretch? We don’t think so. In fact if the government is Digital Payments to Shift to Non-card Transactions completely successful in pushing its digital agenda, this target will Credit Cards POS Debit Cards POS Ex Rupay M Wallet Rupay Cards POS UPI 100% likely be achieved much more quickly - as we mentioned earlier, its 15% 16% 22% 17% 18% 22% 30% 34% 80% 36% 37% 38% target is to increase digital transactions 6x in this year alone. If suc- 38% 37% 42% cessful, this can cause digital transactions increase to ~30% of GDP 41% 60% by F2027 (a CAGR of ~35% over the next 10 years). 85% 84% 40% 78% 83% 82% 78% 70% 66% 64% 62% 61% 59% 57% Who will be the winner(s) in digital payments? 20% 46% 43%

0% We expect a continued shift in market share away from current F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 leaders - Historically credit cards were the main form of digital pay- F1Q18 ment, but over the last few years debit cards have caught up. Till Source: RBI, NCPI, Morgan Stanley Research 2015, credit and debit cards made up almost 100% of retail digital transactions. Visa and Mastercard were the main payment gateways used for digital transactions.

24 M BLUEPAPER However, this is changing. Mobile wallets and UPI made up ~11% of Exhibit 27: total transactions in F1Q18. Within cards also Rupay card (which is a Mobile Wallets and UPI as Percentage of Retail Digital Transactions; domestic card payment gateway launched three years ago) is gaining Both Are Growing, With UPI Catching Up Quickly (It Was ~40% of Wallet ground and accounted for ~10% of total debit card transactions at Transactions in F1Q18) 15% POS terminals in F1Q18. Mobile Wallet UPI

12% In our view, the alternative payment avenues - UPI, wallets and Rupay 3% debit cards - are likely to be the big drivers of the expected sharp 9% increase in digital payments. We expect the share of these new 1% players/payment systems to increase from the current ~15% to 6%

8% 65-70% over the next 10 years. In fact, we will not be surprised if the 7% 3% pace of the market share shift is even quicker. 5% 3% 0% 1% 0% 1% While we are building in the market share decline for the incumbents F2012 F2013 F2014 F2015 F2016 F2017 F1Q18

- Visa / Mastercard - we still expect them to show reasonably good Source: RBI, NCPI, Morgan Stanley Research growth. The launch of Bharat QR, which is interoperable and does not need POS terminals, should help them grow despite transactions However, unlike China where wallets dominate the digital payment, moving to mobile-based platforms. The table below shows that even in India banks should do well. Mobile wallets have the first mover if Visa / Mastercard market share reduces to 35%, they will still likely advantage, but we expect banks to catch up. The reason is that, unlike grow at a CAGR of ~16%. in China, regulators in India have pushed for a real-time digital payment infrastructure set-up before wallets became ubiquitous. Within non-card payments, who will win - wallets or bank-backed sys- Hence, while wallets will grow, it is unlikely that all of them will tems? prosper. There will be a handful of winners, with the others being consolidated by larger players or banks (for instance in August 2017 Globally tech companies have led the digital wallet space. In both the Axis Bank bought out Freecharge, one of the bigger wallet players). US and China, traditional banks have been virtually non-existent in the digital wallet space. In the US, PayPal leads the digital wallet Over the last few months, the value of digital payments in India has space. In China, Alipay leads the space followed by Tenpay. Tech com- increased sharply, with digital retail transactions up ~36% in F1Q18 panies have been able to establish a dominant position given their (annualised) compared with F17 - this is despite F1Q being the unique value proposition of managing card-not-present (CNP) risks weakest quarter seasonally. Within this, mobile wallet transactions for the merchants and a simplified checkout process for the con- are up ~50%. UPI on the other hand has moved from negligible to sumers. ~3% of total digital transactions. This level is likely to rise fairly quickly as UPI is rolled out to consumers and merchants.

Exhibit 26: Sensitivity of Visa / Mastercard Share and Ten-Year CAGR Current Digital Payments in India, F2017, USD Bn 107 V/MA Share 90.0% MS Estimate on 2027 Digital Payments Market Size, USD Bn 1208

V/MA's share in 2027, scenario 10% 25% 35% 50% 75% 90% Implied CAGR 2.3% 12.1% 15.9% 20.1% 25.1% 27.4%

Source: RBI, Morgan Stanley Research estimates

MORGAN STANLEY RESEARCH 25 M BLUEPAPER GST Is Arguably India’s Most Important and handle tax administration functions such as registration approval, assessment, audit, and so on. For taxpayers, GSTN provides Reform Since the Early 1990s common and shared IT infrastructure and caters to functions facing taxpayers, such as filing registration applications, filing returns, cre- GST, with its ability to clean up India's messy indirect taxation and lift ating challans (invoices) for tax payment, settling IGST payment (like government revenues, has the potential to boost growth. From the a clearinghouse), and generating business intelligence and analytics. corporate sector's perspective, warehousing costs, freight costs, and Checking Input Tax Credit (ITC) claims is one of the important aspects inventory levels could all fall. Together with a more transparent and of GST, for which data of business to business (B2B) invoices would efficient input tax credit system, this should lead to improved profita- be uploaded by taxpayers and matched by GSTN. The GST system has bility. Over the next two to three years, large companies could also a G2B portal for taxpayers to access the GST Systems. gain share from MSMEs, whose hidden tax subsidy (if they never paid taxes earlier) is likely to vanish under GST. That said, it is not all bad What advantages does GST bring to India? news for MSMEs. Their entry into the formal economy, in addition to the transparency of the GST system, will enable flow-based lending GST implementation is a likely paradigm shift in the economy and to these entities, eventually lifting their growth as well as India's ratio corporate sector. The tax change aims to weed out the cascading of aggregate credit to GDP. effects of taxes on cost at the aggregate level and thus bring down the cost of doing business. While currently there are several tax rate What is GST? slabs, it is quite possible that in the future the number of tax rates will be reduced. We outline below the impact on the corporate sector: The goods and services tax (GST) implementation marks a major reform in India, as the initiative will transform India's indirect tax land- A single tax will lead to centralization and consolidation of ware- scape. GST was launched on July 1, after a prolonged discussion in housing facilities for businesses: The removal of inter-state tax bar- political spheres and several rounds of amendments to the draft law. riers, combined with seamless input credit, will make India one GST has subsumed all indirect taxes (VAT, excise, cenvat, entertain- common market for goods and services, leading to economies of ment, octroi, service and so on) and done away with India’s old scale in production and improved efficiency in the supply chain. GST system of multiple central and state taxes by replacing it with a single implementation implies lower warehousing costs and smoother tax system. GST is a tax on consumption and will weed out the cur- movement of goods between states. rent series of cascading taxes on the production chain. GST implementation will gradually shift the informal sector into the formal Productivity gains for the transportation sector: With the giving sector, leading to greater compliance and higher tax collection for away of entry or state border taxes, freight movement is likely to the government. Compliance also improves because GST is com- speed up, leading to improved cost and time efficiency for the sector. pletely digital and an online platform – arguably the largest digital The removal of inter-state barriers is important as roads account for tax network in the world. The scale of operations is staggering. The almost three-quarters of freight traffic in India. However, efficiency GST network is expected to manage 30-40 billion transactions annu- remains low in the road transportation sector. In a day, trucks in India ally paid by almost 10 million tax-paying entities, which implies drive just one-third of the distance of trucks in the US (280 kilome- 100mn transactions daily. ters versus 800km) as only 40% of the time is spent in driving. Thus on overall costs, inter-state trade costs in India exceed inter-state GST’s digital infrastructure trade costs in the US by a factor of six. Eliminating checkpoint delays could keep trucks moving almost six hours more per day, equivalent India’s central government has set up a Special Purpose Vehicle – to 164km per day. GSTN – to manage the IT infrastructure under GST. GSTN is the front-end IT ecosystem under the GST network and will be an impor- Inventory management to grow more efficient: The net effect of tant part of GST compliance with respect to filing data. GSTN’s man- reduced warehousing and reduced transportation time is that inven- date is to establish, develop and manage the required infrastructure, tory levels will likely decline. In the short run this hurts growth as systems, technology, partnerships, and eco-system for the imple- well as credit formation (with a one-time decline in both). Beyond the mentation of GST. Thus, it enables a uniform interface between the first year, however, it will reduce costs for Corporate India. taxpayer and the government (both central and state). The IT systems of the central and state tax departments function as back-ends

26 M BLUEPAPER Claiming input credit for GST paid on services: Under GST, companies izes India's taxation, while abolishing the Planning Commission in can set off taxes paid on inputs. For sectors with high competitive 2014 had already set the stage for decentralization of expenditure. intensity, input credit would likely be passed on to the consumer This shift, in our view, could prove cathartic for Indian government while in cases where competition is not efficient or the industry is finances. India’s ratio of tax revenues to GDP is on the low side as fragmented, companies could experience margin expansion. compared to other EMs – and this has been a key reason why the fiscal deficit has been relatively high on a comparative basis. The Formal sector to gain market share from informal sector owing to tax implementation of GST will increase tax compliance amongst corpo- rationalization: The biggest advantage that the unorganized sector rates, leading to a boost in tax revenues even while the tax rates are enjoys currently is the tax arbitrage (due to being outside of the tax in and of themselves revenue-neutral. Over time, as tax revenues ambit). GST aims to integrate the informal sector into the formal improve, it should lead to a improvement in fiscal deficits and public sector over the medium term. The incentive of availing input credit debt to GDP trends in India. Niti Aayog (India's policy think tank) on supplies would force the organized players to deal with suppliers expects tax revenues to grow 50% over the next three years. Indeed, that are registered with GSTN. In the GST regime, it will be almost we estimate that if the primary fiscal deficit stays at 1.2% of GDP (as mandatory to pay taxes because the entire supply chain will now compared with an average of 2.1% over the past five years and 1.6% have to be accounted for in the tax chain. The key rationale is that in F2017), the ratio of public debt to GDP should decline to below firms up the value chain would only contract with those companies/ 60% by F2027. dealers whose taxes they can set off. This will encourage more companies from the informal sector to join the formal sector. As the Digitization will help bring down expenditure too. This will help the unorganized sector gets into the formal system and thus gets intro- government conduct welfare spending in a much better way – duced into the tax system, the tax arbitrage (their source of competi- quicker and more efficient implementation (with little leakage). The tiveness) is expected to abate, benefitting the formal sector and government has already implemented a direct benefit transfer (DBT) allowing it to gain market share. The organized sector has a further program, which is a social welfare scheme under which subsidy bene- advantage from lower warehousing costs, freight costs and inven- fits are directly transferred to the beneficiary’s bank account rather tory levels. We highlight industries and stocks from our coverage than via the issuance of checks / cash payments / rebates. Cumula- that we believe will benefit from this shift into the formal economy. tively, the total direct benefit transfer so far has been approximately Rs2 trillion and savings have been ~Rs500bn. GST enabler for credit to the unserviced: Another big change that GST could potentially drive is credit to the unserviced segment. Out What are the challenges? of the eight million entities that are potential users of the GST network, 65-75% are small and micro businesses that do not have easy Despite the long-term positive implications, there could be short- access to bank credit. GSTN could act as an enabler for flow-based term problems: lending for these businesses as they form part of the GST network. With GST coming on board, tax filing would move from a back-end Growth impact: The one-time inventory adjustment has already activity to a front-end activity. The process of GST would require caused some hiccups to growth. filing of returns thrice a month (36 times a year). With the help of ASPs and GSPs, the government expects the filing process to be Is India digitally ready? GSTN is fully digital but not all businesses in simple and convenient. GST records would have a full integration of India are equipped to undertake digital transactions. This has the purchases and sales of GST users. This information could be used by potential to create disruptions. businesses as a means to avail working capital loans. This in our view could lead to unserviced entities coming into the credit system, Informal sector set for a shock: The 'subsidy' to the informal sector in thereby boosting credit to GDP and the growth of these businesses. the form of tax evasion goes away, with a likely concomitant impact on jobs. The informal sector creates four out of five jobs in India. How The Government Gains Too long this may persist is difficult to estimate. There could be associated supply chain disruptions that could last for several months. GST completely alters the way government finances are managed in India. Hitherto, tax collection in India was decentralized while expenditure allocations were centralized through the Planning Com- mission, which set expenditure priorities for the states. GST central-

MORGAN STANLEY RESEARCH 27 M BLUEPAPER Implication 1: GST Plus Digital Payments = Significant Big Data Opportunity

Key Takeaways

1. Lenders (banks / NBFCs / fintech companies) will get access to cash flow data of MSMEs (micro, small and medium enterprises) through multiple sources - for instance, digital payment platforms but more importantly GSTN (the Goods and Services Tax Network). This will likely help them get access to high-quality, high-frequency data that will allow them to lend to this sector. While MSME is a big part of the economy (over 30% of GDP), formal system lending to the MSME industry is just 10% of GDP - we expect MSME lending to have a ~17% CAGR over the next decade. 2. Consumer loans is the other sector that is likely to do well. As data availability becomes better and quicker, banks will be able to roll out new products (for instance HDFC Bank's "loan in 10 seconds"); in our view this, along with fairly strong economic growth, will help drive consumer loan growth at a 17% CAGR over the next decade. Even after this growth consumer loan penetration will be low at ~25% of GDP. This growth in MSME and consumer lending will in turn boost economic growth and discretionary spending.

Exhibit 28: Material Shift in System Loan Mix Ahead: Consumer and MSME to Replace the Lower Share of Large Corporates Mix (%) CAGR (%) As % of GDP F2017 F2027e F2017-27e F2017 F2027e Total Loans (Rs Bn) 102,524 344,681 12.9% 67% 78% Consumer 24% 32% 16.4% 16% 25% Housing 14% 19% 16.0% 10% 15% Non Housing 9% 13% 17.0% 6% 10% Non Consumer 76% 68% 11.6% 51% 53% Large Industry 29% 17% 6.8% 20% 13% MSME Industry (Banks) 5% 7% 17.0% 3% 5% MSME (NBFCs) 4% 5% 17.0% 2% 4% Services (Banks, including MSME**) 18% 20% 14.4% 12% 16% Agri & Others 21% 19% 11.7% 14% 15% Memo: Overall MSME 15% 22% 17.0% 10% 17%

Source: RBI, NHB, Company data, Morgan Stanley Research (e) estimates. ** Of bank credit to service sector, we have assumed 40% is towards MSME. NBFCs and HFCs lending to MSME are Morgan Stanley estimates

28 M BLUEPAPER

Exhibit 29: Exhibit 30: India Loans: Corporate Loans as Percentage of Total India Credit Multiplier at Multi-decade Lows 40% 3.0

36% 33% 33% 32% 33% 2.0 32% 32% 31% 31% 29% 28% 28% 28% 1.0

24%

20% 0.0 F1970 F1972 F1974 F1976 F1978 F1980 F1982 F1984 F1986 F1988 F1990 F1992 F1994 F1996 F1998 F2000 F2002 F2004 F2006 F2008 F2010 F2012 F2014 F2016 FY08 FY09 FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17

Source: RBI, Company Data, Morgan Stanley Research Estimates Source: RBI, Morgan Stanley Research. The above is based on banking system loans

The past 15 years for banks have been good - Indian financial system 3. SOE banks still make up ~55% of total loans (banks and NBFCs) in assets have grown at a CAGR of 16%. There was a period of strong the system and, in our view, they cannot grow at more than low single growth (F2002-F2012) and then a period of fairly lackluster growth digits for several years. We expect SOE banks to lose share at the rate (F2012 onwards given the NPL issues faced by the corporate sector). of ~3-4% a year for the next few years. However, in the near term Overall loan penetration went from 31% of GDP to 67% of GDP given that they make up ~55% of loans, their weakness will drag during this period. down system growth numbers.

Corporates were the big driver (first of an acceleration and then of a But, what we expect over the next 10 years is a material shift in deceleration) in the past 10 years - First, corporates levered up to credit mix for India's lenders. We expect consumer loans to remain create capacity - both in infrastructure and industries such as steel - big drivers of growth and expect them to increase at a CAGR of ~17% and then hit a bottleneck on profitability and have been delever- to Rs. 112 trillion over this period. The other big driver, in our view, will aging. This caused loan growth to be strong for three or four years be the MSME sector. Banks have been reducing exposure to this seg- and then become subdued for the past three years. ment over the past few years given weaker underwriting and growth in other businesses. But as data availability on these businesses Our view is that banks' and NBFC's growth in terms of loans to corpo- improves, we expect a strong pickup. We expect MSME loans to grow rates will be muted (even when the economy picks up, we expect at a CAGR of ~17% (compared with low double-digits in the past these corporates to access bond markets for funding given RBI's reg- decade). ulations on exposures to large corporates). The sharp slowdown in corporate loans and weakness in most other lending segments Historically, banks were unable to cater to vast segments of the (excluding consumer loans) have caused the credit multiplier in India economy fully. The bulk of the banking focus has historically been on to be <1 - this is a multi-decade low. the elite, the affluent and the middle class. The economics of expanding distribution and the relatively low value of transactions If corporate demand will be relatively weak, then what will drive asset from the strata below the middle class has meant that access to growth for the lenders? We expect loan growth in India to average banking services has been tougher for this part of the population. around 13% over the next decade. F2018 and F2019 will likely remain Like other parts of the financial system, the segment below the slow for three reasons: middle class has been left relatively unaddressed by banks.

1. We think the economy will continue to pick up from relatively low We expect this to be the big driver of loan growth - in both the con- levels, implying fairly muted demand for credit from the non-consumer and MSME segments. As the economy grows, more of these sumer segment. families will become middle class or affluent. Also, the digital initiatives will help the financial system to get access to more data on 2. The near-term impact of GST could cause some slowdown in the underlying cash flows for these areas. This will allow lenders to overall economy / loan growth for a few quarters. underwrite businesses on the basis of cash flows.

MORGAN STANLEY RESEARCH 29 M BLUEPAPER

Exhibit 31: India's Banking Pyramid – The Banked, Underbanked, and Unbanked

Source: India Software Product Industry Round Table (iSPIRT)

AlphaWise - In our survey also, almost half of the respondents said that they don’t have access to formal credit. The biggest source of funds for these businesses is own funds or funds from family / friends. For the smaller businesses which can get formal credit, the biggest component is Loan Against Property (LAP), which is obviously a collateral backed loan. Another takeaway was that businesses with less share of cash in revenues get easier access to credit.

Exhibit 32: Exhibit 33: Sources of Formal Credit for Businesses Who Has Access to Formal Credit Usage Share Access to formal credit Share of formal credit

25% Bank Limits / Credit / Overdraft facility 35% 12% Share of cash - 76%+ 26% 25% Loan against property 11% 56% Share of cash - 51-75% 15% 32% Loan against order 6% 61% 5% Share of cash - 0-50% Loan against investments 37% 4%

Source: AlphaWise, Morgan Stanley Research Source: AlphaWise, Morgan Stanley Research

30 M BLUEPAPER Why does data matter? Exhibit 34: Large Private Bank: Personal Loans for a Similar Profile Borrower but Historically, India has been a data-poor country from the perspective Working at Different Firms of credit underwriting. Until about a decade back, there was no credit 25.0% 19.8% bureau for consumer loans, which exacerbated the problem of NPLs 20.0% in consumer lending, especially unsecured loans in 2008/2009 15.0% 14.0% 12.0% (along with unfettered growth without proper credit assessment). 11.0% On corporate lending too, banks did not have proper data about mul- 10.0% tiple borrowings by the same corporate across banks. This accentu- 5.0% ated the problem with corporate NPLs that we have seen over the 0.0% last five years. Employee of a large IT Employee of a Mid Sized Employee of a Employee of a MSME company Bank Manufacturing Firm Company CAT A CAT B CAT C CAT D However, as proper data has become available, underwriting deci- Source: Bank Bazaar, Morgan Stanley Research. Note: Interest rates are on an annualized basis sions have become much better. An example would be consumer lending and CIBIL [Credit Information Bureau (India) Limited], the consumer credit score provider. This was set up in the mid 2000s and Exhibit 35: became increasingly important after the consumer credit cycle in HDFC Bank: YoY Growth in Unsecured Personal Loans; Strong Pickup 2008-2009. Now, it is the primary data provider for consumer loans Since 2015 - Now Makes Up ~10% of Group Loans in India (with over 2,400 members including banks, NBFCs, HFCs 75%

[housing finance companies] and so on, and maintaining credit 60% records on over 550 million individuals and businesses). This has allowed CIBIL to generate proper credit scores on retail borrowers 45% and information that banks can access. 30%

This improvement in data allows banks to 1) lend properly to partic- 15% ular areas - like unsecured - and 2) have differentiated pricing. For 0% instance, if we look at data for a large private bank, there is significant Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 Jun-15 Jun-16 Jun-17 divergence in interest rates offered on the basis of credit profile of a Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 Dec-15 Dec-16 customer. Historically, this was tough to achieve especially for a bank Source: Company data, Morgan Stanley Research with a weaker data base.

The third advantage relates to cost of credit delivery. Both consumer and MSME loans require a significant number of people selling the products, collecting data and so on. However, as on-tap data availability increases, lenders can start offering these loans electronically. This will bring down sourcing costs very sharply.

An example of a product piggybacking on this increased data availability in consumer segment has been HDFC Bank's unsecured loans in ten seconds.

This essentially allows existing customers of HDFC Bank to be able to borrow unsecured personal loans in less than ten seconds. This enables the bank to increase volumes and reduce costs.

In our view, this will be the template for more loan products (consumer and MSME) over the next few years, across banks.

MORGAN STANLEY RESEARCH 31 M BLUEPAPER

Exhibit 36: Timeline of Steps Taken in Setting Up Digital Infrastructure and Policies That Will Help Drive MSME Loan Growth

Source: Morgan Stanley Research

32 M BLUEPAPER Surge in Data Lies Ahead - Consumer and MSME will be a big driver of loan growth over the next decade for lenders MSME To Be the Big Growth Drivers MSME as a sector has to do well for the economy to pickup. This seg- With a large part of the loan book, i.e. large corporate loans, unlikely ment consisted of ~45 million enterprises as of March 2012 (source: to grow, lenders will likely search for new lending avenues. We Ministry of MSME) and provides employment to about 100 mn believe that consumer and MSME loans will be the big growth people (~10% of the country’s population). Also, as per the Fourth drivers. We expect digitization to cause a significant increase in the Census of the MSME sector (2006-07), the sector accounts for number of data sources that banks can tap into to lend across the about 45% of the manufacturing output and around 40% of the total retail and MSME sectors. exports of the country. An RBI Working Group report estimates that the economic output growth in this segment has consistently out- For instance, the use of digital payments leaves a data trail for banks paced overall GDP growth. to assess cash flows of smaller enterprises / traders and so on. This will allow the banks to lend to them at lower rates given the better However, despite the fact that it is a fairly critical part of the economy, data history. This will enable banks to make quick lending decisions. credit availability to the sector has been weak - Banks have traditionally shied away from this segment due to factors like a lack of proper GSTN is likely to be a big source of data for the MSME sector - accounting records and necessary documents, erratic cash flows, the As mentioned earlier, GST is the common indirect tax code for the higher cost of operations, the need for customization during disbur- entire country. The implementation of GST requires significant sals/repayments and past experience of higher delinquencies and, investment in the creation of IT infrastructure. This is the responsi- hence, credit costs. bility of GST Network (GSTN). It is the technology backbone for GST and will administer registration, invoice uploading, tax return filing In fact, if we look at bank lending to micro and small industry compa- and the tax payment system. The businesses will have to upload nies (where investment in plant and machinery is less than Rs 50 mn) invoices for individual transactions on the GSTN network. It will map it has been a fairly weak story for almost two decades. As a per- these invoices and then the taxpayer can claim a tax credit. The GSTN centage of bank lending, loans to these businesses contracted from expects to handle ~3 billion invoices every month. 14% in F2000 to about ~7% in F2010 and then ~5% in F2017. If we look at loans given to medium-sized industry companies (investment As the new tax system picks up there will be a significant impact on

However, if somebody wants to access a loan, they can allow lenders to have access to their tax filings in the GSTN. This will provide banks with significant cash flow information of the borrowers, thereby enabling them to price loans in a much better way with much more cer- tainty on likely credit performance.

We look at both consumer and retail loan markets and assess how big they can be in 10 years.

MORGAN STANLEY RESEARCH 33 M BLUEPAPER

Exhibit 37: Exhibit 38: Bank Credit to Micro and Small Industry Companies* as Percentage of Bank Credit to Medium Industry Companies as Percentage of Total Total Non Food Credit Non Food Credit and GDP Bank Credit to Micro & Small Industry as % of Total Non Food Credit Bank Credit to Medium Industry as % of Total Non Food Credit Bank Credit to Medium Industry as % of GDP Bank Credit to Micro & Small Industry as % of GDP 15% 6%

5% 12%

9% 3%

6% 2%

1% 3%

0% F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

* This does not include loans to MSME in the services sector. Source: RBI, Morgan Stanley Research F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

* This does not include loans to MSME in the services sector. Source: RBI, Morgan Stanley Research

Banks have been fairly averse to lending to this sector. Losses in a Exhibit 39: down-cycle in the economy have been fairly elevated. Also, given the NBFC Credit to MSME as Percentage of Total Credit (Banks and lack of data, most of these loans have been collateral backed lending NBFCs) to MSME Industry and not cash flow based lending. And with foreclosure of collateral 35% 29% not very easy, banks have shied away from this lending. 30% 26% 25% 23% 23% 24% 20% 20% Most of the lending to this space in recent years has been in the form 17%

15% 13% of Loan Against Property (LAP), which requires the borrower to 12% 11% pledge their own property (for instance their residence) to borrow. 10%

This field had been left open for NBFCs to tap into. While it is difficult 0% F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 to get proper NBFC data in terms of lending to this sector, we added Source: Company Data, RBI, Morgan Stanley Research Estimates up the 20 large HFCs' and NBFCs' exposure in these segments. This shows fairly sharp growth in lending towards these sectors, especially over the last five years. Their loans have increased over 10x in Exhibit 40: the last ten years and have increased to ~29% of total credit exposure SCUF's Lending Yields Minus AAA 1 Year Bond Funding Costs 24.0% (banks and NBFCs) to this segment, from ~11% a decade back. 22.0%

Why should it matter whether loans are coming from NBFCs or 20.0% 17.7% banks? 15.9% 15.7% 16.0% 14.4% 14.4% 13.6% 13.6% 13.5% 12.8% 13.3% One reason is that NBFCs are wholesale funded with limited leverage 12.9% 11.1% capacity - implying that they can never be able to meet the full 12.0% demand. Also, with banks absent, the interest rates charged by NBFCs can be very high. They are lower than in the informal sector 8.0%

- moneylenders - but still materially higher than those charged by F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 banks. For instance, if we look at the average yield earned by Shriram Source: Company Data, Bloomberg, Morgan Stanley Research. We have used enterprise loan yields of SCUF which is largely MSME but also includes two-wheeler and gold loans City Union (SCUF) on its entire book - its spread over 1 year AAA paper is almost 14%. Also, since F2012 (when banks were affected by NPL issues) this spread has expanded by >200 bps.

34 M BLUEPAPER

Exhibit 41: Capital Float's Key Value Proposition to an SME Borrower (Per the Company) Capital Float PSU Banks Private Banks Traditional NBFCs Money Lenders FAST Time to get funds <1 Week 4-6 Weeks 1-2 Weeks 1-2 Weeks <5 Days Apply anywhere, any time! Yes No No No No 10 minute online Hassle-free documentation & processing 3 months 1 week Doc intensive Unregulated application FRIENDLY Zero Collateral Yes No Maybe Maybe Yes Minimum years in business 1 year 3-5 years 3-5 years 3-5 years Varies More inclusive for 'New-Age' business Yes No No No Maybe AFFORDABLE Loan Tenure 1-12 Months Not less than 1 year Not less than 1 year Not less than 1 year <1 year Preclosure Penalties None 5% of loan 5% of loan 0-5% of loan Varies Repayment options Flexible EMI only EMI only EMI only Weekly, Monthly Advertised interest rate 18% - 24% 14% - 16% 14% - 22% 16% - 24% 24% - 120% Processing fees up to 2% 2% 2% - 3% 2% - 3% 2% Hidden charges None Varies 2% Insurance 2% Insurance None

Source: Company Data, Morgan Stanley Research

In our view this will change over the next 10 years as data availa- 2. Refine models - When CIBIL was launched, it took three to four bility of these companies' cash flow becomes much broader and years for banks to be sure of data points they needed before they faster. This is a space where a number of fintech companies in India could lend on the basis of these models. Similarly, the models here are operating - for instance, Capital Float & Lending Kart. Capital will need to be refined, especially after one credit cycle. For instance, Float has various products, for instance supply chain finance loans for the data from GSTN will enable banks to capture data on sales on a SMEs. The borrowers who qualify can apply online within ten min- monthly basis. They will be able to identify any potential issues on an utes and get credit within three days at reasonably low rates (consid- almost real-time basis. ering their financing cost). Sharply lower lending rates in the MSME sector over the next Over time, we expect banks to start using these models to lend in the five years - The real time flow of data is likely to reduce riskiness of space. This in turn will likely cause a significant pickup in lending to these loans meaningfully. Also, the cost of underwriting will reduce the MSME sector by banks. This will not be easy and its unlikely that fairly sharply given the dependence on technology. All this will feed we see a big growth pickup for two or three years as formal lenders into lower pricing. It is tough to take a call on how much spreads can have to do the following: come down in this product - but we could easily see average lending rates declining by 300-400bp over the next five years. 1. Build lending models based on the incoming data - Although data sources will increase significantly, banks will take time to build Right now lenders are not sure of how bad credit costs can be as busi- models to lend on the basis of these new data points. nesses slow down. Given the lack of data, lenders cannot identify problematic loans until they have almost become NPLs. This makes them price loans at a fairly elevated level - resulting in high ROEs even in a normalised credit cost scenario. However, as data availability improves, lenders will be able to have more clarity on businesses on a much more frequent basis. Also, the cost of delivery will come down. In this scenario, we would expect lenders to cut pricing significantly and still earn ~20% ROE.

MORGAN STANLEY RESEARCH 35 M BLUEPAPER

Exhibit 42: Indian Lenders Can Cut Spreads on MSME Loans by ~300bp as Data Quality Becomes Better and More Reliable Present With gains from higher customer data MSME Dupont Bank NBFC Bank NBFC Interest Income 16.5 20.0 13.8 17.0 Interest Expense 5.0 7.3 5.0 7.3 NII 11.5 12.7 8.8 9.7 Fee Income 0.0 0.0 0.0 0.0 Total Income 11.5 12.7 8.8 9.7 Operating Costs 5.6 4.8 4.9 3.8 Operating Profit 5.9 7.9 4.0 5.9 Loan Loss Provisions 1.9 2.5 1.1 1.5 Standard Provisions 0.06 0.07 0.06 0.07 Profit before Tax 3.9 5.3 2.8 4.3 Tax 1.3 1.8 1.0 1.5 Profit after Tax 2.5 3.5 1.8 2.8 ROA 2.5% 3.5% 1.8% 2.8% Leverage 11.0x 7.1x 11.0x 7.1x ROE 28.0% 24.8% 19.8% 20.2% Loans 76 100 76 100 CRR 4 4 SLR 20 20 Interest Earning Assets 100 100 100 100 Equity 9.1 14.0 9.1 14.0 Applicable Risk Weights 100% 100% 100% 100% RWAs 76 100 76 100 Tier I Ratio 12.0% 14.0% 12.0% 14.0% Borrowings 90.9 86.0 90.9 86.0 Yield on Loans 20.0% 20.0% 16.5% 17.0% Cost of Funds 5.5% 8.5% 5.5% 8.5% Spread 14.5% 11.5% 11.0% 8.5% Yield on SLR 6.5% 6.5% NIM 11.5% 12.7% 8.8% 9.7% Fees / Avg Loans (Assumption) 0.0% 0.0% 0.0% 0.0% Operating Costs / Avg Assets (Assumption) 5.6% 4.8% 4.9% 3.8% Standard Provisions / Loans (Assuming 20% loan 0.1% 0.1% 0.1% 0.1% growth) Loan Loss Provisions / Avg Loans (Assumption) 2.5% 2.5% 1.5% 1.5% Effective Tax Rate 34.6% 34.6% 34.6% 34.6%

Source: Morgan Stanley Research. Note: The above calculations are made on an asset book of 100 units of a given currency. For MSME loans by banks, risks weights can be lower at 75% subject to certain conditions, making these loans more profitable. However, for this exercise we have assumed a 100% risk weight.

A stronger economy and better access (quicker and at lower rates) As we discuss in the economics section, better credit availability is a to credit should help the MSME sector to do very well, in our view. big driver to MSME profitability and investment. China saw this We expect these loans (from banks and NBFCs) to grow at a CAGR transformation, and if our thesis plays out, we will likely see this of 17% over the next ten years and increase to ~22% of total loans. sector as a big driver of India's economic growth. Some of this growth will also be taken up by the fintech based lenders (we consider them to be a part of the NBFC book).

36 M BLUEPAPER

Exhibit 43: Case Study: Role of SMEs in China's Consumer Loans to GDP 18% development 15.9% 15% 13.5%

SMEs have played a crucial role in driving economic 12% development in China, primarily in two areas: first, job 9% creation, in particular, during periods of significant structural transitions, and second, the promotion of 6%

industrial upgrading (see Cunningham [2011], SMEs as 3% 3.8% motor of growth: A review of China’s SMEs development 0% in thirty years, 1978–2008). F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

As a major driver of employment growth, SMEs allowed Source: RBI, NHB (National Housing Board), Company Data, Morgan Stanley Research Estimates for the swift absorption of agricultural labor during the country's initial industrialization process in the 1980s, Exhibit 44: with the blossoming of township and village enterprises. Consumer Credit Multiplier and Total Credit Multiplier Consumer Credit Multiplier , 3YMA Total Credit Multiplier The OECD estimates that Chinese SMEs accounted for 5.0 60% of GDP in 2013, and 68% of China’s total exports in 4.4 4.0 2011. SMEs' operational flexibility and small-scale have meant that they can better cater to the diverse 3.0 2.3

outsourcing demands from foreign firms. 2.0 1.9 Access to credit has been vital to the development of 1.0 0.8 China’s SME sector. SME loans account for 57% of GDP 0.0 (broadly similar to their economic size), and 39% of total

loans in China as of 2Q17. According to the World Bank F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 Source: RBI, NHB, Company Data, Morgan Stanley Research Estimates enterprise survey database, only 2.9% of firms in China report having difficulty in accessing credit, as compared with 26.2% globally and 15.1% in India. Moreover, the loan We expect consumer lending to grow at a CAGR of ~17% over this rejection rate in China is also materially lower, with only period and to increase to around 25% of GDP. This would imply retail 6.6% of SME loan applications turned down, versus 11.5% loans will increase by around 5x to US$ 1.5trillion. globally and 12.9% in India in 2014.

Various factors are driving this move.

Consumer Loans - Likely to grow at a CAGR of ~17% over the next 1. Economic growth and likely sharp increase in per capita GDP - We 10 years expect India's GDP in F2027 to be ~US$6.0 trillion and per capita GDP to be ~US$4,094 (~ INR 330,000) in nominal terms. This would While loan growth has slowed across segments, consumer loans imply an INR CAGR at ~11%, compared to ~12% over the last 10 years. have performed very well. Since 2000, these loans have increased On a PPP adjusted basis, we expect per capita GDP to increase to from ~4% of GDP to around 16% of GDP in F2017 - a CAGR of 23% ~US$16,800 from the current ~US$7,150. If India matches other over 17 years. There was a pause during the GFC when consumer countries' consumer loan penetration versus PPP GDP, we should credit to GDP declined as unsecured loans saw meaningful NPLs - but see consumer lending to GDP reach around 30% by F2027. even then in absolute terms consumer loans grew at a CAGR of ~13% over that three-year period. If we look at the credit multiplier in India, consumer has been reasonably strong at around 1.5x for the last few years.

MORGAN STANLEY RESEARCH 37 M BLUEPAPER

Exhibit 45: Exhibit 46: India Per Capita GDP Should Rise 3x in the Next Ten Years Consumer Credit to GDP Versus per Capita Income Nominal GDP Per Capita (Rs) 80% Malaysia 350000

300000 60%

250000 China Korea 200000 40%

150000 Thailand

100000 20% India by C2026e/F2027e India 50000 Indonesia 2017 2017 Consumer Credit/ GDP

0 /F Philippines 0%

2016 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 C F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

F2018E F2019E F2020E F2021E F2022E F2023E F2024E F2025E F2026E F2027E C2016/F2017 Per Capita Income (PPP, USD)

Source: RBI, World Bank, Morgan Stanley Research (e) estimates Source: IMF, Central Bank Data, RBI, NHB, Company Data, Morgan Stanley Research (e) estimates

Exhibit 47: Multipronged Push by the Government to Improve Housing

RERA to bring greater transparency in real estate transactions and RERA empower buyers Restrictions on end use of homebuyer funds, penalties for material deviation from the building plan and penal interest / imprisonment for delays in construction, should all result in better buyer sentiment and protection.

Better profit margins on affordable housing projects Tax Sops for Developers for 100% Profit-linked income tax exemption for promoters/developers of affordable housing Affordable Housing scheme on projects getting approved from June, 2016 to March, 2019 and completed in 3 Projects years from the date of approval. In addition, real estate developers will also get tax relief on unsold stock as tax liability will arise on capital gains when the project is completed.

Priority lending status for affordable housing Infrastructure Status to Affordable Infrastructure status to affordable housing will provide cheaper sources of finance to Housing developers and also open up additional avenues for developers to raise funds. The carpet area cap for affordable housing has also been increased to 60 sq. mt. instead of 30 sq. mt. earlier.

Rs290.4bn budgeted overall for FY2017-18; Rs60.4bn for urban housing Budgetary support and Rs230bn for rural housing to Pradhan Mantri The government has announced various schemes to boost supply and demand in affordable Awas Yojana - housing. Prominent scheme on the demand side is the credit linked subsidy scheme which Housing for All by involves an upfront subsidy for first time home buyers up to Rs0.26mn across Economically 2022 weaker sections (EWS), Low income groups (LIG), and Miiddle income groups (MIG1 & MIG2).

Source: Morgan Stanley Research

2. Government push for mortgages - The government is focussed on We are building in a 16% CAGR for banks and HFCs over the next ten increasing the stock of affordable housing in India as this can have a years and are forecasting outstanding home loans to grow from significant trickle-down effect on the economy. In 2017, the govern- ~US$220bn as of FY17 to ~US$905bn as of FY27. This should result ment has announced significant incentives to boost both supply and in mortgages to GDP moving from ~10% to ~15% during the period. demand in housing. This coupled with the increase in income levels will likely drive mortgage growth.

38 M BLUEPAPER

Exhibit 48: Exhibit 49: Mortgage Penetration in India is Quite Low Compared with that in Many Mortgage Penetration (i.e. Mortgages / GDP) Relative to Per Capita Developed / Developing Countries GDP 100% 50% 90%

Malaysia 80% 40% 68%

64% 2016 60% Korea 47% 47% 30% China 42% 41% 40% 32% 27% 26% 20%

20% 13% 10% Thailand 10% India Mortgage Penetration, Penetration, Mortgage 0% Philippines Indonesia UK USA India

China 0% Japan

Thailand 0 10000 20000 30000 40000 Malaysia Denmark Germany Singapore Hong Kong Hong Per Capita GDP (PPP, USD), 2016 South Korea South

Source: Countries' central banks, RBI, NHB, CEIC, IMF, Morgan Stanley Research Source: Countries' central banks, RBI, NHB, CEIC, IMF, Morgan Stanley Research

3. The other big driver of consumer loans will be loans that are Exhibit 50: higher up in the risk spectrum - The area where banks have been Proportion of Maruti Customers Using Finance to Purchase a Vehicle more aggressive on using data mining to lend has been the unsecured consumer loan space. Given the data banks have (CIBIL, transaction history and so on), they have started offering the following: a) Quicker processing: Now lenders can offer these loans with a much quicker turnaround. For instance, HDFC Bank's loan in ten seconds essentially allows it to credit a customer's account with the loan amount within ten seconds of application. b) Significant operating leverage: Since most of the processes - application, underwriting and actual credit delivery - happen online, the cost of originating these loans is fairly low.

This has caused banks to cut unsecured retail lending rates quite sig- Source: Maruti Suzuki nificantly over the last few years. For instance, for a good borrower unsecured personal loans are being offered at ~11-12%, compared with 16-17% which was the norm three to four years ago. Exhibit 51: Proportion of Buyers Using Finance to Buy Two-Wheelers; Likely To We are also seeing households using more leverage to buy discre- Increase Sharply 38% tionary items. For instance, the proportion of buyers using leverage 37% 2W Finance Penetration (%) to buy cars has increased from 72% in F13 to 80% in F17, and we 36% expect this to continue to increase. Other segments like two- 34% wheelers will also see a reasonably large pick-up in financing, in our 32% view. This again is driven by banks' willingness to lend at a price that 30% borrowers can afford. 28% 27% 26% 26% 26% 26% 25% 24% 24% 23% 23%

22% F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

Source: Crisil, Morgan Stanley Research

MORGAN STANLEY RESEARCH 39 M BLUEPAPER

Exhibit 52: Consumer Loan Mix Across Key Loan Products and Forecasts Rs Billion Mix F2012 F2017 CAGR F2027e CAGR F2012 F2017 F2027e Housing Loans 6,193 14,754 19% 65,343 16% 56.8% 60.8% 58.8% Banks 3,971 8,601 17% 29,320 13% 36.4% 35.4% 26.4% HFCs 2,222 6,153 23% 36,023 19% 20.4% 25.4% 32.4% Non Housing Loans 4,701 9,519 15% 45,694 17% 43.2% 39.2% 41.2% Banks 3,858 7,599 15% 32,142 16% 35.4% 31.3% 28.9% NBFCs 843 1,920 18% 13,552 22% 7.7% 7.9% 12.2% Non Housing Loans (Product wise) Auto Loans (Banks only) 891 1,705 14% 7,212 16% 8.2% 7.0% 6.5% Credit Cards (Banks only) 204 521 21% 2,205 16% 1.9% 2.1% 2.0% Other Retail Loans 3,606 7,292 15% 36,277 17% 33.1% 30.0% 32.7% Banks 2,763 5,373 14% 22,725 16% 25.4% 22.1% 20.5% NBFCs 843 1,920 18% 13,552 22% 7.7% 7.9% 12.2% Total 10,893 24,273 17% 111,037 16% 100.0% 100.0% 100.0% As % of GDP 12.5% 15.9% NA 25.1% NA

Source: Company Data, RBI, NHB, CSO, CEIC, Morgan Stanley Research (e) estimates. Note: Numbers for NBFCs and HFCs are based on our assumptions

Are these very aggressive assumptions? In our view, these are not Exhibit 53: aggressive as even after building in this growth, we are likely to end Annual New Loan Originations in Consumer Sector Likely To Be ~9% of up with consumer credit to GDP at ~25% by F2027 compared with GDP by F2027 and Grow 4.5x to Rs 40 trillion by F2027 ~16% now. As mentioned above, given that lenders are likely to focus Retail Disbursements (Rs Bn) 45,000 on consumer / MSME loans for growth and also continued improve- 40,000 ment in technology / data collection, this should be achievable. 35,000

30,000 16% CAGR Another way to look at this would be what would this imply in terms 25,000 of new loan originations every year. Given that housing will still be 20,000 around 60% of total retail loans and it runs off at ~14% every year, 15,000 the annual originations needed by F2027 will be around Rs40 trillion 10,000

18% CAGR compared to around Rs9 trillion in F2017 (16% CAGR). Given nominal 5,000

GDP growth at ~11% and the focus on consumer loans, this should be 0 2012 2017 2027e achievable in our view. Source: RBI, NHB, Company Data, Morgan Stanley Research (e) estimates

We can also compare the experience of the US when its GDP rose from US$2.1 trillion in 1977 to US$6.8 trillion in 1993. India will likely Exhibit 54: be on a faster path than the US was in that period given the significant In the US, Consumer Credit Penetration Surged When Nominal GDP difference in GDP per capita. During this period, retail credit penetra- Went From US$2.1 Trillion to US$6.8 Trillion (1977-1993) 100% tion in US went from 42% of GDP to 59%. Again, given the difference US Nominal GDP went from ~USD 2.1 trillion in in GDP per capita between the US then and India now, this is not 1977 to USD 6.8 trillion 80% in 1993. During this time strictly comparable. But it does show that penetration can increase retail credit increased from 42% to 59% of GDP meaningfully as populations get richer. 60%

40%

20%

0% 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

Source: Central Bank data, World Bank, Morgan Stanley Research

40 M BLUEPAPER Implication 2: Digitization = Economic Growth Booster

Key Takeaways

1. Digitization should provide an additional boost to an already compelling structural growth story. We estimate that digitization will provide a boost of 50-75bps to GDP growth and forecast that GDP growth will average 7.1% between F2018 and F2027, with India growing to a US$6.0 trillion economy and achieving upper-middle income status by F2027. 2. As digitization takes hold, SMEs and consumers will have better access to credit, which should help boost domestic demand. Specifically, SMEs will be able to lift their capital spending, thereby raising sustainable growth, while households will be able to increase their discretionary spending. 3. Digitization will help improve public finances via stronger tax compliance and enhanced efficiency in welfare spending. All else equal, digitization should help improve the country's fiscal deficit, and public debt to GDP should decrease over the next ten years.

Exhibit 55: Key Forecasts F2018-F2027E F2017 Bull Base Bear Real GDP, YoY (avg) 7.1% 8.4% 7.1% 5.9% Nominal GDP, US$ bn (eop) 2,279 7,026 6,040 5,180 Nominal GDP, YoY (avg) 11.3% 13.0% 11.2% 9.5% GNI Per Capita, US$ (eop) 1,702 4,809 4,135 3,546 Investment as % of GDP, (avg) 29.8% 34.3% FDI, US$ bn (eop) 60 120

Source: Morgan Stanley Research, Note: GDP growth rate in INR terms.

MORGAN STANLEY RESEARCH 41 M BLUEPAPER India’s Structural Growth Story Remains Exhibit 56: India's Demographic Profile Remains Favorable 80% Age Dependency Ratio (%) Very Compelling India World ex India 75% US China 70% Demographics Deteriorating We have, on a number of occasions (e.g., see The Next India - From a 65%

Cyclical Downturn to a Structural Upturn), presented our view that 60% India is likely to continue to enjoy structurally strong rates of growth. 55% Indeed, since the mid-1980s, trend-line growth rates in India have 50% been rising steadily. GDP growth has averaged 7.1% in 2000-17, 5.8% 45% 40% in the 1990s, and 5.7% in the 1980s. Over the coming decade, India is 35% likely to remain one of the fastest-growing large-sized economies in 30% the world. 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014E 2016E 2018E 2020E 2022E 2024E 2026E 2028E 2030E 2032E 2034E 2036E 2038E 2040E 2042E 2044E

Source: United Nations Population Database, Haver, Morgan Stanley Research. E=UN estimates The structural growth story is based on a combination of three key drivers: demographics, globalization, and reforms. First, favorable demographic trends provide surplus labour that can be mobilized. Exhibit 57: Second, globalisation provides exports as an additional source of India Likely to Contribute 27% of the Addition to the World's Working aggregate demand and financing that can be used to boost growth. Age Population in the Next Ten Years Working Age Third, economic reforms spur the corporate sector to invest, which Population as of Additions to Working Age Population, People mn 2016 in turn leads to utilization of surplus labour and unleashes faster pro- (2017-2026E) World4872 457 (100%) ductivity growth. Africa* 676 208 (46%) India 874 122 (27%)

ROW 735 110 (24%)

In fact, it is this combination – favorable demographics, riding the LATAM 428 43 (9%) wave of globalization, and implementing policy reforms that boost Indonesia 175 19 (4%) Rest of AXJ^ 199 7 (2%) investment – that has been a key factor in sustaining growth across US 214 3 (1%) countries in Asia. Throughout the region, a virtuous cycle of falling Japan 76 -5 (-1%) China 1005 age dependencies (i.e., the ratio of non-working age population to -24 (-5%) Figures in () indicate share in world working age Europe 490 -27 (-6%) population addition working age population), along with an improving saving (and invest- -80 -30 20 70 120 170 ment) to GDP ratio, has led to long phases of strong GDP growth. Source: United Nations Population Database, World Bank, Haver, Morgan Stanley Research. E=UN estimates Japan was the first to experience a positive demographic wave in the early 1950s, followed by the collective experience of Hong Kong, Korea, Singapore, Taiwan from the 1960s and China from 1974 onwards. According to UN estimates, India will continue to undergo this positive demographic trend of an improving age dependency until 2040, and, thus, we believe it is well positioned to experience a phase of rapid growth.

42 M BLUEPAPER Investment - The Key to Unlocking the Exhibit 58: India's Saving …

55% Savings as % of GDP India Structural Story Asian tigers 50% China Japan 45% Indonesia We note that favorable demographics is a necessary but not suffi- 40% 35% cient condition for a phase of rapid growth. While a cycle of falling 30% age dependencies typically leads to a virtuous cycle of improving 25% saving to GDP rates, the key to unlocking a phase of strong GDP 20% 15% growth lies in investment. A rise in investment-to-GDP ratio is 10%

needed to drive a virtuous cycle of higher investment – job growth 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Note: Asian tigers are Hong Kong, Korea, Singapore and Taiwan. Source: IMF, CEIC, Morgan Stanley – higher income – and higher saving. Research.

Structurally, though, India’s investment-to-GDP ratio remains below that which other Asian economies experienced during their rapid Exhibit 59: development phases ( Exhibit 59 ). Indeed, despite India's favourable … and Investment Rates Remain Weak Relative to Those of Other Asian demographic trends, the only instance in which there had been a Economies

50% Investment as % of GDP India strong investment cycle was during the 2003-07 period, when Asian tigers 45% China investment-to-GDP picked up to a peak of 38.1% in 2007, as com- Japan 40% pared with an average rate of 22.6% between 1980 to 2002. Unsur- 35% prisingly, this was also the period during which India's GDP growth 30% rate had been the strongest in the country's history, as the contribu- 25% 20% tion of investment growth accelerated significantly. However, out- 15% side of this period, India’s investment-to-GDP ratios have remained 10%

lackluster, averaging just 27.9% of GDP, as compared to an average 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Note: Asian tigers are Hong Kong, Korea, Singapore and Taiwan. Source: IMF, CEIC, Morgan Stanley of between 30-35% for most other Asian economies, when they were Research. experiencing their favourable demographic waves.

One of the key drivers affecting India's investment trend is the lack of saving within the economy. India’s savings trend has been consistently below that of the other Asian economies, despite the country's more favourable demographic profile. Indeed, India’s 2017 savings ratio of 27.2% is almost identical to that of Japan, despite the fact that India’s median age is 27, as compared with 47 in Japan, and that the age dependency ratio in India is at 51.0%, as compared with 66.5% in Japan.

While India's private capex cycle has been weak in recent years, con- strained by a challenging macro environment, we expect it to pick up in 2018, driven by recovery in end-demand, improvement in corporate balance sheets, and the positive momentum provided by strong public capex and FDI flows. Over the medium term, we expect the share of investment to GDP to rise, supported, as noted, by a cycle of favourable demographics and rising income and savings rates. Ulti- mately, in our view, it is the strength of the capex cycle that will determine how strong and sustained the growth cycle will be. Over the next decade, we expect investment to GDP ratios to average 34.3%, contributing 3% points to real GDP growth.

MORGAN STANLEY RESEARCH 43 M BLUEPAPER Digitization - Three Channels of Impact Exhibit 60: Real GDP Growth to Average 7.1% YoY; We estimate a US$6 Trillion Economy by 2027 Nominal GDP, USD bn It is against this backdrop of structural forces that the recent devel- 8000 opments in digitization should be evaluated. We see three key chan- 7000 7026 Bull nels of impact as digitization takes hold: improved access to credit for 6000 6040 Base

SMEs (boosting investment), stronger lending to households 5000 5180 Bear

(boosting discretionary consumption), and strengthening tax com- 4000 pliance and better efficiency in welfare spending (fostering improve- 3000 ment in public finances). The key impact will come via increasing the 2279 2000 2043 availability of financial saving – and thereby boost the investment 1000 952 cycle. Digitization could provide greater economic benefits than the 0 50-75bps boost to growth that we have accounted for. For instance, 0 F2007 1 F2015 2 F2017 3 F2027E 4 5 intangible benefits (e.g., reduced costs and improved efficiency in Source: CEIC, Morgan Stanley Research. E = Morgan Stanley Research estimates. doing business) would yield a key productivity boost. Digitization could also have the potential to disrupt and create new growth chan- Exhibit 61: nels, and could thereby provide additional growth engines. India Fares Relatively Poorly with Respect to Other EMs on Credit Access to MSMEs

35% % of MSMEs with access to credit 1. Corporate sector: Improved access to credit, especially for 30% 31% 30% MSMEs 25% 20% 18% 16% 16% 15% 14% 14% 12% 13% 10% 11% 11% 11% On the corporate side, the key impact will be via increased access to 10% 5% credit, particularly for the underserved MSME sector. This will princi- 0% South Africa China Republic of Thailand Indonesia Philippines India Mexico Malaysia Russia Argentina Brazil Poland Korea pally help boost investment, leading to stronger and sustained Note: Latest available data as of 2016 Source: IFC, Morgan Stanley Research growth.

Traditionally, the lack of formalisation and access to credit has been Exhibit 62: the key impediment to growth in the MSME sector. The lack of for- Lack of Credit Access to MSMEs Remains a Barrier

50% % of MSMEs listing access to finance as major/severe barrier 44% malisation (i.e. lack of legal status) limits access to new markets such 45% 42% 40% 34% as the public sector and MNCs and deters the formation of business 35% 30% 25% 25% 23% linkages with registered companies. Informal MSMEs are less likely 20% 20% 18% 18% 15% 15% to invest in human or physical capital, given the limited size of their 15% 12% 12% 10% 5% end-markets. At the macro level, this implies sluggish productivity 0% Malaysia Philippines Indonesia South Africa China Mexico Turkey India Poland Argentina Brazil Russia growth and lower aggregate growth rates. Note: Latest available data as of 2016 Source: IFC, Morgan Stanley Research

The other key constraint is the lack of available credit. With better access to credit, the MSME sector can create more jobs as they can Exhibit 63: start new businesses more easily, make larger investments, and indi- Credit Access for MSMEs Remains Weak for India

25% 23% 23% rectly create employment via supply and distribution chains as firms % of MSMEs that have loans 20% expand operations after gaining access to finance. 15% 13% 13% 11% 11% 10% 10% 9% 9% 9% 9% 10% 8%

0% India China Korea South Africa Thailand Indonesia Philippines Argentina Russia Malaysia Mexico Brazil Poland

Source: IFC, Morgan Stanley Research

44 M BLUEPAPER What has changed and will change for the MSME sector? Exhibit 64: India Only Fares Better than a Few EMs with Respect to MSME Credit In the case of India, the implementation of the GST is addressing both Value Gap the issues of formalization of SME sector and access of credit, which 15% Total MSME credit value gap, % of GDP 13% 12.2% 12.5%

11% will significantly improve SMEs’ growth prospects. The GST has 8.9% 9% 8.2% 8.5% 6.5% already led to the registration of 5.9 million taxpaying entities in the 7% 6.1% 6.2% 4.5% 5% 3.8% 3.9% 3.0% 3.0% GST network, with an additional 1.4 million yet to complete their pro- 3%

1% cedural formalities (in total, approximately 1.9 million more entities -1% Indonesia China Philippines Russia South Africa Mexico India Malaysia Poland Turkey Republic of Argentina Colombia Korea than before GST), as of 29 August 2017. While the push for registra- Note: Credit value gap: Degree of access to credit and use of deposit accounts. Source: IFC, Morgan tion in the MSME segment might have been predominantly driven by Stanley Research the need to facilitate the claiming of input tax credits, it has the positive spillover benefits of providing a kickstart to the formalisation Exhibit 65: process. Households in India Spend Almost 50% of Their Expenditure on Food 50 IN We expect digitization to improve flow-based lending, with the SME 45 sector being the key beneficiary in the corporate space. We expect 40 RU overall credit growth (from banks and NBFCs) to the MSME sector 35 30 HU HK BR JP TW to grow at a CAGR of 17% over the next ten years. SMEs will then be PL 25 TR better able to achieve scale economies and lift their investment. We ZA MX ES 20 IT FR CN expect this to provide a boost to capital deepening, and to provide a 15 DE US boost to growth of about 30-40bps over the next decade. 10 KO CPI CPI Weight for Food and

Beverages Beverages (Parts per 100) UK 5 2) Households: Better consumption smoothing, stronger con- 0 0 10000 20000 30000 40000 50000 60000 sumer durable spending GDP per Capita (Current International Dollar)

Source: CEIC, Haver, IMF, Morgan Stanley Research. GDP per capita and CPI weights as of 2016 for Households, too, will benefit from improved credit access. Credit Korea, 2017 for all other countries penetration in the household segment in India remains low relative to the country's per capita incomes, limiting the ability of households Exhibit 66: to smooth consumption over their lifecycle. Rising per Capita Incomes Indicate Shifting Consumption Patterns away from Food In recent years, shifts have begun to occur in the consumer credit 90.0 space. India's household debt to GDP ratio has risen from 11.8% in 85.0 2010 to 16.2% in 2016. This has also led to an increase in the private consumption to GDP ratio, from 56.0% in F2012 to 58.8% in F2017. 80.0 US 75.0 The major boost to consumer activity will likely occur in the discre- China* tionary consumption space. Based on the CPI basket weights, food 70.0 Japan Korea** accounts for 46.8% of annual household expenditure, which is CPI CPI Weight ex Food and 65.0 Beverages Beverages (Parts per 100) Germany amongst the highest such levels globally ( Exhibit 65 ). While this is likely a reflection of India's current level of economic development, 60.0 0 20000 40000 60000 as per capita incomes rise, the share of expenditure will likely shift GDP per Capita (Current International Dollar) towards non-food consumption. A cross-country analysis of a Note: *China excluding food only, **Korea excluding food and non-alcoholic beverages only Source: CEIC, Haver, Morgan Stanley Research. Based on data for the following time frames for various countries - US: number of DM and EM economies as of 2016-17 ( Exhibit 66 ) sug- 1980, 1987-2017, China: 2006-2017 Japan: 1970-2017, Korea: 1994-2016, Germany: 1995-2017 gests that food tends to account for an average of 21.6% of the CPI basket, with the shares ranging from a low of 9.7% in the UK to a high of 35.5% in Russia.

MORGAN STANLEY RESEARCH 45 M BLUEPAPER Thus, as the experience of other countries suggests, the share of Exhibit 67: expenditure towards non-food items tends to rise as per capita India's Tax Revenues to GDP Levels Appear Low in an EM Context

35 Tax revenue as % of GDP, 2016 incomes rise. Indeed, India's own experience has begun to reflect this 31.9 32.1 30.2 transition, too. As per Euromonitor data, the share of consumer 30 25.2 25 23.8 expenditure on food and non-alcoholic beverages has been system- 21.1 19.3 19.6 20.0 20 atically declining since the late 1990s, and we expect this transition 17.2 17.7 15 to gather pace as per capita incomes rise. 11.4 10

5 3) Government: Increased tax compliance and revenues, more effi- 0 cient public spending Indonesia India China Korea Colombia Poland Turkey Mexico S.Africa Russia Brazil Argentina Source: CEIC, Haver, Morgan Stanley Research

The third channel in which digitization helps from a macro perspective is via its impact on government finances. Two positive effects Exhibit 68: should emerge: 1) boost to tax revenues and 2) improved efficiency Public Debt to GDP to Decline: Debt to GDP Ratios in F2027 Based on in spending. Various Primary Deficit and Nominal GDP Assumptions Nominal GDP Growth (YoY %) Primary Deficit (% of GDP) 11 12 13 14 15 a) Improved tax compliance helps control fiscal deficit and 1.0 57 53 48 44 41 reduce public debt 1.2 59 54 50 46 42 1.4 61 56 52 47 44 1.6 63 58 53 49 45 India’s tax revenue to GDP ratio is on the low side as compared with 1.8 65 60 55 51 47 2.0 67 62 57 52 49 levels in other EMs, and this has been a key obstacle in achieving a Note: Current primary deficit as of F2017 is 0.3% and public debt to GDP is at 67.6% in F2017. Source: fiscal deficit more in line with that of EMs. The recent implementa- Morgan Stanley Research estimates tion of the GST (while not necessarily a direct beneficiary of digitization – though aided by it, and it should drive further digitization), will help increase tax compliance amongst corporates, leading to a boost The government has already implemented a direct benefit transfer in tax revenues, even while the tax rates are, in and of themselves, (DBT) scheme – a social welfare scheme under which the subsidy revenue neutral. benefits are directly transferred to the beneficiary’s account – rather than issuing checks / cash payments / rebates. Presently, 290 Over time, improvement in tax revenues should drive improvement schemes in 50 ministries are under DBT, and the government plans in fiscal deficits and public debt to GDP trends. Indeed, we estimate to scale that up to 533 central payout schemes in 64 industries by that if the primary fiscal deficit stays at 1.2% of GDP (as compared March 31, 2018 (see Livemint, September 7, 2017). Cumulatively, the with an average of 2.1% over the past five years, and at 1.6% in total direct benefit transfer so far has been approximately F2017RE), public debt to GDP should decline to below 60% by INR2025bn, and savings have amounted to INR496.5bn over the pre- F2027. vious method. b) Improved efficiency in public spending

On the expenditure side, digitization efforts also enable the government to introduce a more targeted approach towards disbursing welfare spending to the country's citizens. A more targeted approach would help in preserving financial resources – in contrast to the previous approach of engaging in transfers to households, which was prone to leakages.

46 M BLUEPAPER Global Implications Exhibit 69: India to Achieve Upper Middle Income Status within the Next Ten Years GNI Per Capita, USD 5000 4809 Bull It took nearly 60 years from independence for India’s GDP to reach Base US$1 trillion (in F2007). However, the confluence of positive factors 4000 4135 3546 Bear (demographics, reforms and globalization) helped the country reach 3000 the next trillion milestone within nine years (by F2015). With sustained strong rates of growth in the coming decade and the added 2000 1702 boost from digitization, India's economy, we believe, could reach 1560 1000 US$6 trillion by F2027. By our estimates, this would make India the 788 third-largest economy by then, up from seventh currently. 0 0 F2007 1 F2015 2 F2017 3 F2027E 4 5 On a per capita income basis, the changes would be even more pro- Source: World Bank, Morgan Stanley Research. E=Morgan Stanley Research estimates. GNI=Gross national income. We use the income status classification as defined by the World Bank. found. We estimate that per capita annual incomes will likely reach US$4,135 by F2027, lifting India into an upper-middle income status. Against this backdrop, we project that India will receive gross FDI To put this in perspective, per capita incomes in India by 2027 would inflows amounting to US$120bn by F2027, almost double the cur- be similar to those of China in 2010. Such a rise of a large-sized, rent 12-month trailing run rate of US$64bn. As a marketplace, India, upper-middle income society is likely to have profound implications in our view, offers significant new opportunities in the discretionary on how India interacts with the rest of the global economy and consumption and infrastructure spaces. As highlighted previously, becomes an even larger market for many of the world’s companies. with the rise in per capita incomes, discretionary spending growth in India is on track to accelerate and take a larger wallet share in con- From a macro perspective, India’s markets have been more open at sumer expenditures. Similarly, in investment, the needs for infra- this stage of development. As it is, India is attracting FDI inflows structure buildup will also have to be met. Aside from the funding amounting to 2.7% of its GDP on a 12-month trailing basis in July 2017. requirements, capital goods companies can also be a significant sup- India accounts for 2.5% of total global FDI flows (as of 2016) and is plier to meet India’s infrastructure needs. the 11th largest recipient of global FDI inflows. This is up from 0.8% of global FDI flows in 2005 when India was the 32nd largest recipient As a case in point, the recently announced Mumbai-Ahmedabad of global FDI inflows. FDI flows have been concentrated thus far in bullet train project (costing Rs108bn [US$1.7bn], and slated for com- the telecom, insurance, wholesale and retail trade & IT sectors. pletion by 2022), which is being 80% financed by a loan from the Japan International Cooperation Agency, will likely be beneficial for In particular, foreign investment for the financial and technology sec- Japanese companies that can leverage India as an export market and tors (which traditionally are more protected) are more liberal as com- transfer of Japan's technology. In the high-speed rail space, six other pared with China at a comparable development stage (per capita projects are in the pipeline, highlighting the infrastructure develop- incomes), and financial liberalization efforts have also gathered ment opportunities in India. apace relative to India’s stage in economic development. This open- ness has meant that companies will be in a better relative position to capitalize on the growth opportunities that India is offering.

MORGAN STANLEY RESEARCH 47 M BLUEPAPER Implication 3: A Massive eCommerce Opportunity

Key Takeaways

1. The eCommerce market should receive a major boost, with rising internet penetration, improving data speeds and a fast developing handset ecosystem, an ever-growing assortment of goods and services online and a palpable need for convenience. The government’s drive to digitize payments and the introduction of GSTN are significant enablers for the growth of the online economy. These coupled with the regular flow of growth capital from both financial institutions and strategic players will be key to the rapid growth of the eCommerce industry. 2. We expect eCommerce sales to grow at a 30% CAGR through F2027 (C2026) and to touch ~US$200bn of GMV, translating into a 12% online penetration within retail sales, versus ~2% in F2017 (C2016). 3. Between 2014 and August 2017, the eCommerce sector received funding of US$10.3bn. The importance of funding was demonstrated in 2016 when it fell 66% yoy and led to a slowdown in growth in Gross Merchandise Value (GMV). The level of funding has improved in 2017 YTD (reaching US$4.3 bn by the end of August), but a regular inflow of large amounts of money will be critical for growth of the segment, in our view.

Exhibit 70: Scenario Analysis for India's eCommerce Sales Suggests that the Market Can Compound Significantly from Current Levels F2027e Scenarios for India online retail market F2017 BASE BULL BEAR Total Internet users (million) 432 915 1,010 841 Internet Penetration (%) 33% 62% 68% 57% Total Online shoppers (million) 60 475 630 378 Online shoppers as % of total internet users 14% 52% 62% 45% Total online retail market size including food delivery (US$ bn) 15 200 251 105 Total online retail as % of total retail (%) 2.2% 12.1% 14.0% 7.7%

Source: IAMAI, Morgan Stanley Research (e) Estimates. Please note that F2027 represents calendar year C2026 ending December 2026 and so on for the previous years; For F2017 online retail market size is based on Morgan Stanley Research estimates and is an approximate figure .

48 M BLUEPAPER We believe the eCommerce industry in India will grow extremely fast Exhibit 71: in the next ten years for the following reasons: AlphaWise Survey (2015) Showed that Individuals with Over Five Years of Use on the Internet Were More Likely to Transact Online 1. Rising internet penetration - India is adding over two new 80% internet users every second, driven by falling prices of smartphones, 70% the higher adoption of mobile devices, and the availability of telecom 60% infrastructure that supports 3G/4G speeds. 81% of all smartphones 50% shipped to India in 2016 were 4G LTE enabled, and the price of the cheapest 4G handset touched Rs3,000 (a 36% decline in 15 months). 40%

Internet penetration was 33% in F2017 (C2016), with 432mn internet 30% users, and we expect penetration to go over 60% by F2027 (C2026), 20% with about 915mn internet users. India has a large population of millennials who will be exposed to technology and the internet at an 10% early age and could drive rapid online adoption. 0% 1-2 yrs 3 yrs 5 yrs > 5 yrs Not transacting online Transacting but not shopping New Product Purchase 2. Availability of higher assortment as more merchants/prod- Source: AlphaWise, Morgan Stanley Research ucts come online - We believe a healthy marketplace is one where there are several sellers on one side and several buyers on the other. users with less than two years of activity on the internet were less Currently, the number of merchants selling on online platforms such susceptible to transacting and primarily indulged in emails, mes- as Flipkart is around 150,000, however, this number could signifi- saging and search. However, users with over five years of experience cantly increase. Amazon India has 160mn products listed on its plat- on the internet were more likely to transact online. A large chunk of form, versus 400mn+ for Amazon in the US. We can see two key the internet base in India has emerged only in the last few years and developments taking place that can increase the number of sellers/ hence has possibly not matured yet to the stage of transacting products online: a) Availability of credit to merchants - large eCom- online. However, in 2019 and beyond, more than half of the internet merce companies such as Flipkart and Amazon have implemented population will have matured over five years and that could possibly seller lending programs for providing working capital loans to sellers be when online shopping sees a massive inflection. Further, India's on their platform. With these loans, not only will the sellers be in a millennial population (including past millennials of 2016) by 2020 position to bring more inventory online but also this development will be 36% of the total population but will be a much larger force to will help bring new sellers on the platforms. In addition to these pro- reckon with in terms of the internet and online shopper population; grams, various P2P marketplace lending platforms such as Capital- this could be a key driver of online activity. As an example, as per our float, Indifi and Lendingkart are providing loans to merchants that survey of millennials in India in 2016 (The Next India: Technology: The were historically unable to take loans from banks; and b) GST - Millennials Series: The Disruptive Wave in the World's Seventh increased availability of credit due to GST will be a key enabler of Largest Economy (Feb 19, 2017)), the online shopper penetration is eCommerce. already high at close to 40% (versus 14% for the overall population in India in 2016) across some key segments; we expect online shopper 3. Faster growth of online shoppers and per capita spends - India penetration in India to cross 50% by F2027 (C2026). With close to had only ~60mn online shoppers in F2017 (C2016), which was ~14% 475mn online shoppers in F2027 and an online spend of ~10% of the of the total internet population, versus 64% in China. Our analysis of per capita income, we believe the eCommerce market can grow rap- some global eCommerce companies highlights that two-thirds of the idly from the current base. Furthermore, with the help of data, we growth in their eCommerce sales happened due to new users coming expect banks, NBFCs and fintech companies to provide short-term online and shopping, while the balance was driven by existing online financing online, which could significantly boost consumerism in the shoppers buying more frequently and/or driving up order values. As country. per an AlphaWise survey conducted in 2014, we saw that internet

MORGAN STANLEY RESEARCH 49 M BLUEPAPER

Exhibit 72: Exhibit 73: India's Online Shopper Penetration in F2021 Will Be Similar to What Convenience Outstrips Price Attraction Once Users Have Over Five Brazil and Russia Have Today, and by F2027 it Will Be 50%+ Years of History on the Internet; by C2019-20, About 50% of the Indian 70% Internet Population Will Have Been on the Internet for Over Five Years 60%

50%

40%

30%

20%

10%

0% UK US Brazil China Japan Russia Germany Source: AlphaWise, IAMAI, Morgan Stanley Research; E = Morgan Stanley Research estimates India in in F2027 India India in in F2016 India in F2021 India

Source: CNNIC (China Internet Network Information Center), IDC (International Data Corporation), For- rester, IAMAI (Internet and Mobile Association of India), Morgan Stanley Research estimates for India in F2021 and F2027. Data as of 2015. Please note that F2027 represents calendar year C2026 ending December 2026 and so on for the previous years.

4. Evolving payment methods thus reducing dependence on cash Exhibit 74: - Cash on delivery is the main payment method, accounting for about Cash on Delivery Is a Prominent Payment Method for eCommerce 55-60% of total eCommerce sales. One of the key reasons for the Sales; However, New and Easy Payment Methods Along with Growing high dependence on cash is the trust deficit that exists between cus- Customer Confidence in Online Retailers Could Cause Payments to tomers and online merchants. However, with the passage of time, we Become More Digital believe cash could lose market share to credit/debit cards and pay- 70% ment wallets. We have already seen a significant surge in the usage 60% 55-60% of digital wallets from about two or three years back when their usage was small. This should improve the online shopping experience 50% by making payments more seamless and therefore easier. 40%

30% 5. GST could boost logistics and warehousing capabilities - We 24-27% see GST enabling the free flow of goods across the country as hur- 20% dles at state borders (in the form of tax payments and other adminis- 11-14% 10% trative measures) are done away with. Not only will surface 4-7% transportation become a possibility, but the change may also cause 0% eCommerce and logistic companies to relook at their warehousing Cash on Delivery Credit Cards Debit Cards and Wallets/UPI/Others Net Banking strategies and possibly invest closer to demand centres, thus Source: Morgan Stanley Research estimates improving delivery timelines and customer experiences.

50 M BLUEPAPER eCommerce Opportunity Could Be US online as big data, the availability of funding to both merchants and customers, and the rapid adoption of digital payments will make it a $200bn in F2027 more convenient and fulfilling experience. eCommerce sales in F2017 (C2016) touched ~US$15bn; however, this eCommerce was about 2% of the total retail market in 2016 and we marked a significant slowdown from what had occurred in the pre- see this climbing to 12% of the retail market by F2027 (C2026), vious years. One key reason for the slowdown was soft VC/PE (ven- reaching a size of US$200bn (including food delivery). While smart- ture capital and private equity) activity leading to companies scaling phones, electronics and apparel are some of the prominent catego- back on growth plans and instead focusing on profitability. Further- ries today, we expect grocery, personal and beauty products, more, with funding drying up and becoming restricted to only a few furniture and food delivery to become important as well. The key big players in each segment, several start-ups folded up. The demon- drivers of this growth will essentially be a growing population of etization drive in November 2016 also impacted growth given the online shoppers aided by a large assortment of merchandise avail- massive reliance on cash-on-delivery. However, funding has resumed able online. in 2017 and with that we see companies are focusing on growth again. The digitization push from the government however, reinforces our view that a lot more transactional activities are likely to be done

Exhibit 75: Exhibit 76: Top eCommerce Apps in India (in Terms of Downloads) Engagement Levels on eCommerce Platforms Looking Robust in India; Rank App Category Average Time Spent on Various Online Platforms in India in July 2017, 1 Paytm General merchandise/Digital Wallets Along With a US Comparable (in Minutes) 2 Amazon India General merchandise 160 mins

3 Flipkart General merchandise 140 700

4 Snapdeal General merchandise 120 452

5 Myntra Fashion/Apparel 100

6 Limeroad Fashion/Apparel 80

7 Voonik Fashion/Apparel 60 40 8 Shopclues General merchandise 20 9 Jabong Fashion/Apparel 0 10 AJIO Fashion/Apparel Oyo Yatra MMYT Paytm Paypal

Source: Sensor Tower, Morgan Stanley Research; the above rankings are based on downloads for the 12 Goibibo Flipkart Booking Expedia Amazon Phonepe Youtube Mobikwik Bigbasket months ending August 2017 Snapdeal

Video eCommerce Payments Travel Average time spent overall - US (mins) Average time spent overall - India (mins)

Source: ComScore, Morgan Stanley Research; Data for July 2017; Average time spent calculated as total time spent in minutes divided by total unique visitors/viewers

MORGAN STANLEY RESEARCH 51 M BLUEPAPER

Exhibit 77: Exhibit 78: eCommerce Sales in India to Cross US$200bn by 2026 and Constitute eCommerce as a Percentage of GDP: The Effect on the Indian Economy ~12% of Total Retail Sales in India in our Base Case Scenario (US$bn, Is Likely To Be High %) 9.0% China 8.0% 300 14.0% 7.0% 12.0% 6.0% 250 Bull Case 5.0% 10.0% 200 Base Case 4.0% 8.0% 3.0% India 150 2.0% US 6.0% Bull Case Bear Case 1.0% 100 4.0% 0.0% Base Case 50 2.0%

Bear Case F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F2018e F2019e F2020e F2021e F2022e F2023e F2024e F2025e F2026e F2027e

0 0.0% Source: IMF, IDC, Euromonitor, Morgan Stanley Research. eCommerce estimates are Morgan Stanley Research estimates. Please note that F2027 represents calendar year C2026 ending December 2026 and so on for the previous years. F2014 F2015 F2016 F2017e F2021e F2027e

eCommerce Sales (US$ bn) ( L.H.S.) eCommerce sales as % of total Retail sales - Base Case (R.H.S.)

Source: Euromonitor, Morgan Stanley Research Estimates. Please note that F2027 represents calendar year C2026 ending December 2026 and so on for the previous years.

Exhibit 79: Exhibit 80: India's Online Spending Per Capita as a Percentage of GDP Per Capita VC/PE Funding: General Merchandise Dominates Within eCommerce in 2026 Could Be Similar to That of China in 2010 in Our Base Case Sce- (January 2014 to YTD 2017) nario, Based on Parity in GDP Per Capita (US$, %) 10,000 20% General merchandise

8,000 16% India - Bull case scenario Fashion, shoes and accessories 6,000 12%

Furniture and home décor 4,000 8%

India - Bear case 2,000 scenario 4% Baby care

- 0% Groceries F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F2018e F2019e F2020e F2021e F2022e F2023e F2024e F2025e F2026e F2027e China GDP per capita (US$) India GDP per capita (US$) Jewelry

India - Online spend as % of GDP per capita (RHS) China - Online spend as % of GDP per capita (RHS)

Source: CEIC, International Monetary Fund, World Economic Outlook Database, Morgan Stanley Research (e) Estimates. Please note that F2027 represents calendar year C2026 ending December 2026 and so on for the previous years. Source: Economic Times, VC Circle, Inc42.com, Crunchbase, company data, Morgan Stanley Research

Exhibit 81: Key Investors in Indian eCommerce Sector

• SoftBank • Microsoft • Steadview Flipkart • Tiger Global • eBay • Accel India • Naspers • Tencent • SoftBank • SAIF Partners Paytm • Alibaba Group • Silicon Vallley Bank

• Nexus India Capital • Tiger Global

• GIC

• Abraaj Group • Helion Venture • Bessemer Venture Partners Partners

•SoftBank •• NexusSoftBank Venture Partners • Kalaari Capital • eBay •• KalaariNexus Capital Venture • Bessemer Venture Partners Partners

Source: Economic Times, VC Circle, Inc42.com, Crunchbase, company data, Morgan Stanley Research

52 M BLUEPAPER Implication 4: Digital to Drive Equity Bull Market

Key Takeaways

1. We expect India to continue to be among the world's best-performing markets, with a potential 24% CAGR in US dollar returns over the coming five years. 2. We expect such a return to be driven by acceleration in earnings growth (12% CAGR, 2017-27, in US dollar terms), as well as by multiple expansion. 3. The starting point of under-penetration combined with digitization almost guarantees growth in financial assets and liabilities on the household balance sheet, driving up the size of the domestic mutual fund industry and household ownership of equities. 4. We would focus on investing in the financials, consumer, and emerging large caps categories, while avoiding global sectors, such as technology and pharmaceuticals, to play this story. 5. We expect considerable expansion in market activity, with trading likely to rise from US$21 trillion to US$41 trillion in 2027, we estimate.

Exhibit 82: India Equities at a Glance Key Forecasts Current F2027e Bull Bear BSE Sensex 31,627 1,00,000 1,30,000 65,000 10-year CAGR 7% 12% 15% 7% Market Cap (US$ billion) 2,100 6,100 8,500 3,700 10-year CAGR 6% 11% 15% 6% Aggregate Trading Turnover (US$ bn) 21,289 40,951 57,375 23,618 Cash Volumes (US$ bn) 1,131 4,575 6,375 2,220 Derivatives Volumes (US$ bn) 20,152 36,376 51,000 21,398 10-year Equity Issuances (US$ bn) 166 500 850 259 10-year Domestic Equity Flows (US$ bn) 60 420 706 211 Domestic Equity Mutual Funds AUM (US$ bn) 111 1,000 1,530 481 ETF AUM (US$ bn) 9 200 300 100

Source: Morgan Stanley Research (e) estimates

MORGAN STANLEY RESEARCH 53 M BLUEPAPER India Could Continue to Be a Strong Exhibit 83: India - The World's Top-performing Market Over Two Decades Equity Market over the Next Ten Years Top markets Market Cap (US$ Market Cap 20 year bn) to GDP returns

India’s digitization program, as discussed, has the potential to lift UNITED STATES 7,691 40% 164% GDP growth. The concomitant impact is that corporate earnings CHINA 7,691 65% -5% growth would also likely accelerate. India's history shows a good cor- JAPAN 5,754 119% 23% relation between nominal GDP growth and earnings, albeit earnings UNITED KINGDOM 3,593 144% 18% also depend on other factors such as the investment and saving rate. FRANCE 2,493 103% 114% We expect this relationship to be sustained going forward. Ulti- GERMANY 2,350 69% 109% mately, earnings growth is what drives share prices, but we also CANADA 2,255 141% 188% sense that Indian multiples could expand in the years ahead. Thus, INDIA 2,105 86% 396% Indian stock markets could be among best-performing in the world in SWITZERLAND 1,750 265% 159% the coming decade. KOREA 1,544 103% 387% AUSTRALIA 1,327 98% 140% Indian equities have been strong performers over the past 20 years. TAIWAN 1,180 208% -4% In US dollar terms, the MSCI India index ranks as the top-performing BRAZIL 944 44% 88% market among the world's largest economies. This is consistent with SPAIN 830 67% 88% India’s GDP growth, which also ranks among the best over this period. SWEDEN 829 163% 177% Indeed, in local currency terms, the market has done even better. Source: Bloomberg, IMF, WEF, MSCI, RIMES.

With such trailing performance, the temptation may be to be more cautious about the coming decade. However, in our view, factors Exhibit 84: seem ripe for India's outperformance to continue in the coming Earnings Growth Versus GDP Growth - Strong Relationship 80% decade. Broad Market Earnings growth 2- year MA 70% Why do we hold such a positive view? 60% 50% First, we expect US dollar earnings to compound at 40% 12% over the next ten years 30%

20%

We think earnings can easily match or better our forecasts for nom- 10% inal GDP with a 10.1% CAGR in US dollars over the next 10 years. The 0% next five years could witness significantly more growth than the 8% 10% 12% 14% 16% 18% -10% Nominal GDP growth 2-year MA latter part of the ten-year period, given the low starting point of Source: Morgan Stanley Research, Morgan Stanley Research estimates trailing earnings. India’s digital story could also lift India’s credit- to-GDP ratio. Rising credit to GDP is good for earnings growth. Rising nominal growth is also good for earnings. There is also another metric that should support the coming earnings cycle: the starting point of profit share in GDP is itself attractive – sitting well below trend, and we expect mean reversion as explained below. That the 10-year trailing CAGR in earnings is well below historical levels is only an added advantage to the earnings outlook for the coming period.

54 M BLUEPAPER Our framework to understand earnings comes from the Kalecki Exhibit 85: Equation, which itself is derived from the identity for the current We Expect Profits to GDP to Rise over the Next Decade account: 11% 10% Profits / GDP % 9% l Current Account = Investment - Saving (as % of GDP) 8% l Current Account = Investment - (Government Saving + 7%

Household Saving + Corporate Saving) (all as % of GDP) 6% l But Corporate Saving = Corporate Profits - Corporate Divi- 5% dends, so, rewriting, we get 4% l Corporate Profits (as % of GDP) = Investment - Current 3% 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Account - Government Saving - Household Saving + Divi- 2017e 2019e 2021e 2023e 2025e dends (as % of GDP) Source: CEIC, Capitaline, Morgan Stanley Research (e) estimates

This gives us the share of profit in GDP, and, multiplying it by nominal Exhibit 86: GDP growth, yields the nominal earnings growth rate. This analysis Rising Bank Credit Will Support Earnings Growth 40% Bank Credit 70% helps explain the source of earnings. Thus, profit share in GDP is a 60% 35% Broad market (Ex oil) earnings growth YoY - function of the investment cycle, the current account, and the sav- RS 50% 30% 40% ings rate of the government and households. We expect the invest- 30% 25% ment cycle to turn in the coming months, as explained earlier. While 20% 20% 10% a declining fiscal deficit is a drag on earnings, the changing mix of 15% 0% -10% household saving, from physical to financial, should provide signifi- 10% -20% cant support for profits. Digitization is helping accelerate the shift in 5% -30% household saving to financial assets. The current account appears Mar-06 Mar-07 Mar-08 Mar-09 Mar-10 Mar-11 Mar-12 Mar-13 Mar-14 Mar-15 Mar-16 Mar-17 Sep-05 Sep-06 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 Sep-15 Sep-16 range bound, at around 1.0-1.5% over the next 10 years, which is neu- Source: RBI, Capitaline, Morgan Stanley Research tral for profits.

Factors that favor earnings growth include a rising investment rate Households are leveraging-up their balance sheets, a prom- and household balance sheet leveraging. ising development for consumption, and a counter to a relatively strong savings rate. The shift in the household l We forecast that the nation's investment rate will rise, led balance sheet is being supported by India's digitization pro- by public investment, especially in infrastructure. We also gram, which will increase penetration of the financial expect the private capex cycle to turn next year. The gov- system as well as improve transparency thereby reducing ernment spending mix has changed favorably towards the attractiveness of gold and property as asset classes. investments, but aggregate government spending is still The starting point of under-penetration almost guarantees muted, and, to some extent, offsets the positive impact of growth in financial assets and liabilities on the household a rising investment rate. balance sheet. For example, mutual fund assets are just 9% l Positive real rates have helped reduce the current account of GDP compared with 100% in the US and consumer and are positively affecting the mix of household savings loans are only 16% of GDP - albeit partly explained by low towards financial saving. We expect the current account per capita income. deficit to stay in a narrow range, unlike circumstances in the recent past, when India ran a very high deficit. When the current account is in deficit, it means local demand is being serviced by global production, reducing domestic production and earnings.

MORGAN STANLEY RESEARCH 55 M BLUEPAPER Second, we expect Indian multiples to remain above Exhibit 87: historical averages - and a move towards 100,000 Our Macro Model Points to Recovery in Earnings Sensex, or US$6 trillion market cap 80% 70% Corporate Profit Growth Based on the Macro model 60% Two factors are likely to drive the multiple: 50% 40% 30% 1) Rising demand for shares from domestic investors: We think 20% India is in the midst of a domestic liquidity supercycle. We project 10% 0% equity saving of between US$420bn and US$525bn over the next 10 -10% Actual Broad Market Profit Growth years, versus the respective US$60bn and US$120bn that household -20% and foreign portfolio investments had invested over the previous 10 F1994 F1996 F1998 F2000 F2002 F2004 F2006 F2008 F2010 F2012 F2014 F2016 F2018E F2020E years. The three main sources of these flows are mutual funds, pen- F2023e F2025e F2026e Source: Morgan Stanley Research, Morgan Stanley Research estimates sion and provident funds, and insurance companies. The drivers of this change (see: Asia Insight: India's Domestic Liquidity Supercycle, May 2, 2017) are falling age dependency and low risk averseness (typ- Exhibit 88: ical of a young population) on the demand side, combined with macro India's EPS Growth Appears to Have Troughed 30% stability and initiatives to educate investors, as well as progressive MSCI India 5-year CAGR trailing EPS growth 25% regulations and digitization, on the supply side. Consensus is still 20% largely overlooking such drivers, and continues to be concerned 15% about sustainability of a domestic equity savings trend. We expect 10% five factors ("DREAM") to play an important role in driving an equity 5% savings boom in India: 0% -5% D – Demographics: A young population tends to be less risk averse, -10% and, combined with accelerating income growth, is likely to save more, especially into equities. Source: RIMES, MSCI, Morgan Stanley Research

R – Regulations: Multiple changes have made the equity markets more palatable to retail investors. Rule changes for institutions such M – Macro: Positive real rates make financial assets more attractive, as the National Pension System (NPS) and the Employees Provident while improving growth underpins equity performance. Fund Organisation (EPFO) are also driving flows into stocks. The demand for equities should be seen in light of the expanding E - Education, or investor literacy: This has helped create better sav- financialization of India's economy, as driven by digitization, rising per ings and investing behaviour among retail investors, who, for the past capita incomes, and increasing financial literacy, among other factors. 20 years, were almost absent from equities, leading to a big drop in We estimate that financial savings could swell by US$5 trillion over accumulated savings' exposure to equities. the next 10 years, premised on the assumption that the financial savings share in GDP rises 25bps in each of the next 10 years. We view A - Appetite for risk: Trailing returns establish equities as a favored this as a reasonable assumption because, despite the annual rise, the asset class over other categories, such as gold, bonds and property. share of financial savings to GDP would still be lower than the previous peak of 17.7% in F2007. By F2027, the financial-savings-to-GDP ratio would rise to 13.5%, we estimate, higher than the long-term average of 11%.

56 M BLUEPAPER 2) Stock multiples: The effect of declining public debt to GDP: As The sum total of higher earnings growth (which we estimate at we argue, robustness in government revenues arising from suc- 20% for the coming five years) and higher multiples implies that cessful execution of GST is likely to lead to a lower public debt to the leading indices could compound at around 24%, in US dollar GDP ratio. It clearly follows that this would lead to a crowding in of terms (25% in INR terms), in the coming five years. This type of private investments, with a positive impact on interest rates and front-ended returns could lead to moderation of returns in the inflation. This will also lead to higher multiples, given the historically subsequent five years. Thus, the coming ten-year period could inverted relationship between the change in public debt and P/E mul- witness an 11% CAGR in US dollar stock returns, just a shade tiples. behind our forecast for earnings growth, at 12%, for the same period.

Exhibit 89: Exhibit 90: Declining Debt-to-GDP Ratio Helps Valuation Multiples Rising Profit Share in GDP Drives Market-Cap-to-GDP Ratio

6 160% Market cap to GDP P/B 140% 5 120% y = 0.1207e20.779x 4 R² = 0.8384 100% 3 80%

2 60% 40% 1 Annual Change in Debt to GDP 20% Net Profit to GDP 0 9.0% 10.0% 11.0% 12.0% 13.0% 14.0% 15.0% 16.0% 17.0% 0% 4% 6% 8% 10% 12% Source: CEIC, MSCI, RIMES, Morgan Stanley Research. .Data as of August 31, 2017 Source: BSE, NSE, Capitaline, CEIC, Morgan Stanley Research . Data as of September 20, 2017

Exhibit 91: Exhibit 92: Market Multiple in the Middle of Historical Range India Market Forecasts 25% Annual 10-year fwd Current P/B of 3.1 implies a 10Y Avg. Trailing Peak Current F2027E MSCI India returns 10-year annual return of 13.8% Nominal GDP growth in US$ 14.9% 20.2% 8.7% 10.2% 20%

Nominal GDP - US$ bn 2503 6040

15% Net Profit to GDP 8.9% 10.7% 7.0% 8.5% MSCI India Trailing P/B 10% Net Profit growth (in US$) 12.8% 56.0% 1.8% 12.4%

R² = 0.766 Mkt Cap to GDP 80% 149% 84% 101% 5% y = -0.0358x + 0.246

India's Market Cap - US$ bn 1239 2105 2105 6100 0% 1 2 3 4 5 6 Source: BSE, NSE, Morgan Stanley Research estimates Source: RIMES, MSCI, Morgan Stanley Research. Data as of September 20, 2017

MORGAN STANLEY RESEARCH 57 M BLUEPAPER Portfolio Implications Exhibit 93: India's Index Weight Set to Rise in the Coming Years

11% India GDP weight in EM msci ff market cap weight in MSCI EM a) Buy Emerging Mega Caps 10%

9% In our view, as India gains performance versus emerging markets and the world at large, its MSCI EM index weight will rise. Already, India 8% is under-represented in the index relative to its GDP rank in the EM 7% world. As India's index weight rises, foreign flows will become larger, 6% and the equity market will also attract "tourist" money – larger quan- tities of opportunistic flows. This will result in the search for large 5%

liquid names. Simultaneously, domestic institutions are likely to 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 grow in size. Hitherto, domestic institutions derived a lot of their per- Source: MSCI, IMF estimates, Morgan Stanley Research formance from buying stocks down the cap curve. This was a viable strategy when domestic funds were small relative to the market. But, Exhibit 94: going forward, larger size will increase the need to make the right Domestic Mutual Fund Assets: Likely to Rise Tenfold, to US$1 Trillion calls on large liquid names. Counterintuitively, getting these in 10 Years large-cap calls right may become key to beating the market. The third DMF equity assets under management projection (US$ bn) 1000 reason to favor large caps is likely growth in exchange-traded index funds. We forecast that ETFs will grow 30x in the coming decade, to 800

US$200bn in assets by 2026. From a relatively small size of 600 US$100mn in 2009, ETF assets have grown to US$8.4bn in size in a 400 span of seven years. We expect ETFs in India to gain further traction as provident funds pursue a policy of channelling their investments 200 in equities via ETFs. In our view, all these trends point to the growing 0 importance of mega and large caps in India. F2012 F2013 F2014 F2015 F2016 F2017 F2018e F2019e F2020e F2021e F2022e F2023e F2024e F2025e F2026e

Source: AMFI, Morgan Stanley Research. e=Morgan Stanley Research estimates

Exhibit 95: ETF AUM Has Risen 8x in the Past Few Years 600 ETFs AUM (Rs bn) 528 500 Provident fund flows into ETFs 444

400

300

200 161

81 100 45 7 10 25 16 15 0 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F2018

Source: AMFI, EPFO annual report, Morgan Stanley Research

58 M BLUEPAPER b) Buy Financials and Discretionary Consumption; Avoid Global Exhibit 96: Sectors Consumption and Financial Sectors to Gain Share in Total Market Cap Financial sector Consumer sectors Global sectors 100% share in India's market cap Rising GDP growth, a rising loan-to-GDP ratio, and market share gains 90% for the private sector financials at the expense of the public sector are 80% 70% likely to drive strong top-line and earnings growth for Indian private 36% 20% 60% sector financials, in our view, boosting the sector's performance rela- 50% 40% 33% tive to other sectors in the coming years. We expect the share of 25% 30% financials in India's total market to rise from 22% in 2017 to 30% 20% by 2027, with their absolute market cap likely rising from 10% 22% 30% 0% US$460 billion to US$1.8 trillion in the coming ten years. 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 We expect discretionary consumption to surprise on the upside Source: Capitaline, Morgan Stanley Research estimates. because of rising GDP growth and increasing consumer credit pene- Exhibit 97: tration, aided by digitization. We expect consumer credit to com- Currently, India Has a High Food Share in Its Overall Consumption pound at around 17% annually over the next ten years. Given the high CPI Basket Urban India USA share of food in India's consumption basket, we expect the food share Food and beverages to decline as per capita incomes rise. Not only is nominal income set 36% 15% Housing 26% 37% to grow faster in the coming decade, but future consumption is being Clothing and footwear 6% 3% advanced through leveraging, and non-food consumption itself is Fuel and light 6% 5% likely to grow faster than overall consumption. Further, digitization Transport and communication 10% 19% will likely lead to an increase in the share of the organized sector in Healthcare 5% 9% the consumption space. Versus an overall annual GDP growth of 11%, Recreation 2% 6% it is quite possible that organized non-food consumption registers Education 6% 3% mid-teen growth in the coming decade. Thus, the share of consump- Others 5% 4% tion-related sectors (discretionary plus staples) in the market capi- Source: BLS, CEIC, Morgan Stanley Research talization could rise from just over 25% to 33%, at the expense of global sectors. This means that these sectors could add up to US$2 trillion in market cap by 2027, up from US$500 billion currently.

MORGAN STANLEY RESEARCH 59 M BLUEPAPER Global sectors, such as software services and healthcare, have Exhibit 98: already lost market appeal over the past few years. The valuations Tech and Pharma Underperform, But Could Keep Experiencing Trading for some large-cap stocks in these sectors have turned attractive, and Pops they find a place in our Focus List, with a 12- to 18-month time horizon, 30% Performance of Technology and Healthcare sector relative to MSCI 20% with a portfolio perspective in mind. However, widening our perspec- India 10% tive to a longer, multiyear view, these sectors could continue to lose 0% share in the overall market as their earnings lag those of the domestic -10% sectors. So, to fund overweight positions in Financials and Consumer -20% Discretionary, portfolio managers could aim to be underweight any -30% of these sectors: Materials, Energy, Software Services, Pharmaceuti- -40% cals. -50% -60% c) Big growth in capital markets 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: Capitaline, Morgan Stanley Research

We note two additional arenas that could grow along with India's Exhibit 99: market capitalization: stock market trading and equity issuances. At The Coming Financial Savings Boom a cash trading to market cap level of 56%, India is well below the 88% 20% average for the world's top 20 markets (by market cap). We expect Gross financial saving to GDP 18% trading to rise as financial savings penetration improves. Assuming 16% that, over the next ten years, India's cash volumes trade at 75% to 14% market cap (still below the world average) and derivatives volumes 12% trade at 8x cash volumes (the trailing average), this would imply that India could be trading about US$4.5 trillion in cash and another 10% US$36 trillion in derivatives, up from current levels of US$1.1 trillion 8% and US$20 trillion, respectively. As market cap grows, we expect 6% new listings, as well as existing listed companies, to raise fresh cap- F1980 F1982 F1984 F1986 F1988 F1990 F1992 F1994 F1996 F1998 F2000 F2002 F2004 F2006 F2008 F2010 F2012 F2014 F2016 F2018E F2020E F2022E F2024E F2026E ital. India's 20-year trailing equity issuances are 1.7% of market cap. Source: RBI, Morgan Stanley Estimates However, during the previous bull market (2004-08), equity issuances amounted to 2.3% of market cap. We think issuances could head back toward that level in the coming years. Using the historical average of 1.7%, equity issuances could add up to US$500 billion in the coming decade, as compared with US$166 billion in the trailing 10-year period.

60 M BLUEPAPER Implication 5: Financials – US$1.8 Trillion Market Cap in 10 Years

Key Takeaways

1. Strong asset growth should help Indian lenders to grow earnings at a good pace. Technology will impact various sources of revenue. However, Indian lenders lack economies of scale – as more business moves online, we expect operating leverage to be very large. The gap between digital leaders and laggards will cause continued shifts in market share. 2. We expect the market cap of financial services stocks to reach US$1.8trn over the coming decade, from US$370bn in F2017. This will be driven by the growth of existing businesses and new listings. 3. We believe the combined market cap of non-banks will grow faster than banks as they continue to gain share and as there are new listings for insurers, AMCs and other capital-markets businesses. We expect SOE banks to decline to less than 10% of the sector's market cap.

Exhibit 100: Indian Financials: Key Indicators Over Next 10 Years and Likely Move in Market Shares F07-17 F17-27E F2007 F2017 F2027E F2007 F2017 F2027E CAGR CAGR USD Bn %% % of GDP Digital Payments 11 107 1,208 25.1% 27.4% 1.2% 4.7% 20.0% Total Loans 512 1,530 4,704 11.6% 11.9% 51.9% 67.1% 77.9% MF AUM 74 273 1,935 13.9% 21.6% 7.5% 12.0% 32.0% Life Insurance Premiums 36 62 185 5.7% 11.5% 3.6% 2.7% 3.1% General Insurance Premiums 6 18 59 10.9% 13.0% 0.6% 0.8% 1.0% E-Commerce 0 15 200 41.5% 29.7% 0.0% 0.7% 3.3% Market Turnover (Cash) 667 1,131 4,575 5.4% 15.0% 67.6% 39.6% 75.7% Market Cap (Overall) 805 2,105 6,100 10.1% 11.2% 82.5% 79.6% 101.0% Market Cap (Financials) 95 371 1,844 14.5% 17.4% 9.7% 16.3% 30.5%

Source: RBI, IRDA, Bloomberg, Morgan Stanley Research (E) estimates. USD/INR for FY17 we have used is 67. MF AUM for FY17 is 4Q Average AUM. F2017 Market Turnover (Cash) and Market Cap (Overall) are current data

Exhibit 101: India Financials Market Cap as Percentage of GDP - Comparing with US and China as it Stands Now USA (Current) CHINA (Current) INDIA (F2017) INDIA (F2027E) As % of As % of As % of As % of Market Cap US$ Bn GDP US$ Bn GDP US$ Bn GDP US$ Bn GDP Financials 4,795 25.7% 2,431 21.7% 371 16.3% 1,845 30.5% Banks 2,044 11.0% 1,529 13.6% 226 9.9% 966 16.0% Insurance 1,062 5.7% 410 3.7% 10 0.5% 216 3.6% Asset Management 452 2.4% 135 1.2% 0 0.0% 77 1.3% Others 1,237 6.6% 357 3.2% 134 5.9% 586 9.7%

Source: Bloomberg, Morgan Stanley Research (E) estimates. USD/INR for F2017 we have used is 67

MORGAN STANLEY RESEARCH 61 M BLUEPAPER

Exhibit 102: India Financials Market Cap - Comparison with Large Economies as They Stand Now

G-7 Nations (Current) CHINA INDIA In US$ Bn ITALY GERMANY FRANCE CANADA JAPAN UK USA (Current) (F2027E)

Financials Market Cap 222 273 374 645 688 788 4,795 2,431 1,845 % of GDP 12% 8% 15% 42% 14% 30% 26% 22% 31%

Source: Bloomberg, Morgan Stanley Research (E) estimates

Exhibit 103: Exhibit 104: Indian Financials: Market Cap to GDP Market Cap Progression: Private Banks Will Dominate Market Share Mkt Cap (Rs Trillion) As % of GDP, RS Formation in Nominal Terms 30.0 21.0% 2000 SOE Banks Private Banks 18.0% NBFCs + HFCs AMCs 25.0 Life Insurance (ex. LIC) General Insurance Others 1500 15.0% 20.0

12.0% 1000 15.0

9.0%

10.0 500 6.0%

5.0 3.0% 0 F2017 F2027E

0.0 0.0% Source: Bloomberg, Morgan Stanley Research estimates. Note: Data in US$ Bn F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

Source: Bloomberg, Morgan Stanley Research

Indian financial stocks have been strong performers over the past Exhibit 105: two decades. Their market cap increased by >80x over this period as Indian Financials: Change in Market Cap, Rolling 12 Months YoY the underpenetration of the financial services industry was reduced. 160%

Even over the last 10 years (from elevated levels before the global 120% financial crisis as the average P/E of the sector in February 2008 was 31x), the market cap of Indian banks has increased almost 5x (private 80% banks have seen a 7x increase). 40%

0% We expect continued strength for the financial services sector over the decade to come. In this section, we look at the lending, insurance -40% and capital markets-related businesses separately and conclude that -80% they can achieve a combined market cap of US$1.8trn in 10 years – a Jun-00 Jun-01 Jun-02 Jun-03 Jun-04 Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 Jun-15 Jun-16 Jun-17 Dec-99 Dec-00 Dec-01 Dec-02 Dec-03 Dec-04 Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 Dec-15 Dec-16 fivefold increase from the F2017 level. This will come primarily from Source: Bloomberg, Morgan Stanley Research growth and new listings across the financials space.

We continue to expect a shift in market share towards private players (banks, NBFCs and other financial businesses), with SOE banks likely to be pushed further out to the fringes.

62 M BLUEPAPER A point to note here is that while the structural story for Indian finan- Exhibit 106: cials is very strong, there will be cyclical downturns that may be fairly Indian Banking System's ROA Has Compressed Sharply Since F2012 aggressive. Since 2000 Indian financials have delivered very strong Due to NPLs returns, but there have also been periods of big corrections. The next 1.5% 10 years will likely be similar, with strong returns but significant vola- 1.2% tility. Given the structural thematic we would advise investors to buy any dip in the non-SOE part of the financial system. 0.9%

0.6%

Banks, NBFCs and HFCs – Upside Ahead, 0.3%

Driven by Growth and Share Gains for 0.0%

Better Lenders F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017E

Source: RBI, company data, Morgan Stanley Research (E) estimates An interesting thing has happened over the past five years. System's Exhibit 107: return on assets fell to 0.4% from 1% – and yet the banks' combined Profit Share of Bank Groups market cap rose from US$135bn in 2012 to US$265bn as of August SOE Private Foreign 2017. This has essentially been driven by the big shift in market share 200% towards the private players, which trade at much higher multiples. 150%

As a result system P/E multiples for the sector have increased. 100%

50% Over the next 10 years, we expect banks' loan growth to increase by 0% 12% annually – implying an average credit multiplier (nominal credit -50% growth as a percentage of nominal GDP growth) of ~1x. As mentioned in an earlier section, this will primarily be driven by loan -100%

growth in the consumer and MSME sectors. Corporate loan growth F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017E is likely to be fairly muted in comparison. Source: RBI, company data, Morgan Stanley Research (E) estimates

Exhibit 108: We expect the SOE banks' loan market share to fall to around 35% by Indian Financials: Our Expectation of Loan Growth Over the Next 10 F2027 from 65% in F2017 (this is market share within the banking Years... sector). This is driven by our view that the balance sheets of most % CAGR Rs Bn F07 F17 F27E SOE banks will remain fairly challenged, with capital being infused in F17-27E a trickle-down fashion. We expect the credit growth of SOE banks to SOE Banks 14,401 56,243 89,657 5% be 5% annually over the next 10 years, with the larger banks growing Private Banks 4,148 22,370 144,879 21% Foreign Banks 1,263 4,181 13,650 13% at 7-9% but some smaller ones contracting their loan books. Private NBFCs 1,366 11,556 48,648 15% banks, on the other hand, are likely to see annual loan growth of 21%. HFCs 1,096 8,173 47,847 19% Overall, we forecast that credit growth in the banking system will TOTAL 22,275 102,524 344,681 13% average 12% annually in the next 10 years. GDP 42,947 152,781 442,821 11% Loans / GDP 51.9% 67.1% 77.8%

Source: RBI, NHB, company data, Morgan Stanley Research (E) estimates This is fairly conservative given Morgan Stanley's estimate of nominal GDP growth at a CAGR of more than 11%. But with SOE banks Exhibit 109: being unable to grow, the system overall will struggle to reach loan …As Well As Market Share of Lending growth of much more than 13% annually. Some of this additional % of Total F07 F17 F27E demand will likely be met through strong NBFC and HFC growth SOE Banks 65% 55% 26% (including fintech lenders in this space) and some will move towards Private Banks 19% 22% 42% the bond markets. Foreign Banks 6% 4% 4% NBFCs 6% 11% 14% HFCs 5% 8% 14%

Source: RBI, NHB, company data, Morgan Stanley Research (E) estimates MORGAN STANLEY RESEARCH 63 M BLUEPAPER So essentially a larger part of the banking system will be dominated most of the disruption due to technology has been limited to the pay- by the better-run – and therefore higher-multiple – businesses such ments space as mobile wallets take share from established payment as private banks, NBFCs and HFCs. We look at these segments sepa- gateways. However, moving ahead we see pressure on various rately from an earnings and market cap perspective. sources of revenue – MDR charges, distribution income, low-cost deposits, high-yield loans – which in turn will impact both net interest Banks – Asset growth, stable ROA and private-sector income and non-interest income. The exhibit below shows the level share gain will drive market-cap creation of disruption for key revenue items now and for the next few years.

From a profitability perspective, technology will impact banks in the form of lower revenue yield and lower costs. There are clearly a number of revenue streams where banks will earn less. As of now,

Exhibit 110: Areas of Disruption Versus Revenue Impact as Technology Usage Gets More Widespread Now In Next Few Years

Payments

Distribution - Simple products (MF, protection) > > ---

Distribution - Complicated products (Traditional Insurance)

CASA Level of Disruption

Mortgages Remittances

Unsecured Loans Impact on Revenues --->

Source: Morgan Stanley Research

Exhibit 111: Exhibit 112: Indian Banks: Commissions From MF Sales Will Be Under Pressure Indian Banks: Revenue From MDR Charges Is Another Source of Pres- Rs Mn As % of Fees, RS sure 5,000 10.0% 60.0

4,000 8.0% 50.0

3,000 6.0% 40.0

2,000 4.0% 30.0

1,000 2.0% 20.0

0 0.0% 10.0 IIB SBI YES AXIS ICBK HDBK 0.0 KOTAK

Source: Company data, AMFI, Morgan Stanley Research F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

Source: Company data, Morgan Stanley Research estimates

64 M BLUEPAPER Revenue yields will come off for the system. But, the positive side will Exhibit 113: be lower costs – both opex and credit costs. Against this backdrop, Indian Banking System Lacks Economies of Scale 250,000 how should we think of ROA for the system as a whole over the next Loan per Capita (in USD) Deposit per Capita (in USD) decade? We look at each of the key lines of ROA: 200,000

150,000 Net interest income: We expect yields on the overall loan book to be relatively stable over the next decade. Loan yields will come down 100,000 across products as underwriting becomes easier and better. How- 50,000 ever, as the loan mix shifts towards MSME/consumer loans from the 0 HK

current domination by corporate loans, there will be some net India China Korea Taiwan Thailand Australia Malaysia Indonesia Singapore interest margin (NIM) support. Philippines Source: Central banks, CEIC, Morgan Stanley Research The other risk to net interest income (NII) will come from competi- Exhibit 114: tion for deposits. Since the RBI deregulated saving account rates, HDFC Bank: Sharp Increase in Balance Sheet Per Employee and Per most banks have stayed away from competing aggressively for these Branch deposits. Some of the smaller banks have offered higher rates (and Employees (Number) Assets / Employee (Rs Mn), RS 100,000 125 gained significant share) but larger players have generally stayed away from competing. As it becomes easier to move money between 80,000 100 banks, larger banks may have to start thinking more about competition. 60,000 75

We are assuming a contraction in NII/assets from 2.9% in F2012 to 2.7% by F2027. 40,000 50

Fees: This is the part of revenues that will likely come under pressure 20,000 25 from the increased use of technology. Clear pressures will be felt in fees from the distribution of financial products – mutual funds, insur- 0 0

ance, and so on. We expect a larger share of these products to be sold F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 through direct channels over the next decade. Cyclically, this busi- Source: Company data, Morgan Stanley Research estimates ness is booming for banks given the sharp pick-up in equity markets. But structurally this will likely decline. 2) The implementation of technology in almost all facets of financial services will reduce the requirement of branch expansion. We have Various other parts of fees like MDR should also reduce over time – already started seeing signs of this, with branch sizes getting smaller, however, a pickup in volumes (which will be much faster than asset not necessarily occupying premier retail space, and cuts in manpower growth) should help this part of fees to be stable as a percentage of requirements. assets. Costs as a percentage of assets for Indian banks are among the Costs: Indian banks run fairly subscale businesses. This is driven by highest in Asia, and we expect banks to easily take out 40-50 bps of the country's relatively scattered branch network given the vast assets over the next decade. This would imply that underlying ROA geography and smaller banking system pie per capita. Over the next should be fairly stable for banks. There will be pressure on revenues 10 years, banks will be in a position to control this meaningfully. for sure - but cost control will help drive profitability growth.

1) As GDP expands, banking system assets should expand at a CAGR This is evident in the case of HDFC Bank, which is at the forefront of of 12%. This will increase economies of scale. This increase in the size offering digital solutions. Over the last three years, its balance sheet of the system balance sheet itself will also help improve costs to has grown at a CAGR of 21% while its branch network has grown at assets. But this will be accentuated by the rapid increase in the digiti- an 11% CAGR and its employee base at a 7% CAGR. In fact, since Sep- zation of banking services. tember 2016, its employee base has declined by ~10,000 people despite the balance sheet expanding by 8%.

MORGAN STANLEY RESEARCH 65 M BLUEPAPER This is likely to cause costs as a percentage of assets to come off fairly Exhibit 115: meaningfully. Hence, while there are pressures on revenues, the pri- HDFC Bank's Versus SOE Banks' Cost Income Ratio: Sharp Divergence vate banking sector as a whole should be able to reduce costs to meet Between Winners and Laggards Ahead HDBK SoE Banks these pressures and keep ROAs broadly stable. This, coupled with 65% strong asset growth, will help drive earnings progression. 60%

We expect banks with strong digital adoption to gain share in reve- 55% nues and control costs very effectively. Carrying on the HDFC Bank 50% example from above, if we compare HDFC Bank's cost income ratio 45% evolution over the last five years with SOE banks' (where costs are 40% sticky given state ownership), the differential will likely widen to 15-20% points over the next two to three years. This exemplifies how 35%

the gap between winners and "data laggards" will likely expand. F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017E F2018E F2019E

Source: Company Data, Morgan Stanley Research. Note: We have taken core income as NII+CEB We also build in normalized credit costs at around 50 bps of assets. Currently , credit costs are running at elevated levels given the corporate credit cycle - which has been going for the past few years. This Exhibit 116: has impacted corporate lenders - IFRS implementation should make Indian Banks: Credit Costs (as Percentage of Loans); We Are Building banks provide for bad loans quickly. The retail lenders have main- in Normalization at ~85 bps of Loans tained credit costs at 50-60 bps and stronger data can keep costs 250 lower for longer. For F2027, we assume credit costs at 85 bps of loans 200 (50 bps of assets).

150 Banking sector's market cap will likely rise to US$965bn: The above estimates for asset growth, ROA and mix shift imply that 100 profits for the Indian banking system will be around Rs4.7trn by 50 F2027. This compares with Rs550bn in F2017 – a year that was hit by huge loan losses and a lack of profitability at SOE banks. In F2012, 0 profits for Indian banks were Rs817bn, implying a CAGR of 12% over F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 15 years. Source: Company Data, Morgan Stanley Research

More importantly we expect the profit share gains of private banks to grow even faster than their likely gains in asset share. This will be Exhibit 117: driven by stronger profitability as costs are controlled even though Indian Banks: We Expect the Industry to Earn an ROA of 1.2% Over the revenue yield comes under some pressure. Five years back the SOE Next Decade, With Fees and NII Pressure Negated by Cost Efficiencies banks' share of system profits was around 60% – we expect this to % of average assets F2012 F2017 F2027E fall to around 20% over the next 10 years. Private banks, on the other Net Interest Income 2.9% 2.6% 2.7% Non-Interest Income 1.1% 1.2% 1.1% hand, should be about 70%, with foreign banks making up around Fee Income 0.7% 0.6% 0.6% 10% of system profits. Others 0.4% 0.6% 0.5% Total Revenues 4.0% 3.8% 3.7% Total costs 1.8% 1.8% 1.5% We assign P/E multiples to various types of banks – 8x for large SOE Pre-Provision Operating Profits 2.2% 2.0% 2.3% banks; 6x for smaller banks; 15x for private banks – and arrive at a Bad Debts 0.5% 1.3% 0.5% potential market cap of Rs71trn for the banking sector over the next Other Provisions incl Tax 0.7% 0.3% 0.6% ROA 1.1% 0.4% 1.2% 10 years compared to the current Rs15trn. Within the subset, we Leverage 14x 13x 11x expect the SOE banks' market cap to be up by about 1.6x – larger ROE 14.6% 5.3% 12.5% banks will likely do better but some of the smaller banks will find it Source: RBI, company data, Morgan Stanley Research (E) estimates tough to hold on to current market caps. Private banks is where the market cap pickup will be fairly material – we expect it to compound at around 19% in rupee terms.

66 M BLUEPAPER Foreign bank profits should also increase reasonably strongly over Exhibit 118: the 10-year period to Rs390bn (US$5.3bn). Within foreign banks, the Indian Profits* of Foreign Banks: Citi Has Been Most Consistent PAT (Rs Bn) profits above are based on regulatory filings and to that extent do not 40.0 capture the India linked earnings, which are booked abroad (for Citibank StanC HSBC instance, lending to an Indian corporate in US dollars from a foreign 32.0 branch). The bulk of these profits are concentrated in three banks - Citi, HSBC and Standard Chartered - with Citi delivering consistent 24.0 trend. This will likely continue, with DBS potentially becoming a 16.0 bigger player as Digibank gets rolled out.

8.0 We are assigning no market cap as the Indian entities may not be listed (unless they decide to go for a wholly-owned structure and list 0.0

– DBS has recently got approval in principle to open a WOS (wholly- F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 owned subsidiary) in India). But for these banks, implied market cap * This is per local regulatory filings. Source: Company Data, Morgan Stanley Research will be reflected in their parent businesses.

In our bull case, we have assumed 2.3ppt stronger loan growth com- Exhibit 120: pared to the base case, driven by stronger GDP growth, greater Market Cap of Indian Banks by F2027 – Scenario Analysis market share gains in retail/MSME lending relative to NBFCs, and Banks Base Bull Bear lower-than-expected competition (in fees and deposits) from fin- Assets CAGR (FY17-27E) 12% 14% 8% Assets (FY27E) techs. We have also assumed higher RoA by 10bps, driven by higher INR Bn 427,327 524,640 309,376 cost efficiencies and lower competition in margins. Consequently, USD Bn 6,375 7,827 4,615 this in our view will also help improve valuations to 16x versus 14x in ROA (FY27E) 1.2% 1.3% 0.9% our base case. We estimate a US$1.5trn market cap in our bull case, PAT (FY28E) 5,344 7,350 3,089 Forward P/E 13.8x 15.5x 11.8x 55% higher than our base case. MCAP (FY27E), Listed INR Bn 70,794 109,326 34,642 In our bear case, we have assumed 3.5ppt slower loan growth com- USD Bn 966 1,491 473 pared to the base case, driven by weaker GDP growth and greater Source: RBI, company data, Morgan Stanley Research (E) estimates. To arrive at market cap, we have first computed the valuations of the standalone entities (PAT * core P/E as we see in the above table) and competition from NBFCs and fintechs. More importantly, we have then added valuation of stakes in other entities. assumed lower ROA by ~25bps versus our base case, driven by significant compression in fee income and also margins (lower float and hence moderation in CASA ratios). Consequently, this in our view would also reduce valuations to 12x versus 14x in our base case. We estimate a US$475bn market cap in our bear case, 50% lower than our base case.

Exhibit 119: Indian Banks – Total Market Cap of Private and SOE Banks Could Reach Rs71trn Over the Next 10 Years, With Private Banks Gaining More Share PAT PE Market Cap Rs. Bn 10 Year F2012 F2017 F2027E F2017 F2027E Average F2017 F2027E SOE 495 5 881 959.6x 8.8x 10.0x 4,546 8,120 Private 227 434 3,424 24.7x 15.5x 15.0x 10,706 62,674 Foreign 94 118 386 NA NA NA NA NA Total 817 557 4,691 18.7x 14.3x NA 15,252 70,794

Source: RBI, company data, Morgan Stanley Research (E) estimates. To arrive at market cap, we have first computed the valuations of the standalone entities (PAT * core P/E as we see in the above table) and then added valuation of stakes in other entities.

MORGAN STANLEY RESEARCH 67 M BLUEPAPER NBFCs and HFCs – Strong outlook driven by market Exhibit 121: growth and share gains NBFC and HFC Loan Book Grew at a 23% CAGR in FY07-17, to Rs19.5trn This is one space that has continued to do well over the last few years. NBFC+HFC Loan Book (Rs Bn) A number of market participants are usually concerned about the 20,000 sustainability of this space given the fact that these are wholesale- 16,000 funded. 12,000

Because of these concerns, these stocks tend to be fairly volatile. 8,000 Whenever the market slows, these stocks tend to have big draw- 4,000 downs as concerns around asset quality and NIMs increase. 0

However, despite these concerns and the slowdown seen in India F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 over the last five years by the broader financial system, this space has Source: Company data, Morgan Stanley Research seen loans increase from Rs2.5trn in F2007 to Rs19.5trn in F2017 – a CAGR of 23%. Consumer/SME NBFCs: These NBFCs will also face pressure on revenue yields, but we expect lower costs led by technology rollouts Given the likely increase in loan demand over the next 10 years and and a reduction in credit costs from all-time highs for some of the the fact that SOE banks will be unable to grow by much more than more cyclical NBFCs to result in slightly better ROAs than FY17 5% given capital issues, NBFCs and HFCs will likely grow at a CAGR levels. of 17% – taking their share in the system to 28% from 19% in F2017. Within this segment, the corporate NBFCs – dominated by infra- Diversified NBFCs: These NBFCs, being more diversified with structure lenders – will likely lag, in our view, with consumer/SME respect to both lending (mix of corporate and retail loans) and rev- NBFCs, diversified NBFCs and HFCs growing much faster. enue streams (non-lending businesses like asset management,wealth management, broking, investment banking, insurance, With respect to profitability, below are our views on various seg- and so on), are better positioned to counter both the competition and ments: cyclicality in businesses. We are building in some NIM compression and assuming some credit cost normalization from low levels, HFCs: In our view competition in home loans will continue to inten- although we expect this to be offset by operating leverage in both the sify, weighing on yields and NIMs. We are also seeing pricing competi- lending and capital markets businesses. This should result in some tion in developer loans intensify. Further, as data availability marginal ROA expansion from FY17 levels. improves and as more avenues open up for MSMEs to borrow, their dependence on Loans against Property will go down, which could Corporate NBFCs: We expect some improvement in ROA as better potentially impact the HFCs, which generally do only secured quality credit is written, resulting in lower credit costs offset to a lending. Hence, we have assumed compression in ROAs from FY17 large extent by lower yields. levels. Credit costs and operating costs are low and hence there is limited scope for meaningful change.

68 M BLUEPAPER

Exhibit 122: We Expect Profitability at HFCs to Decline on Lower Margins due to Competition. We Expect Overall Profitability at NBFCs to Improve, Driven by Operating Leverage at the Consumer/SME/Diversified NBFCs and a Reduction in the Share of Large Infrastructure Loans in the Overall NBFC Loan Book A) HFCs B) NBFCs (Aggregate of C,D,E) C) Consumer / SME NBFCs D) Diversified NBFCs E) Corporate NBFCs As % of Average Assets F2012 F2017 F2027e F2012 F2017 F2027e F2012 F2017 F2027e F2012 F2017 F2027e F2012 F2017 F2027e Net Interest Income 3.1% 3.0% 2.7% 5.2% 5.6% 5.4% 7.7% 8.0% 7.1% 3.9% 4.8% 4.3% 4.0% 4.1% 3.1% Non-Interest Income 0.5% 0.3% 0.3% 1.3% 1.8% 2.6% 0.5% 0.9% 0.9% 6.3% 6.4% 6.4% 0.1% 0.2% 0.2% Total Revenues 3.6% 3.3% 3.0% 6.5% 7.4% 8.0% 8.2% 8.8% 8.0% 10.2% 11.2% 10.7% 4.1% 4.3% 3.3% Total costs 0.5% 0.5% 0.5% 2.5% 3.0% 3.7% 3.2% 3.6% 3.0% 7.0% 7.2% 6.6% 0.4% 0.3% 0.3% PPOP 3.1% 2.8% 2.6% 4.0% 4.4% 4.3% 4.9% 5.2% 5.0% 3.2% 3.9% 4.1% 3.7% 4.0% 3.0% Provisions 0.2% 0.3% 0.3% 0.6% 1.5% 1.1% 1.4% 1.9% 1.6% 0.6% 0.6% 0.8% 0.1% 1.6% 0.5% PBT 2.9% 2.6% 2.3% 3.4% 2.9% 3.3% 3.5% 3.3% 3.4% 2.6% 3.3% 3.4% 3.6% 2.5% 2.6% Taxes 0.8% 0.8% 0.8% 1.1% 1.0% 1.1% 1.4% 1.1% 1.2% 0.9% 1.1% 1.2% 1.0% 0.9% 0.9% ROA 2.1% 1.8% 1.5% 2.3% 1.9% 2.1% 2.1% 2.1% 2.3% 1.7% 2.2% 2.2% 2.6% 1.6% 1.7%

Source: Company data, Morgan Stanley Research; e = Morgan Stanley Research estimates. The above computations are for underlying businesses and excluding numbers of holding companies

Exhibit 123: Exhibit 124: HFCs: One-year Forward P/E NBFCs (ex HFCs): One-year Forward P/E 25.0 35.0 1Y Fwd P/E 1Y Fwd P/E 30.0 20.0 25.0

15.0 20.0

15.0 10.0

10.0 5.0 5.0

0.0 0.0 Feb-08 Feb-09 Feb-10 Feb-11 Feb-12 Feb-13 Feb-14 Feb-15 Feb-16 Feb-17 Feb-08 Feb-09 Feb-10 Feb-11 Feb-12 Feb-13 Feb-14 Feb-15 Feb-16 Feb-17 Aug-07 Aug-08 Aug-09 Aug-10 Aug-11 Aug-12 Aug-13 Aug-14 Aug-15 Aug-16 Aug-17 Aug-07 Aug-08 Aug-09 Aug-10 Aug-11 Aug-12 Aug-13 Aug-14 Aug-15 Aug-16 Aug-17

Source: Thomson Reuters, company Data, Morgan Stanley Research Source: Thomson Reuters, company Data, Morgan Stanley Research

These players usually trade at a discount to private banks given the Exhibit 125: lack of a retail funding franchise. We assume the same trend will con- With Diversification in Revenue and Profit Streams and an Increase in tinue despite fairly strong growth. the Share of Better-run Businesses, We Expect an Overall Forward P/E Multiple of 14x for NBFCs and HFCs Versus a 10-Year Average of 12x

Profits Forward P/E Market Cap In our base case, we assume that NBFCs will sustain a valuation mul- Rs. Bn 10 Year F2012 F2017 F2027e F2017 F2027e Average F2017 F2027e tiple of 14x forward earnings (compared with a 10-year average of NBFCs 117 235 1,116 14.2x 15.7x 11.0x 4,717 20,142 Consumer / SME NBFCs 39 90 538 17.9x 15.0x 12.9x 2,010 9,277 11x), driven by diversification of business and an increasing share of Diversified NBFCs 10 60 402 14.6x 14.9x 22.5x 1,753 8,756 Corporate NBFCs 68 86 175 9.4x 10.0x 7.2x 954 2,108 more profitable businesses, and HFCs trade at 14x forward earnings HFCs 64 139 761 15.0x 14.1x 14.9x 3,697 18,363 Total NBFCs / HFCs 182 375 1,877 14.5x 17.6x 12.2x 8,414 38,505

(10-year average of 15x) – implying a potential market cap of Rs39trn. Source: Company data, Morgan Stanley Research; e = Morgan Stanley Research estimates. To arrive at market cap, we have first computed the valuations of the standalone entities (PAT * core P/E as we see in the above table) and then added valuation of stakes in other entities. In our bear and bull cases, we assume a 2.5ppt lower/higher CAGR compared to the base case, key drivers being economic growth, the Exhibit 126: credit cycle and competitive intensity from banks. We have assumed NBFCs and HFCs: Valuation Scenarios an ROA differential of 20-25bp (mainly driven by pricing power and NBFCs + HFCs Base Bull Bear NIM) across the scenarios and a P/E multiple differential of 2x. In our Assets CAGR (FY17-27E) 17% 20% 12% base case, the combined market cap of NBFCs and HFCs works out Assets (FY27E) INR Bn 109,438 135,058 70,877 to US$525bn versus US$795bn in the bull case and US$260bn in the USD Bn 1,493 1,842 967 bear case. ROA (FY27E) 1.8% 1.6% 2.0% PAT (FY28E) 2,183 2,986 1,179 Forward P/E 14.2x 16.2x 11.7x MCAP (FY27E) INR Bn 38,505 58,129 19,060 USD Bn 525 793 260

Source: Morgan Stanley Research (E) estimates

MORGAN STANLEY RESEARCH 69 M BLUEPAPER Thus, in our base case, the lending businesses – banks and NBFCs – Exhibit 127: will likely go from a market cap of Rs24trn now to Rs109trn in F2027. Expect Insurance Penetration to Stabilize 5.0% Almost the entire increase in market cap is being driven by earnings First Year Premium Total Insurance Premiums (% of GDP) 4.2% Renewal Premium progression and a gain in share by better-run businesses. 4.0%

2.3% 2.8% 2.7% 3.1% 3.0% 2.2% 1.7% Insurance (Life & General) to See Strong CAGR 1.7% 2.0% CAGR 1.7% CAGR 1.6% 12% CAGR 10% 23% 8% CAGR 1.6% Growth and a Spate of New Listings 24% 1.2% CAGR CAGR 1.0% 1.9% CAGR CAGR 4% 13% 14% 1.3% 41% 1.1% 1.1% CAGR 0.7% 0.5% 32% 0.0% Until 2016 there was only one pure play on the insurance sector in F2001 F2003 F2008 F2014 F2017 India (Max Financial Services). Subsequently, ICICI Prudential Life F2027E Phase I - Entry of Phase II - Expansion Phase III - Phase IV - Growth Downturn in ULIPs Insurance got listed and now there are multiple life and general Private Players of Private Players Revival Source: IRDA, company data, Morgan Stanley Research; E = Morgan Stanley Research estimates insurers which have filed DRHPs with the intent of going public. The market cap in this space is fairly small right now given that there are Exhibit 128: only two listed entities – but we expect a sharp pickup driven by Large Bank-backed Insurers to Grow Faster Given Better Operational growth and new listings. Metrics 90% We have kept LIC, India's biggest insurer, out of this analysis because ICICI Pru

there is not enough data on its operations. SBI Life Max Life

2017 Kotak Life 80% Life insurance – We expect a market cap of US$126bn LIC Birla Sunlife in 10 years (excluding LIC) 70% Reliance Life Bajaj Life

Insurance penetration has stabilized over the past two years after F Persistency, th Month 13 declining in F2009-14. From a top-down perspective, this has been 60% driven by the stabilizing regulatory environment and the improving 5% 10% 15% 20% 25% 30% share of financial savings. Adjusted Cost Ratio, F2017 Source: Company data, Morgan Stanley Research

Incrementally, our view is that significant regulatory changes are likely done, and the shift towards financial savings is not only cyclical However, large bank-backed insurers will grow faster (15-18% CAGR) but also structural - given formalization of savings. This, coupled with than nominal GDP, driven by market share gains. They have access to relatively low penetration, should help the insurance sector to grow banks and good operational metrics – both of these are critical in the at ~13% CAGR over the next decade compared to nominal GDP new normal environment. While the regulator has allowed open growth of 11%. architecture, we expect gradual moves by the large banks. On profitability, the current differential between large bank-backed insurers and other private insurers is quite high and we see it being sustained at elevated levels as the large bank-backed insurers increase their protection mix and leverage technology to improve efficiency and cross-selling.

70 M BLUEPAPER Technology will create the next leg of differentiation between Exhibit 129: private players Focus on Protection, Which Is Under-penetrated, Will Drive Profitability 3.0

US Japan We expect a big change in the way insurance is distributed over the 2.5 next 10 years, led by technology. The current generation of insurance Singapore 2.0 buyers (35-40 years old) has a significant portion of customers who Korea are less tech savvy and still visit branches for banking services, where 1.5 Malaysia Germany a lot of insurance sales pitches actually take place. 1.0 Thailand

Sum Assured Sum Assured / GDP (2013) India 0.5 However, over the next 10 years, the prospective insurance buyer could be quite different and much more digital. Indeed, branch foot- 0.0 0 500 1000 1500 2000 2500 3000 3500 falls will come down significantly. This is a key challenge for private Life Insurance Density (Premiums per capita in USD), 2014 insurers as 50-60% of their insurance sales happen via bancassur- Source: Swiss Re and ICICI Prudential Life Presentation, Morgan Stanley Research ance. In this context, building a relationship with prospective customers via strong digital infrastructure becomes very important. Exhibit 130: Private insurance companies are creating infrastructure for this Market Share Comparison - Sum Assured Versus APE

75% shift Market Share (%) Sum Assured APE

60% Online sales for simpler products have been picking up – term insurance has picked up well and we expect low cost ULIP products to 45% follow: For relatively simple/pull insurance products like protection, online insurance penetration has picked up quite significantly. This 30% could be sales via a company's own online portal, partner or aggre- 15% gator websites and even call centers (assisted online sales). How- ever, online sales are significantly under penetrated for most other 0%

savings products given that these are relatively complex products. F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 F2013 F2014 F2015 F2016 F2017 LIC IPru Life SBI Life Max Life Kotak Life Reliance Life Birla Sunlife Bajaj Life

Source: Company data, Life Insurance Council, Morgan Stanley Research. Note: Data is for F2017. Sum For relatively less simple or push products, digitization could assured system data does not include Sahara Life and India First Life. empower distributors with better lead generation and sales – Key initiatives that insurers are working on include a) upselling insurance policies to existing customers at the time of inbound calls, policy maturity and also based on analytics; b) innovative digital marketing tools via social media, viral marketing and online display banners; c) integrating digital platforms with those of distributors – eg, straight through processing integration; d) emerging financial marketplaces and moving away from just placing products in partner ecosystems with differentiated pricing and promotion schemes for customers using analytics.

MORGAN STANLEY RESEARCH 71 M BLUEPAPER As distribution evolves, we expect private players which are more Exhibit 131: nimble to continue to gain significant share from state-owned entity. Private Life Insurance Valuations Life Insurance Base Bull Bear EV CAGR (FY17-27E) 13% 16% 10% We expect the market cap of the private life insurance sector to EV (FY28E) 3,533 4,715 2,627 rise to US$126bn by F2027 and for the top five private bank- Forward P/EV 2.6x 3.0x 2.0x MCAP (FY27E) based insurers to contribute 89% of that. A sector valuation exer- INR Bn 9,247 14,226 5,299 USD Bn 126 194 72 cise is not easy given the lack of detail around EV for all players. In our Source: Company data, Morgan Stanley Research (E) estimates exercise, we have estimated EV for all private players based on a) reported net worth, and b) our estimates of VIF/policyholders' AUM General insurance – We expect a US$90bn industry in for players where EV is not reported. We estimate the private sector 10 years EV to be US$14bn as of F2017, with the top five bank-based private life players contributing 57% of that. We believe the general insurance industry can grow at a 15% Our view is that the top five bank-based private players (combined) CAGR (following a 13% CAGR over the past five years), even after could compound EV (pre dividends) at 15-20% over the next 10 years, which penetration will remain low at 1.0% of GDP (F2027) driven by a) strong premium growth and b) improving profitability helped by an increasing share of protection and greater economies of As with most financial services, general insurance penetration in scale. India is low at 0.8% (the world average was 2.8% in 2016) and is supported by favorable demographics and stable to improving economic For other private life insurers, we have assumed an RoEV (pre divi- growth. Since 2000, the sector has grown through three broad dends) CAGR of 10%, much higher than the low single digits we cur- phases: a) Phase 1 (2000-07): High growth for private players helped rently estimate. The improvement is driven by a) a cyclical uplift to by the liberalization of the sector; b) Phase 2 (2007-11): Industry con- growth, even as they lose market share, and b) an improvement in solidation with detariffing of key products and high losses in Indian profitability given lower cost overruns. However, we expect RoEV to Motor Third Party Insurance Pools (IMTPIP); and c) Phase 3 (2011 remain below CoE for multiple years, driving our low valuation for onward): Improving profitability helped by tariff revisions in third this group. party motor insurance and new growth segments like crop insurance. Incrementally, we believe there are two key catalysts: 1) the Motor The bull case for the sector is a surprise recovery in the agency Vehicles (Amendment) Bill 2016, which can be a significant catalyst channel and potential regulatory changes that relax some of the to growth and profitability as it intends to a) fix a maximum time limit stricter product regulations. Another driver would be higher-than- for claim requests of six months (no limitation now), b) fix maximum expected financial savings growth and a stronger-than-expected insurer liability on claims, and c) encourage renewal registrations of pickup in protection. We assume stronger EV growth (pre dividends) motor vehicles; and 2) Digitization aided reduction in combined by 3ppts and also expect valuations to be sustained at higher levels. ratios. Another potential catalyst of profitability that needs to be watched is a potential for increased focus on profitability at PSU The bear case for the sector is a) significant margin compression in insurers given the government's drive. the protection business, b) a significant growth slowdown in bancas- surance sales, driven mainly by lower footfall, and c) a potential increase in corporate income tax rates. Consequently, we assume weaker EV growth (pre dividends) by 3ppts compared to our base case. Against this backdrop, valuations moderate to 2.0x EV vs. 2.6x in our base case.

72 M BLUEPAPER Leading private players are better placed on distribution, under- Exhibit 132: writing/product mix and also technology General Insurance Penetration Non Life Insurance Penetration, 2016 (%)

5.0 4.7 To start with, the leading private players (ICICI Lombard, Bajaj Gen- 4.3

4.0 3.5 eral, HDFC Ergo and so on) have a better product mix (more retail), 3.3 3.3 3.2 2.8 distribution mix, underwriting capabilities, and more conservative 3.0 2.7 2.6 2.4 2.0 reserving practices. Consequently, their combined ratio is much 2.0 1.8 1.8 1.4 lower than peers'. This should continue to help them gain market 1.3 1.0 0.8 share and improve profitability. 0.0 UK Italy USA India Brazil Spain China Korea Japan Ireland France Moreover, we believe the general insurance industry could see a Taiwan Australia Germany Hong Kong Hong meaningful impact from technology adoption. In our view, tech- Africa South nology will play an even greater role in a) the underwriting process Source: Swiss Re and policy selection and b) reducing costs. Unless competed away, Exhibit 133: the successful players could reduce combined ratios by at least 2-3% General Insurance Penetration to Improve to 1.0% of GDP over the next 10 years. The key benefits of technology could be as Non Life Insurance Penetration, India (%) follows: 1.2%

1.0% 1.0% a) Lower operating costs: Insurance is a cost-intensive business 0.8% (with F2017 cost ratios of 20-40%), and digital platforms could be 0.8% used to reduce cost of acquisitions, improve turnaround times, 0.6% survey claims digitally (via video streams for motor claims or drones 0.4% for wind turbines, and so on), and automate sales and customer service (chatbots, other apps). 0.2% F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F2027E b) Distribution: Not only can a greater amount of insurance policies Source: General Insurance Council, Morgan Stanley Research (E) estimates be sourced digitally, but insurers can partner with various manufac- turers/dealers and provide end-to-end digital platforms for product innovation/solutions (mobile-based cashless claims, virtual surveys) Exhibit 134: – and this can help build sticky relationships. This also applies to Combined Ratios banking and other distribution tie-ups. Claims Commission Expense Combined Ratio (F2017) 150%

120% c) Better risk management systems and underwriting: Insurers 90% can utilize artificial intelligence and machine learning techniques to 60% improve underwriting practices amid increasing data availability. 30% 0%

-30% In our base case, we assume that industry RoEs average 10% over the

next 10 years, and we also assume a P/E of ~15x. For the private Reliance HDFC Ergo HDFC Bajaj Allianz Bajaj IFFCO Tokio IFFCO players, we assume higher P/E multiples given better growth and Lombard ICICI United Insurance United profitability outlook. Our base case assumptions yield a valuation for Insurance National New India Assurance India New the industry of US$90bn in F2027. Source: Company data

MORGAN STANLEY RESEARCH 73 M BLUEPAPER The bull case for the industry would be a) significant changes in com- Exhibit 135: petitive practices by large public insurers, b) higher-than-expected General Insurance Market Cap Computations benefits from technology changes, and c) the motor insurance bill General Insurance Base Bull Bear PAT CAGR (FY17-27E) 27% 29% 24% getting cleared by IRDA, reducing the claim intimation period. This PAT (FY28E) 446 557 340 would imply stronger RoEs (11% versus 10% in our base case) and Forward P/E 14.7x 17.2x 12.2x MCAP (FY27E) higher P/E multiples (17x versus 15x). Consequently, the valuation of INR Bn 6,572 9,602 4,153 USD Bn 90 131 57 general insurance in our bull case would be as high as US$131bn, 46% Source: Company data, Morgan Stanley Research (E) estimates above our base case.

The bear case for the industry would be high and increasing competi- Exhibit 136: tive intensity, not only by the public sector, but also within the large Overall Domestic AUM/GDP (2016) players as they all eye profitable sub-segments and compete aggres- 150% Overall AUM / GDP (2016) 133% sively. This would imply weaker RoEs (8.5% versus 10%) and lower 120% 101%

P/E multiples (12x versus 15x). Consequently, the valuation of gen- 90% 78% 76% 59% 57% 56% 50% eral insurance in our bear case is US$57bn, 37% lower than our base 60% 29% case. 30% 12% 11% 10% 9% 0% UK USA INDIA

Mutual Funds – Strong AUM Growth and CHINA SWISS BRAZIL KOREA TAIWAN MEXICO FRANCE GERMANY Listings Over the Next 10 Years AUSTRALIA SOUTH AFRICA SOUTH DMs EMs

The fund industry in India has seen significant AUM growth over the Source: IMF, IIFA, Morgan Stanley Research past decade (with a F2007-17 CAGR of 22%) and the past five years (F2012-17 CAGR of 24%). The past three years have seen an accelera- Exhibit 137: tion (F2014-17 CAGR of 29%). Moreover, the composition of AUM Domestic Equity AUM/GDP (2016) 56% Equity AUM / GDP (2016) has moved towards equity (F2014-17 CAGR of 43%). Equity has 60% 50% increased to 33% of AUM from 21% three years ago. 45%

26% 30% Despite this strong growth, the asset management industry remains 24% 13% 12% 15% 8% very underpenetrated relative to other countries. Overall domestic 4% 3% 3% 3% 1% 1% AUM/GDP was at 9% as of 2016 and equity AUM/GDP was at 3% – 0% UK one of the lowest levels among large countries. USA INDIA CHINA SWISS KOREA BRAZIL TAIWAN MEXICO FRANCE GERMANY AUSTRALIA SOUTH AFRICA SOUTH DMs EMs

Source: IMF, IIFA, Morgan Stanley Research

74 M BLUEPAPER We expect the growth in AUM to be sustained at a very strong pace. Exhibit 138: This is being driven by higher financialisation of savings – a trend that Systematic Flows Gaining Share will accelerate, as discussed earlier in the report. Apart from a pickup SIP Flows Flows in Equity MFs 1,000 887 in the economy and stronger financial savings growth, which we have 839 discussed in other places, two trends will help AUM growth. 800 744

600 a) Rising share of systematic flows – Flows into systematic invest- 439 400 317 ment plans have logged a CAGR of 30% in the last five years. The 205 200 118 143 145 share of SIP flows in overall equity flows has been 28%, 38% and 49% 52 92 18 3 in the last three years. Given that these are regular monthly invest- 0 -103 ment plans, these are fairly predictable and are good for long-term -200 -125 -145 growth. F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 Source: AMFI, Morgan Stanley Research b) Rising contribution from smaller cities – In the past, mutual fund flows and AUM have been largely concentrated in the top cities. Even today, the top five cities account for over 70% of mutual fund AUM. However, in recent years the share of cities beyond the top 15 cities has been growing.

This is happening amid extremely strong growth in the top 15 cities. In our view, the share of the smaller cities in equities could have picked up even more sharply, given that debt flows are more likely to be from institutions and HNIs in urban areas. Technology will continue to drive investors in these smaller cities towards mutual funds as it becomes easier to offer them products in a cost-effective way.

Exhibit 139: Contribution of B-15 Cities Has Risen Despite Strong Growth in the Top 15 Cities

AUM by Geography - Consolidated F2013 F2014 F2015 F2016 F2017 data for MF Industry Total AUM - Industry (Rs bn) 7,014 8,252 10,828 12,326 17,529 Top 5 Cities (Metros) 5,193 6,018 7,786 8,897 12,694 Next 10 Cities 900 1,108 1,477 1,675 2,254 Beyond Top 15 Cities 921 1,126 1,565 1,754 2,580 As % of Total Industry AUM Top 5 Cities (Metros) 74.0 72.9 71.9 72.2 72.4 Next 10 - (Tier 1 Cities) 12.8 13.4 13.6 13.6 12.9 Beyond Top 15 Cities 13.1 13.7 14.5 14.2 14.7 YoY Growth (%) Top 5 Cities (Metros) 24% 16% 29% 14% 43% Next 10 Cities 9% 23% 33% 13% 35% Beyond Top 15 Cities 6% 22% 39% 12% 47%

Source: AMFI, SEBI, Morgan Stanley Research. Note: B-15: Beyond top 15 cities.

MORGAN STANLEY RESEARCH 75 M BLUEPAPER Our strategy team (Ridham Desai and Sheela Rathi) forecasts a 27% Exhibit 140: CAGR in equity AUM during F2017-27 – or a tenfold increase in con- We Expect the Mutual Fund Industry's Domestic AUM to Post a 23% stant currency terms, from US$100bn to US$1trn (discussed earlier CAGR in F2017-27; AUM/GDP Should Increase from 11% to 28% 160,000 32% in the report). We build in a 20% CAGR in debt AUM during the same AUM - MF Industry, Rs bn (LHS) as % of GDP (RHS) period, compared with a 25% CAGR in the last decade. This translates 140,000 28% 28% 120,000 24% into a 23% CAGR for overall AUM in INR terms. Adjusted for currency 22% 100,000 20% depreciation, we forecast overall AUM at US$1.85trn by F2027 (a 80,000 16% 22% CAGR). AUM/GDP should increase from 11% to 28% over the 11% 135,920 60,000 12% coming decade. 7% 40,000 5% 8% 59,735 20,000 4% 2,319 5,872 17,529 Industry valuation: As of now there are no listed asset management - 0% FY07 FY12 FY17 FY22e FY27e companies. Some companies have already expressed an intent to list Source: AMFI, Morgan Stanley Research; e = Morgan Stanley Research estimates and have moved in that direction.

1. 'Reliance Nippon Life AMC files IPO papers' Exhibit 141: 2. 'UTI Asset Management Company fairly ready for IPO, says Assuming CAGRs of 27% and 20% for Domestic Equity AUM and Debt MD Leo Puri' AUM Over the Next Decade, the Share of Equity in Overall AUM Would Rise to 47% Versus 33% in F2017 Debt Funds Equity Funds It is possible that more leading AMCs in India may list over the next 100% decade. Most of the bigger ones are owned by financial conglomer- 32% 33% 80% 40% ates, and there may be continued moves towards price discovery 45% 47% through IPOs. 60%

40% What could be the valuation for AMCs? Given that there are no listed 68% 67% 60% 55% 53% entities in this space, there is no peer group in India to compare valua- 20% tions with. However, there have been quite a few secondary market 0% deals over the years (the last being the sale of a stake in Reliance FY07 FY12 FY17 FY22e FY27e Mutual Fund to Nippon Life in October 2015). Some of these deals Source: AMFI, Morgan Stanley Research; e = Morgan Stanley Research estimates have been done at valuations of around 5-6% of AUM. These are higher than global averages. But given the stronger growth Exhibit 142: prospects in India, it is possible that these levels may be sustained for AMC Industry: Valuation Scenarios several years. AMCs Base Bull Bear AAUM CAGR (FY17-27E) 23% 30% 13% AAUM (FY27E) We use a base case valuation of 4.0% of AAUM to arrive at an INR Bn 141,869 250,364 64,419 USD Bn 1,935 3,415 879 industry valuation of US$77bn by F2027. Our bull and bear case sce- Valuation Multiple (% of AAUM) (FY27E) 4.0% 5.5% 3.0% MCAP (FY27E) narios assume AAUM CAGRs of 30% and 13% respectively, resulting INR Bn 5,675 13,770 1,933 in valuations of US$188bn (@5.5% of AUM) and US$26bn (@3.0% USD Bn 77 188 26 of AUM). Source: e = Morgan Stanley Research estimates. AAUM = Average Assets Under Management.

76 M BLUEPAPER How Big Can Fintech Companies Get in Exhibit 143: Fintech: Valuation Scenarios Terms of Market Cap? Payments / Fintechs Base Bull Bear Digital Transactions 1,208 % Share in Digital Transactions 20% 25% 15% We are valuing other financial companies – comprising exchanges, MCAP / Transactions 0.15X 0.20X 0.10X rating agencies and fintechs – at US$60bn by F2027. The key piece MCAP (FY27E) within this is fintech, and we discuss our valuation methodology INR Bn 2,657 4,428 1,328 USD Bn 36 60 18 below. Source: Morgan Stanley Research (E) estimates

We look at the potential size of digital payment transactions and assume that they will have a 20% share of the system by F2027, up If we look at the spectrum of fintech companies in India, they straddle from 8% now. That would imply annual transactions worth around multiple parts of the financial services industry. There are multiple US$240bn through these platforms. players in payments, lending, wealth management, loan origination, and insurance. How could they be valued given the nascent nature of Some of the other global payment platforms trade at 13-15% of total these businesses? payment volumes. Applying that to the potential payment throughput in India would imply a potential market cap of Most of these fintechs are operating in areas where banks and non- US$20-60bn for these payment platforms. It is likely that the bulk of bank lenders are already operating, and in that sense they will be a this will be dominated by very few players, with more consolidation part of the market sizing analysis we have conducted above. Hence, in this space. The key here will be monetization – will these compa- we are not trying to ascribe a potential value to them separately. nies start charging for payments or will they keep them free to attract customers and use the data to offer other products like The one area which is not being captured in the earlier areas is market wealth management, lending services, etcetera. share gained by independent digital payment platforms like Paytm. Hence, we have ascribed a potential value to these businesses.

MORGAN STANLEY RESEARCH 77 M BLUEPAPER Risks to our Story

We have high confidence in India's digitization theme Threats to cybersecurity as more data goes online - Repeated evolving over the next ten years with significant benefits cyberattacks and newsflow of data theft could rattle user trust and for investors. Nonetheless, this investment theme does slow growth in digital transactions. While the architects of Aadhaar carry risks. These include: evince great confidence in the security arrangements around Aad- haar, the security of associated platforms is in diverse hands and Political outcome of 2019 elections - A change in government could pose a problem to the system at large if compromised. could pose the risk of change in agenda and unwinding of current policies, developments that could hamper growth and buck our assump- India’s macro climate - Although India is positioned well for strong tions about rising equity savings. The current government is a major economic growth (given demographics and policy initiatives in recent proponent of India's digital economy and is giving regulatory and years), such growth is contingent upon successful execution along administrative impetus to the Aadhaar platform. There is no guar- several fronts, including infrastructure, job generation, healthcare, antee that such an impetus may continue if there is a change in gov- and education, for example. Any impediments in this respect could ernment. adversely affect our outlook. India will add significant people to its work force in the coming decade and generating jobs to these people Risk of Aadhaar being subsumed by the ongoing debate around is crucial to the economic, social and political environment. privacy - A key driver of the success of digitization is the seamless use of Aadhaar in both financial and nonfinancial transactions. Hence, Deep global recession or crisis - A deep global recession or crisis any legal hurdles that might severely limit usage of Aadhaar could could negatively impact capital flows, currency, investments and hinder the digitization process. India's Supreme Court recently thereby hurt the progress of many of the above-mentioned initia- judged that privacy is a fundamental right for Indian citizens. tives. Whether Aadhaar infringes privacy or not is an ongoing debate. E-commerce growth risks: Funding is key to the growth of this Smooth transition to GST - The general expectation is that, over a sector, which in turn depends on technology stock valuations in the period of time, tax compliance will rise significantly, thereby US. So adverse developments in either the global VC/PE space and/or resulting in a stronger fiscal position. If GST implementation were a sharp correction in technology stocks could affect India's e-com- not to be as expected, then the resultant expected benefits of merce growth story. Also as the sector gains size, government inter- improved credit access for MSMEs and for economic growth might vention may increase in the form of regulations or taxes. not unfold as we envision. We should also not lose sight of the near- term risks that GST brings to the fore, which is that MSME profitability suffers, leading to job losses and an overall slowdown in economic activity.

78 M BLUEPAPER Individual Stock Section

Theme 1: Indian Financials Theme 3: Global Stocks With Indian Optionality

l Bajaj Finance l Alibaba l Edelweiss Financial Services Ltd. l Amazon l HDFC Bank l DBS l ICICI Prudential Life Insurance l Naspers l Kotak Mahindra Bank l Softbank l LIC Housing Finance l TransUnion l Mahindra & Mahindra Financial Services l Visa / Mastercard

Theme 2: Indian Consumption

l Asian Paints l Eicher Motors l ITC l MakeMyTrip l Maruti Suzuki l Ultratech Cement

MORGAN STANLEY RESEARCH 79 M BLUEPAPER Bajaj Finance

This is a pure play on consumer and SME financing 3. BAF's strong technology, digital orientation, and risk manage- growth given it has the highest exposure to these segment. In our view, BAF has demonstrated great operational rigour in ments as a share of the loan book. It should benefit delivering a high volume of small-ticket loans, as well as credit risk more from higher consumer / SME data given high management (working with a large inflow of new-to-credit cus- existing use of analytics. tomers). We believe that, given BAF's early adoption, relative to peers, and hence greater familiarity with, for example, a scorecard- Bajaj is the most geared, among our financials coverage, to the based approach to credit underwriting, and credit bureau data-based strong growth potential driven by digitization in consumer and analytics, it should benefit significantly from the coming data democ- SME lending. ratization being fostered by policies such as GST and RERA.

Consumer and SME loans, which we forecast will be the major India's multiyear discretionary consumption growth story, driver of loan growth over the next ten years, constitute >90% driven by millennials, is one of the core drivers of our expecta- of the loan book of Bajaj Finance (BAF). BAF offers a diverse range tion for BAF to excel amid digitization. We believe that the com- of products across both these segments. Continued focus on these pany can sustain high growth (around 30%) and ROE (around 20%) segments has helped it deliver a strong 33% EPS CAGR, with an for the next three to five years. India's retail credit penetration is average ROE of 21%, FY2014-17. much lower than that of other Asian emerging countries and the developed world. We believe this can increase significantly over the We expect the following factors to drive sustained strong perform- next ten years, given the growing Millennial population, alone (i.e., ance at BAF, FY2017-27: the sheer proportion of the population entering the work force). In our view, this augurs very well for consumer financiers such as BAF 1. BAF's expertise and strong positioning in discretionary con- that have made significant investments in this business, possesses a sumption finance, which we expect to accelerate meaningfully pan-India network, experience, and which have spent significant over the next ten years. BAF has been at the forefront of tapping effort in developing risk management systems and service delivery the financing opportunity in discretionary consumption. It finances channels. >20% of all consumer durables (white goods) sales in India. It has also been consistently and aggressively increasing its footprint Where we could be wrong: An underlying assumption is that BAF across channels (both physical and digital). In addition, it has been will be able to maintain asset quality and profitability. If this were not adding products. This has driven strong growth in customer acquisi- to happen and if BAF were to see meaningful compression in profita- tion as well as repeat business with existing customers. bility with or without an increase in NPLs, valuation multiples could come under significant pressure from current premium levels. 2. Ability to deliver product innovation and retain customers through cross sell and customer satisfaction. Rising share and incomes of millennials should result in demand for newer products and services. In addition to consumer durables, BAF has been innovative in identifying opportunities and adding verticals, such as digital financing (mobile phones and so on); lifestyle financing (furniture and so on); and life-care financing (dental care, hair restoration, laser eye treatment, stem cell preservation and so on). Product innovations include pre-approving customers, loyalty cards (called Existing Member Identification i.e. EMI cards), and credit lines for both SMEs and consumers. BAF has also expanded its product suite by adding payments via its tie-up with MobiKwik.

80 M BLUEPAPER

Exhibit 144: Exhibit 145: BAF Screens Higher on ROE Profile Than Retail Private Banks... … and its Share of Consumer Loans + SME in Loan Book (>90%) Is Much 30.0% ROE (%) FY17 FY18e FY19e FY20e Higher Than That of Retail Private Banks, Which Helps Explain its Pre- 25.0% mium Relative Valuations 35 BAF 20.0% 33 31 15.0% 29 Kotak

10.0% 25.0% 27 21.7% 21.7% 21.6% 21.6% 20.9% 20.9% 20.2% 19.4% 19.4% 18.4% 18.4% 17.9% 1Y PEFwd 1Y 16.2% 16.2% 25 HDBK 14.7% 14.1% 14.1% 5.0% 13.8% 23

0.0% 21 BAF HDBK Kotak 19 Source: Company data, Morgan Stanley Research (e) estimates 17 Size of bubble indicates Consumer+SME book (F1Q18) 15 40% 60% 80% 100% Share of retail loans (F1Q18)

Source: Company data, Morgan Stanley Research. Morgan Stanley Research (e) estimates

MORGAN STANLEY RESEARCH 81 M BLUEPAPER Edelweiss Financial Services Ltd.

This is a diversified financial with exposure to many of Exhibit 146: the fastest-growing and highly profitable financial ser- We Believe that Exposure to Diverse and Fast-growing Profit vices businesses - affordable housing and SME lending, Pools (With Less Competition, and Hence High ROE) Position wealth and asset management, niche spaces like dis- Edelweiss for Significant Profit Growth Over the Next Decade tressed credit. BMU, Corporate Structured and Others Collateralised As a diversified financial services competitor, with a good combina- Franchise 10% Credit Business Capital 15% tion of credit and capital-light businesses, Edelweiss is positioned Markets 15% well to benefit from the strong growth we expect in financial ser- Wholesale Mortgage vices, FY2017-27. In FY17, its credit and other businesses contributed Wealth & 22% Asset 75% of group profits, and non-credit businesses (capital market Management 10% linked) contributed 25%. Distressed LAS, SME &Retail Credit Agri Mortgage 12% 9% 6% Its distinctive advantage is that its key businesses are in segments Credit Business that are among the fastest-growing parts of the financial services PAT Mix ex-insurance - FY17 (Pre MI) industry, but which face relatively less competition from banks, Source: Company data, Morgan Stanley Research implying good pricing power (i.e., profitability potential): 4. Structured credit solutions: Like distressed credit, structured collateralized credit is also a niche business with few players of scale. 1. Affordable housing finance: Edelweiss's focus is on small- to Edelweiss's key strength in this area is its investment banking rela- medium-ticket self-employed housing loans, which is not a signifi- tionships, which enable it to better assess and take risk, as compared cant focus of banks. Hence, unlike traditional "prime" mortgages, this with peers who do not focus on this arena. As in its distressed credit product has pricing power and hence greater profitability potential. business, Edelweiss has indicated that it can use its various relation- Over the next ten years, government initiatives for affordable ships to scale this business via the off-balance-sheet route, thereby housing should drive a strong profit CAGR for Edelweiss, in our view. limiting balance sheet risk and deriving management fee income, rather than growing these assets on the balance sheet. 2. MSME credit: Edelweiss provides small-ticket loans to MSMEs, a significant proportion of which are disbursed on the basis of surro- 5. Capital markets: As discussed earlier in the report, we forecast gate means of underwriting (e.g., observing footfalls, talking to the very strong equity flows and trading turnover over the next ten borrowers' vendors and suppliers), owing to a lack of documentary years. Bond markets should benefit from positive regulatory evidence of income. GST implementation should result in higher con- changes. We expect Edelweiss's traditional broking and investment sumer credit data availability, and hence significantly higher vol- banking businesses to derive significant benefit from the growth of umes. While yields will likely move lower, credit costs should also this segment. move lower, and we expect leverage to improve over time. 6. Wealth management: Strong economic growth over the next 3. Distressed credit: Edelweiss ARC is India's largest asset recon- decade should result in significant wealth creation and opportunities struction company by managed assets, and it has ~50% market share for wealth managers. Domestic players such as Edelweiss are in managed assets as of FY17. Capital deployment / ability to mobilize ramping up significantly, especially in the wake of exits by some for- capital via partners and strategic investors will, we believe, be a key eign competitors. Synergies are high with other businesses (capital differentiating factor among ARCs as the industry's focus has moved markets, asset management, insurance, distressed assets and struc- towards resolution and potential complete takeover of stressed tured credit). assets from bank balance sheets. In our view, Edelweiss is positioned well, supported by partners such as CDPQ (a Canadian Pension Fund) Where we could be wrong: Our thesis won't play out if Edelweiss and by its high net worth individual relationships in its wealth man- sees a sharp rise in NPLs with or without a major prolonged decline agement, alternatives and capital markets businesses. in equity markets and sentiment.

82 M BLUEPAPER HDFC Bank

Play on acceleration in retail and MSME loan growth Exhibit 147: HDFC Bank: Balance Sheet Per Employee Is Rising Meaningfully HDFC Bank is clearly one of the biggest beneficiaries of the con- Employees (Number) Assets / Employee (Rs Mn), RS 100,000 125 sumer and MSME lending boom. HDFC Bank has a number of advantages over the rest of the banking system: 80,000 100

1. Cleaner balance sheet relative to many other lenders. This has allowed management to focus on growth areas and benefit from the 60,000 75 market share shift away from corporate lenders. Even in the corporate lending space, the bank is seeing solid growth as it develops 40,000 50 stronger relationships with large corporates. 20,000 25

2. HDFC's low funding cost allows it to price aggressively and still make strong returns. The bank has sought to maintain a strong 0 0

liability base, with a high proportion of funding coming from current F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 accounts and savings accounts. This has kept its funding cost Source: Company data, Morgan Stanley Research amongst the lowest in India, allowing it to achieve reasonably high spreads, even on the high-quality assets. Exhibit 148: On top of this, the bank has been aggressively rolling out its digital HDFC Bank: We Expect Market Share in Bank Loans to More Than strategy, helping it to reduce its cost of operations fairly significantly. Double Over the Next Decade For instance, since September 2016, the bank has grown its balance 18.0% sheet by >20%, while its employee base has contracted from 16.4% ~95,000 to ~85,000 during this period. 15.0% CAGR of 12.0% 20% As we mentioned in the market cap section, we would expect some pressure on revenue yields for banks, as technology is likely to stoke 9.0%

competition. But the compensating factor would be lower operating 7.3% 6.2% costs as a percentage of assets. And HDFC Bank is seeing early signs 6.0% 5.2% 4.7% 4.4% 4.1% 3.9% of such improvement. We expect its cost efficiencies to improve fur- 3.8% 3.6% 2.8% 2.5% 2.4%

3.0% 2.3%

ther, which should help keep ROA fairly stable. 2.2% 1.7% 1.1% 0.9% 0.8%

0.0% F2000 F2001 F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F2027E

Source: Company data, Morgan Stanley Research

MORGAN STANLEY RESEARCH 83 M BLUEPAPER The bank is also using data very effectively in growing its loan book. Exhibit 149: For instance, almost 30% of its unsecured loan book is being offered HDFC Bank: P/E Multiple Is High Versus History HDBK HDBK Core through its "loan in 10 seconds" product. The bank has set up a back 35.0 end that can rapidly process loan applications, access its customer data, retrieve and assess credit bureau data, and then quickly make 30.0 the loan. This is one of the factors that is helping it grow its loan vol- 25.0 umes at a fast pace. Mean + 1SD Mean 20.0 Mean - 1SD We believe that as MSME lending picks up, supported by data mining, 15.0 HDFC Bank will be amongst the major beneficiaries. It should be able to process this data more quickly than many other banks and make 10.0 faster lending decisions, given its digital infrastructure. This, along Mar-08 Mar-09 Mar-10 Mar-11 Mar-12 Mar-13 Mar-14 Mar-15 Mar-16 Mar-17 Sep-07 Sep-08 Sep-09 Sep-10 Sep-11 Sep-12 Sep-13 Sep-14 Sep-15 Sep-16 Sep-17 with funding cost advantage, should help significantly boost its HDFC Bank core adjusts for its stake in HDB Financial. Source: Company data, Morgan Stanley Research growth. We expect HDFC Bank's loan growth to average about 20% over the next decade (as compared with ~27% over the past 10 years).

Strong growth coupled with stable ROA is likely to drive strong earnings. Moreover, HDFC Bank's nonbanking subsidiary - HDB Financial - is also performing well. It should also derive a substantial benefit from a pickup in consumer and MSME loans - the key focus areas for HDB Financial.

HDFC Bank's valuations as compared with its history are elevated and will likely contract over the next ten years. But, given the likely strength of compounding, we expect the stock to perform well.

Where we could be wrong: a) greater than expected competition in retail on pricing; b) slower than expected CASA growth; c) higher than expected competition in processing and other fees.

84 M BLUEPAPER ICICI Prudential Life Insurance

Play on strong growth in equity savings and under Exhibit 150: insurance We Expect ICICI Pru Life's APE Market Share to Improve ~6ppts Over the Next Decade India's life insurance segment, in our view, is likely to be a key ICICI Pru Life: APE Market Share (%) 20.0% beneficiary of: a) accelerating GDP growth, b) formalization of sav- 18.0% ings and increasing share of financial savings within that, and c) accel- 16.0% 12.7% erating protection growth. Large bank-backed insurers are well 12.0% 11.3% 11.3% 11.3% 11.3%

12.0% 10.9% 10.1% 9.9% 9.3% placed to capitalize on such developments. They have better distri- 8.8% 7.3% 7.2% 8.0% 7.0% bution, strong operating metrics (persistency, costs), and are focused 5.9% on various technology initiatives. ICICI Prudential Life is one such 4.0% play. 0.0% F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

We forecast a 17% CAGR for premium growth and a 6ppt APE F2027E market share rise, to 18% in ten years: ICICI Prudential is a Source: IRDA, Morgan Stanley Research estimates leading unit-linked insurance product (ULIP) competitor, supported by a portfolio of relatively competitive products. While Exhibit 151: the regulator has significantly capped expenses under ULIPs, we note Cost and Persistency Ratios - Strong Improvement Over Past five Years that ULIP products at ICICI Prudential have expense charges that are Adjusted Opex Ratio %) 13M Persistency Ratio (%), RS even lower than the requirement, and this has helped its ULIPs stand 16% 90% out relative to the competition, and even versus mutual funds over 80% a longer-term investment horizon. In our view, this will help ICICI capi- 14% talize on the coming growth in equity savings that we expect. 70%

12% We forecast a 25% VNB CAGR, 2017-27, much higher than growth 60% in premiums, given significant improvement in profitability. We 10% 50% highlight two key reasons: F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

Source: Company data, Morgan Stanley Research. a) increasing protection mix: Selling protection in India has been tough historically. Lack of awareness and also limited focus by insurance (10 years back) are the key reasons, and this has meant that life Exhibit 152: protection penetration in India is generally quite poor versus levels IPru Life: P/EV Versus ROEV Forecast 4.5 18.0% in more developed markets. However, awareness levels are now 1Y Forward P/EV (Headline) ROEV (%), RS changing (as seen in a pickup in protection premiums). Moreover, 4.0 17.5% insurers are working on a number of technology initiatives to under- 3.5 17.0% write and cross-sell better. This should help drive protection mix 3.0 16.5% higher. 2.5 16.0%

2.0 15.5%

1.5 15.0% Mar-17 Mar-18 Mar-19 Mar-20 Mar-21 Mar-22 Sep-18 Sep-19 Sep-20 Sep-16 Sep-17 Sep-21

Source: Thomson Reuters, company data, Morgan Stanley Research estimates

MORGAN STANLEY RESEARCH 85 M BLUEPAPER b) better operating leverage. One of the key enablers here would Although ICICI Pru valuations are not cheap, we nonetheless be technology. To start with, ICICI prudential has used technology believe that the combination of the above two factors can drive well to reduce cost and improve persistency ratios. Over the next ten sustainable RoEV of ~17-18% over next ten years. This implies years, we believe technology usage to cross-sell will play a greater strong returns, even as valuations move lower. role in insurance sales, particularly for ICICI Prudential, as it is very focused on protection and low-cost ULIPs. Traditional products, on Where we could be wrong: a) significant margin compression in the the other hand, are relatively complex, and hence technology will protection business; b) an aggressive move to open architecture by play a relatively lower role in such arenas, in our view. We believe that ICICI Bank; and c) potential increase in tax rates. ICICI Pru is prudently focused on smart initiatives, such as a) cross- selling/upselling customers at various touch points using analytics, b) developing innovative digital marketing tools, and c) integrating digital platforms with distributors, among other efforts.

86 M BLUEPAPER Kotak Mahindra Bank

Big beneficiary of acceleration in consumer/MSME loan 1. Sound industry play, given presence across financials value growth and capital market activities chain: Kotak has presence across the entire value chain, with 100% ownership across most business segments. As discussed earlier In our view, among our coverage universe, Kotak Bank will be one herein, we expect capital market businesses to do well, given higher of the biggest gainers in the financial services boom that we formalization of savings and increasing financial savings within that. expect. Kotak kept itself away from lending to riskier segments and has one of the industry's strongest balance sheets (with strong capital and low impaired loans) in the current cycle. More- over, the merger integration process with erstwhile ING Vysya bank is now complete, and the bank can now focus on growth. This, coupled with a strong management team, should help improve loan/revenue growth sharply, as well as boost RoE.

Exhibit 153: Kotak: Present Across the Entire Value Chain

Source: Company presentation

MORGAN STANLEY RESEARCH 87 M BLUEPAPER Banking Businesses: We expect strong 28% growth in earnings over Exhibit 154: the next 10 years, driven by: Kotak: We Expect Market Share in Bank Loans to More Than Triple in the Next Three Years Strong ~25% loan growth CAGR, 2017-27, leading to market 7.5% share of 5.7% (versus 1.8% currently). Kotak's loan book has one 6.0% 5.7% of the highest contributions from the SME and consumer segments. 25% CAGR As discussed earlier in this report, we expect system growth in con- 4.5% sumer and MSME loans to be around 17% each, respectively, over the 3.0% next ten years. For Kotak, we believe growth should be higher, given 1.8% 1.7% 1.5% 1.0%

its continued strong focus and relative expertise in this segment (also 0.9% 0.9% 0.9% 0.7% 0.7% 0.6% 0.6% 0.6% 0.4% 0.4% 0.3% 0.2% helped by its ING Vysya acquisition). 0.2% 0.0% F2002 F2003 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017

We expect bank RoA to improve to 2.0% from 1.7% currently, F2027E driven by higher operating leverage. In our view, the CAGR of costs Source: Company data, Morgan Stanley Research estimates at Kotak Bank over the next 10 years is unlikely to be higher than 16-18%, much lower than the revenue CAGR of ~23% (we expect fee Exhibit 155: income growth to be lower than NII growth). Slower cost growth Market Share in Various Capital Market Segments should be driven by a declining pace/size of branch expansion, as cus- F2017 F2027E 7.5% tomer acquisition and maintenance turns digital. For instance, Kotak

23% CAGR has launched 811, a paperless and digital way to open bank accounts 6.0% within five minutes. More importantly, the acquisition cost via this 4.5% approach is ~90% less than the previous manual method. 20% CAGR

3.0% 6.2% 5.0% Capital market earnings will be strong as well; we expect a 4.5% 1.5% ~20-23% earnings CAGR over the next 10 years. As discussed ear- 2.3% lier, our macro team expects equity savings of US$420bn–US$525bn 0.0% over the next ten years in India, versus the respective US$60bn and Life Insurance AMC Source: Company data, IRDA, SEBI, Morgan Stanley Research estimates US$120bn that households and foreign portfolios invested over the previous ten years. This, coupled with Kotak's relatively low starting point in these segments and strong management, will help it grow Exhibit 156: faster. We expect bottom-line growth to be higher than top-line Kotak (Consol): 1Y Forward P/B growth given the high operating leverage nature of these businesses. 5.0

4.0 Mean+1SD Although Kotak's valuations are not cheap, we believe that strong earnings compounding implies strong returns potential, even as valu- Mean 3.0 ations move lower. Mean-1SD 2.0 Where we could be wrong: The key risk to our thesis on Kotak Bank is higher than expected competition on pricing driving down margins 1.0 and/or weaker than expected loan growth. Mar-04 Mar-05 Mar-06 Mar-07 Mar-08 Mar-09 Mar-10 Mar-11 Mar-12 Mar-13 Mar-14 Mar-15 Mar-16 Mar-17

Source: Thomson Reuters, company data, Morgan Stanley Research estimates

88 M BLUEPAPER LIC Housing Finance

Big beneficiary of strong growth in affordable housing Exhibit157: as it is the lowest-cost producer of housing loans; RERA Low Starting Point of Non Housing Loans, i.e., Flexibility to Drive Loan and GST should drive bigger foray into self-employed Growth and Margins and developer segments Developer LAP, LRD & Loans, 3.8% LICHF is India's second-largest housing finance company (HFC), Other Non with a ~9% market share, with among the strongest core mort- Core, 12.9% gage franchises. Given strong parentage and an AAA rating, LICHF's cost of funds is one of the lowest among HFCs. Further, its cost of operations is lowest among the HFCs. This thus makes LICHF the lowest-cost producer of individual housing loans. Consequently, on prime individual home loans, LICHF's ROE is among the highest.

LICHF has a strong pan-India presence and has traditionally had lower risk preference with among the largest exposures in the Individual salaried housing loan space. LICHF has a geographically diversified Home Loans, loan book covering about 450 key centres across the country. It has 83.3% among the lowest exposures to non-home-loan products, like loans against property (12.9% of the F1Q18 loan book) and developer loans (3.8%). The salaried segment (having predictable cash flows and greater overall tax compliance) forms 84% of the individual segment

(LAP + home loans). Even within the individual segment, the share of Source: Company data, Morgan Stanley Research. Loan mix as of June 2017. loans to government employees is quite high.

Growth in recent years has been sluggish. However, we see a Exhibit 158: strong case for higher loan growth and taking on more risk, i.e., LICHF Has the Highest Share of Lower-ticket Home Loans Among potentially a better profitability mix, driven by the confluence of Large HFCs, i.e., It Is Positioned Well To Be a Big Beneficiary of a Pickup GST and RERA: in Affordable Housing Individual Housing loan more than Rs. 1.5 mn Individual Housing loan upto Rs. 1.5 mn Historically, LICHF's share in the self-employed segment has been low 100.0% relative to some of the other fast-growing large HFCs. But this could 75.0% change as underwriting self-employed borrowers becomes easier with 67.9% data democratization as fostered by GST implementation. This could 87.3% 91.5% 92.9% 50.0% open new segments / markets for LICHF, driving higher loan growth.

Further, loans to self-employed customers also earn a slightly higher 25.0% loan spread, although this should progressively move lower with 32.1% 12.7% 8.5% 7.1% better data availability. 0.0% LICHF HDFC PNBHF IHFL Source: Company data, Morgan Stanley Research. Data for FY17.

MORGAN STANLEY RESEARCH 89 M BLUEPAPER Among the large HFCs, LICHF should be one of the biggest beneficia- Exhibit 159: ries of strong growth in the affordable housing segment. With ~32% Starting Point of Valuation Is Attractive Relative to Most HFCs 4.5 of individual loans having a ticket size of

Driven by the above reasons, we expect a strong housing loan CAGR (around 16%) over the next ten years, with potentially even a stronger non-housing loan CAGR, which could help improve / sustain NIM in the wake of NIM competition in the core housing loan business.

Where we could be wrong: LICHF fails to accelerate growth in home loans and loses market share in housing loans. It also does not increase the share of non-housing loans thereby seeing a decline in net interest margins.

90 M BLUEPAPER Mahindra & Mahindra Financial Services

This is a pure play on growth in rural financing opportu- Exhibit 160: nities given its rural focus and large network. Digitiza- Given its Rural Focus, Expertise, and Strong and Growing Network, tion should enable it to serve its customers better and MMFS Is a Pure Play on Rural Lending Opportunities and Should Grow reduce its cost of operations. Assets at 17-20% CAGR Over the Next Decade, We Expect

640,000 ~600,000 MMFS Presence in Villages 560,000 As penetration and acceptance of digital modes of payments 480,000 rises in rural India, it should boost financial inclusion and lead to 400,000 better and more economical servicing of the 'mass' rural cus- 319,409 320,000 286,007 241,366 tomer. High preference for, and prevalence of, cash in transactions; 240,000 182,264 low proximity to customers, i.e., population spread over large areas; 160,000 133,907 and consequently high costs of customer engagement, have histori- 80,000 0 cally been the key impediments in servicing rural customers in both FY2013 FY2014 FY2015 FY2016 FY2017 Potential Opportunity credit and liabilities. Source: Company data, Morgan Stanley Research.

MMFS is the pure play in this space. While most banks do have a rural presence, MMFS' businesses (vehicle finance, housing finance Exhibit 161: and insurance broking) are focused solely on rural and semi-urban Given High Operating Costs in the Business, Sensitivity of Profitability customers. Further, MMFS' key customer segment is 'new-to-credit' to Reduction in Operating Costs Is High borrowers, which is essentially high-risk / high-yielding collateralized Avg F08-17 Cost to Assets 3.4% loans to 'mass' rural customers, not the preferred segment of banks. Avg F08-17 ROA 2.8% Of the ~600,000 villages in India, MMFS has a presence in % Reduction in underlying Impact on ~320,000. cost of operations ROA (bp) Profits (%) ROE (%pt) 5% 11 4% 0.9% Over the longer term, some of the potential benefits for MMFS from 10% 22 8% 1.8% increasing digitization that are not priced in today, in our view are: 15% 34 12% 2.7% 20% 45 16% 3.6% 25% 56 20% 4.5% a) Digital payments, which should help drive a lower cost of operations and better cash management for MMFS. >50% of Source: Company data, Morgan Stanley Research (e) estimates MMFS' loan installment collections are in cash, and many times these are multiple small value collections requiring repeat visits to the customers, creating a high cost of operations, cash handling charges and treasury costs. As the share of digital payments increase in the rural ecosystem, the possibilities for MMFS to reduce costs are substantial – i.e., collect multiple installments digitally, use the "ask money" feature in UPI to both remind customers and pull funds, manage treasury efficiently, and so on. Digital records of earnings of customers over time, could also help underwrite and monitor them better.

MORGAN STANLEY RESEARCH 91 M BLUEPAPER b) Faster, better and customised credit delivery to existing cus- Rural housing finance – A large opportunity. Rural housing finance tomers, enabling better customer retention. To existing, pre-ap- in India is <10% of overall housing finance, and MMFS' subsidiary is proved and digitally savvy customers, MMFS could consider one of the few players in this space. PMAY (Rural) plans to construct disbursing small-ticket instant loans digitally. 30mn homes in rural areas by 2022. Any meaningful progress in this direction could result in a significant growth opportunity (the subsid- c) GST should help open opportunities in the rural MSME iary is already growing the loan book by 50% annually). financing space. MMFS has been foraying into vendor financing and MSME financing (including the supply chain of M&M). Better transac- Potential to launch more financial products in rural India. MMFS tion data, owing to GST , should help assess borrowers better and has been assessing various opportunities - in addition to credit and deliver higher and faster credit. insurance broking, it has also set up an asset management subsidiary. d) Cheaper deposit funding. MMFS is one of the few NBFCs with a Where we could be wrong: A deep cyclical rural slowdown with an deposit taking license and probably the only NBFC offering a com- associated rise in NPLs is the key risk to MMFS; this could also be a pletely online deposit product. Higher online adoption among users spanner in the works for digital initiatives and so on. could help MMFS garner cheaper deposits (no brokerage and lower admin costs).

92 M BLUEPAPER Asian Paints

Asian Paints is a key beneficiary of increasing dispos- We believe that strong new construction growth trends combined able incomes and strong growth in property demand. It with a shrinking repainting cycle augur well for the decorative paints has a standout business model and is set to benefit industry in India. With 55% market share (in value terms) in the orga- from an improvement in urban consumption growth. nized paints industry in India, we believe Asian Paints is well positioned to benefit from this growth. Repainting demand contributes Euromonitor estimates the Indian decorative paint industry market 80% of APNT's decorative paint segment revenues currently, in our size at Rs262bn in 2015, with 10% revenue CAGR over 2007-16. The view, placing the company on a multi-year trajectory of strong strong growth was driven by a combination of factors: new construc- volume offtake and, consequently, high visibility on earnings growth. tion housing demand, rising urbanization, an increasing number of single-family homes, and a strong repainting demand cycle. Indeed, APNT also has advantages on several fronts, in our view, such as con- over the past two decades, paint industry volume growth has also sumer mindshare, dealer penetration, and supply chain/logistics shown a strong correlation with real GDP growth (1.5-2.0 times GDP strength. An acceleration in urban discretionary consumption growth). Our channel checks with dealers and industry experts sug- growth given visible green-shoots of an urban consumption recovery gest that consumers in India increasingly perceive paint as expendi- would benefit the home improvement category and APNT in partic- ture for home maintenance rather than décor. ular. APNT is one of our top picks in our consumer discretionary coverage and has performed well, with double-digit domestic volume Our analysis of consumption trends for decorative paints across geo- growth over the past three years. This compares favorably with many graphies indicates that the per capita consumption in India is US$3 other discretionary consumer companies that reported sluggish per annum, low compared with the US at US$26 but broadly similar domestic volume growth over the past 2-3 years and found this to China's US$3.4. However, in terms of share of wallet spend, we period challenging from a demand perspective. note that India's consumption trend is higher than that of peers. We attribute this to the weather conditions in India and the reduced Where we could be wrong: Given the strong correlation of paint length of the repainting cycle in India. As per our channel checks, the industry volume growth to real GDP growth, any slowdown in eco- cycle has reduced to 4-5 years currently from 7-8 years in 2006. nomic activity can impact industry growth trends. In the event that competitive intensity increases in the decorative paint segment, we Our India property team expects property industry demand to rise at see risks to APNT’s market share and subsequently our base case a 14% CAGR, 2015-20, and an 18% CAGR, 2021-25. volume growth estimates. Operating margins for the business may be at risk in the event of a sharp increase in input costs.

MORGAN STANLEY RESEARCH 93 M BLUEPAPER Eicher Motors

Eicher is a key beneficiary of higher discretionary c) Exports to be a growth driver: Aiming to capitalize on RE's spending. With rising disposable incomes in India, we British heritage, designed in the UK, made in India: Eicher expects expect a strong upgrade cycle for two wheelers in India exports of its RE brand to be a key focus. The company has set up a and we expect Royal Enfield’s (RE) cruiser bikes seg- team in the UK to make RE a "disruptive force in global motorcycles." ment to be a key beneficiary of this trend. It plans to develop and design models in the UK, keeping a global audience in mind, and offer customers a heritage brand at an afford- We believe that strong demand is also likely to lead to greater able price point. It also seeks to expand the 250-750cc segment glo- investor confidence in Eicher's ability to sustain earnings growth and bally by focusing mainly on 500-750cc motorcycles. We forecast thus support long-term cash flows. Eicher will export around 1,900 units per month of the RE brand in F19, versus Bajaj Auto, which exports a combined 18,000 Pulsars and a) RE is the only aspirational yet affordable brand in India's autos Avengers per month. A faster-than-expected pickup could sharply segment, and we expect it to be a key beneficiary of premiumisa- re-rate the company, in our view, as visibility on export trends and tion: Eicher, with its Royal Enfield brand, has a distinctive position in activity is currently low. India. The brand's aspirational quality helps Eicher stand out among peers, command a loyal customer base, garner strong pricing power, Where we could be wrong: and keep sales promotion spending low. The brand's affordable quality ensures RE is best placed to capitalize from upgrades during "Pull brand" turns into a "push brand": RE is the most profitable com- replacement purchasing. pany now in two-wheelers, and it has attracted competitors, but no competitor has yet succeeded as has RE. Stronger-than-expected b) VECV should see margins improve as market share rises: VE competition could be a key downside risk. Commercial Vehicles Ltd. (VECV) is a joint venture between Volvo and Eicher. We believe growing volumes and the consequent oper- Stronger-than-expected competition: Any strong new entrants in the ating leverage will lead to margin expansion for VECV. cruiser bike segment in India at price points similar to that of Eicher can take away some market share from Royal Enfield.

Any sustained slowdown in commercial vehicles: Also key to watch out for any signs of sustained sluggish CV demand in India.

94 M BLUEPAPER ITC

GST offers the opportunity of structural market share In addition, in a rational tax environment, we compute that ITC’s ciga- gains for ITC from illicit cigarettes; valuations and earn- rette profit pool could increase nearly 5x over the next 10 years. Very ings expectations are relatively benign due to risks from few established categories within Indian consumption offer this sort continuing adverse policy action of growth potential with strong earnings visibility. Also, a shift from the Indian duty paid bidi industry (~365bn sticks p.a. as per our We have long believed that the only way to counter the illicit ciga- channel checks), is declining ~2% p.a. As observed from our Alpha- rette segment in India is a relatively benign tax environment (with Wise survey, the daily consumption frequency of bidi smokers is price hikes only marginally ahead of consumer inflation). However, twice that of core cigarette consumers. We estimate that even if half the implementation of the GST provides the government an oppor- of the bidi volumes translate to cigarette volumes, the industry will tunity to improve tax compliance based on analysis of data collected gain an incremental 3-ppt growth p.a. from this transition alone. It is from the GST network. also interesting to note that the total bidi industry (including unorganized) is much larger, producing 750bn-1tn sticks p.a. Hence, head- We estimate >70% of the illicit cigarettes sold in India are domesti- room for growth could be higher. cally manufactured but duty is evaded. In the absence of a digital trail of transactions, hitherto, it was difficult for government agencies to Historically, a valuation re-rating/de-rating for ITC is linked to visi- establish or monitor tax evasion in the manufacturing value chain. bility of cigarette volume growth. The latest cigarette policy action However, the introduction of the GST may present an opportunity which implies a 11-21% increase in tax for cigarettes versus that in for the government to control tax evasion. To be clear, this opportu- F2017, in our view will impact near-term volume growth and valua- nity is captured only in our bull case scenario. tions multiples for the stock.

Our channel checks suggest that certain raw materials used in the Where we could be wrong: Any incremental adverse government manufacture of cigarettes like filters and specialized paper are policy action (tax, packaging, a loose cigarette ban) will put our ciga- sourced from industries that are relatively consolidated, organized rette volume growth estimates at risk. An increase in awareness and and likely tax compliant. Data collected from these industries may shift to reduced-risk products (RRP) in tobacco consumption as help ascertain the balance between taxes collected on input mate- smokers increasingly look for alternatives to combustible cigarettes, rials and manufactured goods for the industry. There may be an may impact long-term valuations on the stock. However, we point opportunity for the government to track any gap between cigarettes out that affordability (relative to other global markets) and indeed manufactured and duty paid cigarettes sold in India. restrictive government policies on investments in tobacco in India may prove to be insurmountable hurdles for global players over the next decade.

MORGAN STANLEY RESEARCH 95 M BLUEPAPER MakeMyTrip

Dominant OTA platform and direct beneficiary of rising OTAs to gain market share: The online penetration on our esti- internet penetration and acceleration in consumption mates is only 20%, and it can potentially go to 35% by 2021. Within expenditure over the next decade the total online market, the share of OTAs could increase further as they become the dominant online distribution channel for the hotels, MMYT is the dominant online travel agent (OTA) in India, with a as in the case of airlines. Within the domestic air segment specifically, market share of greater than 56% within OTAs in India. We believe MMYT and Ibibo together are well placed to garner a dominant por- MMYT should be a direct beneficiary of rising digitization and an tion of the total OTA market by 2021, in our view. We expect the OTA acceleration in consumption expenditure over the next decade. With market and MMYT to grow gross bookings at a CAGR of 30%+ over an increase in online penetration and improved market share, we the next few years. expect sales and marketing spend to rationalize and the company should achieve profitability by FY20 (2019). Strong bookings/revenue growth and rationalization of sales and marketing spend to drive profitability: With the business mix Online penetration within the domestic air ticketing segment has improving towards the higher-margin Hotels & Packages (H&P) seg- already crossed 50%, and we expect a similar trend to play out within ment, we expect net revenue growth to be stronger than that of domestic hotels in India. We expect overall online penetration to bookings over the next few years. With improving blended net rev- increase from 20% (in 2015) to 35% by 2021 (largely led by the enue margins, better economies of scale for the overheads expenses domestic hotels segment) and dominant OTAs like MMYT should and lower sales and marketing spend with rising dominance of gain further share, taking their total market share to 70%+ by 2021. MMYT platform, we expect the company to turn profitable in FY20, Secondly, the overall travel trends in India (the total number of with steady state margins improving to 20%+ over the longer term. domestic trips) is strongly correlated to nominal GDP growth in India. Further we expect consumption expenditure to increase at a Where we could be wrong: a) Heightened competitive intensity faster clip over the next decade (relative to the last decade), which from new players for longer could delay the turnaround to profits bodes well for discretionary spend items like travel over the coming and drag the company’s cash balance. b) Competition from suppliers’ years. direct channel could increase, such as via airline websites. c) Com- pression in net revenue margins for air and H&P segment would Addressable market opportunity is huge for the OTAs: We esti- impact net revenues and profitability adversely. d) The shift from off- mate the addressable market to increase from US$34bn in 2015 to line to online happens more slowly than expected and requires signif- US$67bn by 2021. This implies a CAGR of over 12% in the overall icantly larger investments than anticipated. e) Rupee depreciation market. versus the US dollar impacts the company’s reported net revenues. f) There could be a slowdown in the macroeconomic environment (leading to slower consumption expenditure on discretionary items like travel) or travel shocks due to unforeseen events.

96 M BLUEPAPER Maruti Suzuki

India is set to be one of the fastest growing car markets Economies of scale: Economies of scale are key to profitability. They in the world and Maruti Suzuki dominates with 80% allow a company to extract value at every touch point. profit pool share. Leading on new technologies: From automated manual transmis- The passenger vehicle (PV) industry has been through a long down- sion (AMT), to mild hybrids to compressed natural gas (CNG) offer- cycle and is showing signs of a demand upturn ahead as economic ings, the company has been launching technologies suited to growth and disposable income growth picks up. We also see consumers financing trends being favorable to MSIL - at 80%, on our F2017 estimates, the finance penetration was at a ten-year high and banks Business Moat 2 – Distribution, more than double that of the remain keen to push auto lending further. While Maruti dominates nearest competitor: Distribution is a key strength for MSIL as it has the Indian car market with 80%-plus share of the profit pool, compet- 2,000-plus sales points – more than double the number of the itive intensity in the industry is at a low with several OEMs stressed nearest competitor. Building this vast distribution is difficult for new due to low scale. Consequently we believe Maruti is best placed to players but it is critical for growth as car growth is mainly coming capitalize on an upcycle in the Indian PV industry. The stock trades at from outside the top ten cities. In addition, the company is upgrading 25x one year forward P/E on our estimates, which we find reasonable its customer reach via new formats like Nexa and now Arena. vis-à-vis our estimate of a 20% EPS CAGR for Maruti over F17-20. Where we could be wrong: Structural reasons to like Suzuki's Indian business: We assume that the Indian market posts a CAGR of 11% in F2017-20 versus 3.3% The Indian car recovery remains patchy: We assume that the Indian in F2012-17. We expect Maruti’s growth will be strong not only on an market posts an 11% F17-20 CAGR versus a 3.3% CAGR over F12-17. A absolute basis but also on a relative basis and hence expect market slower-than-expected pace of economic recovery could lead the share wins to be sustained given the strong moats Maruti has built down-cycle to last longer than anticipated. around its business: Competition setting up a Tesla-like Giga factory in India: Scale will be Business Moat 1 - Strong product proposition: MSIL offers 15 key for success in EVs. Sizable EV investment by an OEM could accel- products, and 7 out of the Top 10 selling models in India are from erate the pace of EV growth in India and Suzuki could remain left out MSIL (F2017). The product proposition is strengthened by: because the company, in our view, will launch in EVs mostly by 2023/24. Low total cost of ownership: This has been the core of the MSIL brand; it offers low total cost of ownership including best in class INR depreciates against JPY: ~11% of MSIL’s costs as a percentage of resale value. net sales are in JPY.

MORGAN STANLEY RESEARCH 97 M BLUEPAPER Ultratech Cement

Ultratech is in a sweet spot amid a strong demand out- Brand power is a sustainable advantage: Unlike in most parts of look given India’s housing shortage and the government the world, cement is a quasi-branded product in India as the leading focus on infrastructure and affordable housing and brands enjoy a pricing premium. Ultratech is among the leading slowing capacity additions brands, which is a sustainable advantage in our view, particularly in the housing segment of the market that currently accounts for With India’s current housing shortage and the government focus on around two-thirds of cement demand in India and more than 75% of infrastructure and affordable housing, we believe cement demand the company volumes. will remain strong over the next decade. This is also reflected in India’s cumulative per capita cement consumption at 3 tons (cumula- Cost focus will further boost the profitability profile: In the last tive since 1950) versus 8-22 tons for developed nations and leading few years, management has focused on initiatives like waste heat developing nations. In addition, with consumption-led loan growth recovery, increased usage of petcoke and alternate fuel and reduc- including housing loans, home upgrades will further support cement tion in lead time, with further additions of grinding units closer to the demand. With a consumption-led housing boom and the right execu- end markets. We think that continued focus on these areas leading tion on infrastructure, India could witness high-single-digit demand to cost efficiencies, will further boost the earnings profile. growth over the next decade, in our view. Where we could be wrong: Weaker than expected cement demand Given our view that timelines to add capacity have expanded in India and higher than expected capacity additions. and the cost of adding capacity has risen, we expect industry funda- mentals to remain attractive for incumbent producers in a strong demand growth environment.

Ultratech is in a sweet spot: Being the largest cement company in India and more importantly with a well-diversified pan-India presence, Ultratech is in a sweet spot in a strong demand environment. In addition, its strong balance sheet and capability to add brownfield capacity, better positions it versus peers from a capacity additions perspective, which will be key to delivering industry-leading volume growth.

98 M BLUEPAPER Alibaba

Exploring for Synergies quality sellers "passing strict quality guidelines and qualification criteria". This selection-based e-commerce listing likely learns from the India is one of the important markets for Alibaba to tackle during positioning of Tmall in China and helps Paytm Mall differentiate itself its global expansion. Alibaba is actively exploring business from other market place e-commerce platforms in India. Logistics- opportunities in this market, starting with the mobile payment wise, all products offered on Paytm Mall will also move through infrastructure (Paytm) and expanding to core e-commerce seg- Paytm Mall's certified warehouse and shippers, similar to how logis- ments (Snapdeal and Paytm Mall), supported by its in-house tics are handled on Tmall. Alibaba also invested in Snapdeal in August mobile browser business (UCWeb). 2015 and took a minority share of less than 5%.

In 2015 Alibaba first invested in Paytm, one of the largest mobile pay- Besides the core e-commerce and payment initiatives, Alibaba's ment platforms in India, and it has about a 9% shareholding as mobile browser App, UCWeb, has provided an additional touch point reported in the latest filing. Alibaba's partner in online payment and with Indian users. UCWeb has captured about 40% market share in eFinance, Ant Financial, is also a shareholder of Paytm, which brings the mobile browser segment in India as of July 2017, according to the total shareholding to around 40%. Alibaba and Ant Financial StatCounter. And UCWeb's mobile news content product, UC News have accumulated rich experience in mobile payment and back-end Feeds, has reached 100mn active users in India and Indonesia as of fintech systems, which could help accelerate the transaction growth March 2017, according to Alibaba. of Paytm in India. Where we could be wrong: India is a more complicated market than In March 2017, Paytm completed the spin-off its e-commerce busi- China, and Alibaba has no operational experience in the market. The ness, Paytm Mall. Including the newly issued preferred shares, Ali- results of Alibaba's initiatives will be affected by product localization, baba in total acquired a 36% equity interest in Paytm Mall on a fully- strategy execution and various other factors, some of which may be diluted basis, for a total cash consideration of US$177mn. Right after out of the company's control. the investment, Paytm launched an updated version of a mobile app that introduced a "Tmall-like" market place, which only includes high-

Morgan Stanley is acting as financial advisor to Cainiao Smart Logistics Network Limited ("Cainiao") in relation to the proposed acquisition of an additional stake in Cainiao by Alibaba Group Holding Limited as announced on September 26, 2017. Cainiao has agreed to pay fees to Morgan Stanley for its financial advisory services. Please refer to the notes at the end of the report.

MORGAN STANLEY RESEARCH 99 M BLUEPAPER Amazon

Making giant strides into the Indian eCommerce market Finally in 1Q17, Amazon announced that it had increased Prime Selec- and should be a key beneficiary tion by 75% since it launched and it debuted Inside Edge on Prime Video, the first of 18 Indian Original Series in 2Q17. Amazon is committed to developing its India business. It has highlighted spend on India numerous times on its quarterly Despite Amazon's presence in India still being relatively small, we do earnings call since 2015, and at a high level, any US$5bn invest- believe that India will be a material driver to Amazon’s long-term non- ment, in our view, speaks to commitment and interest. Germany/UK/Japan international retail business, which we model growing at a 38% forward CAGR, driving 17% of forward company- Focus on customers and selection: It is early, but Amazon is making wide retail revenue growth and 56% of international retail revenue progress. The company has 160mn products across hundreds of cate- growth ( Exhibit 162 ). In all, while we acknowledge Amazon's India gories currently. On its 4Q15 quarterly earnings call Amazon said that business is likely small now, we believe it presents an attractive long- its 4Q India sales were greater than all of 2014, but India is still small term opportunity for the company to tap into the ~US$200bn (by relative to Amazon’s total estimated ~US$180bn global GMV (in cal- 2026) Indian e-Commerce market. endar 2016). In 2Q16, Amazon launched Prime in India in over 100 cities, with one- and two-day shipping on hundreds of thousands of products. In 4Q16, Amazon added Prime Video at no additional cost.

Exhibit 162: We Believe Amazon's International Business (ex Germany/UK/Japan) - Which Includes India - Will Be a Material Driver to Companywide Retail Revenue Growth in $ millions Amazon 3-Year CAGR Contrib. to Growth Retail Revenue * 2013 2016 2019 '13-'16 '16-'19 '13-'16 '16-'19 North America 41,410 79,784 146,233 24% 22% 73% 70%

International Germany/UK/Japan 25,465 34,492 46,871 11% 11% 17% 13% Other Int'l 4,469 9,492 25,109 29% 38% 10% 17% International 29,934 43,984 71,981 14% 18% 27% 30%

Total Retail Revenue 71,344 123,768 218,213 20% 21% 100% 100%

Other Int'l Contrib. to International retail growth 36% 56%

Source: Company data, Morgan Stanley Research estimates for 2019; Note: "Other Int'l" includes India, France, and China; *ex AWS, ex other

100 M BLUEPAPER Investing in fulfillment to drive seller and buyer growth: Amazon Fashion, furniture and grocery: While Myntra and Jabong were is investing in fulfillment in India. We believe this is because the com- considered the early leaders in the fashion category, Amazon India pany sees the fact that the existing fulfillment infrastructure is sub- has increased its focus on the fashion category (in particular around scale. Logistics are likely to be Amazon’s competitive advantage (as mid-2015). It entered the furniture segment in late 2015 wherein it in other countries). Amazon's storage capacity and fulfillment infra- competes with vertical specialists like Pepperfry and Urban Ladder, structure in India has grown to 41 centres in 13 states with a storage as well as with Flipkart. The company has also launched its Grocery capacity of 13 mn cu ft. It also runs a ‘I Have Space’ program (IHS), vertical in Bangalore and reportedly plans to expand across more of wherein it partners with local store owners across different cities to India. As detailed in our latest AlphaWise global eCommerce survey deliver products to customers. Amazon India now has over 17,500 note (Here Comes Online Grocery), we have seen Amazon take an stores under this flagship program in over 225 cities. Like in other early lead in Grocery in other countries (like the US), so it stands to countries, we see Amazon investing in fulfillment and Fulfilled by reason the company will invest in this category in India as well. Amazon (FBA) to drive more user growth, prime subscriber growth, and grow its overall share of consumers’ wallets. Where we could be wrong: India is a completely different market than the ones that Amazon has traditionally been active in and hence exposes it to different operational and regulatory risks. Also, the eCommerce industry is still in its early days, thus making it competitive and demanding from a funding perspective.

MORGAN STANLEY RESEARCH 101 M BLUEPAPER DBS

digibank successful execution could drive strong Exhibit 163: growth in India and provide lessons for use across the DBS India in Numbers (S$m) region

DBS has a long and mixed history in India, but we believe that the medium-term investor focus will be on the build-out of its internet-based retail offering - digibank by DBS. Not only does digibank in India have the potential to become material relative to the rest of the group, but we also expect that the rest of DBS will benefit from what it learns in the build-out phase. DBS is also rolling out the digibank concept elsewhere; an Indonesian version was launched in late August 2017.

DBS in India

DBS Bank India Ltd., is DBS' operating business in India. It operates as a branch and as such faces the usual restrictions around number of branch licenses and so on - it has 12 branches in India.

Given that it cannot offer banking services more than 5km from a branch, this also restricted the expansion of digibank. However, on September 4 DBS announced that it has received an in-principal approval to move to a full operating subsidiary structure, and this will allow it to compete on equal terms with domestic Indian banks - as a result of this we expect DBS will increase its branch network to about 35-40 branches over five years.

Whilst DBS India is a small part of the whole (c.1.5% of loans and deposits, and de minimus profits), it does represent c.2.5% of revenues, in part due to the higher margins available to Indian banks. Even in the medium term therefore a strong roll-out of the digibank franchise in India (as well as Indonesia) could have a noticeable impact on DBS revenues, and could help drive upside to the Indian bottom line given the relatively low cost required to drive the business. Source: Company Reports, Morgan Stanley Research

102 M BLUEPAPER An important part of this argument is the relative size of India and The Product Suite Singapore - in its first 10 months, digibank in India had signed up 840,000 customers, equivalent to 16% of the entire population of The entry product is a digital wallet, which allows the customer to Singapore. If revenues were to expand in line with the branch net- make payments via an e-Visa card. DBS then tries to encourage cus- work, then we could see c.7.5% of group revenues from India in five tomers to open a savings account and offers attractive rates to do years, potentially adding 1% p.a. to our base case forecast of a 10% this (currently 7%). At the moment c.28% of wallet customers have CAGR in the top line. This will be another contributor to our forecast been converted to a savings account, and DBS is restricted to only for DBS' 2018-19 top-line growth of 11% (alongside ASEAN loan using agents for eKYC verification who are within 5km of a DBS growth and wealth management). Longer term, the impact could be branch (although this will be less of an issue when DBS moves to a more meaningful as digibank starts to become a more material part subsidiary structure and can roll out more branches). DBS estimates of the whole. Currently DBS India accounts for just 2% of group that c.71% of digisavings customers are active (i.e. they have carried equity, therefore we believe any capital impact of growing the India out at least one transaction in the last three months). These cus- business will be manageable. tomers are taking full advantage of UPI, with more than 70% of inbound and more than 50% of outbound transactions done through UPI. DBS will roll out more products over the coming months, What is digibank by DBS? including mortgages and unsecured personal loans as well as investment and insurance products. The Concept Where we could be wrong: We see two main risks for the DBS digi- digibank by DBS was India's first mobile-only bank. In DBS' own bank roll-out. The first is that they execute too slowly, and lose out words, it does "away with branches, forms, signatures and call- to larger competitors such as HDFC or Kotak. The second is that they agents. Instead, front and centre are biometrics, artificial intelligence get their credit scoring wrong and build up credit problems for later (AI), analytics and dynamic security". Most customers come in via the once they begin to offer lending products. e-wallet, which is operated by downloading an Android, or iOS App onto a mobile phone. They can then migrate to a savings account. There is no need to go to a branch, and accounts can be opened using a thumbprint and Aadhaar ID, although if customers upgrade to digisavings they need to visit one of DBS partner stores (eg. Feno, or Cafe Coffee Day) for biometric ID verification. There is a call centre for digibank customers, however DBS state that 80% of calls are handled by an AI driven Chatbot. The capabilities of this Chatbot should improve over time. Ultimately, DBS believe that only 12% of interactions will go through agents.

MORGAN STANLEY RESEARCH 103 M BLUEPAPER Naspers

India offers significant value potential for Naspers given Flipkart: As of August this year, Naspers owned a 16.5% stake in Flip- increasing eCommerce penetration and from a shift kart (it has subsequently raised capital from Softbank and others). towards online payments. Naspers has exposure to the Naspers estimates that the Flipkart Group is India’s largest e-com- leading eCommerce players and owns and controls merce marketplace and had a 55% market share of Indian e-com- some payment gateway companies. merce gross sales in March 2017. Amazon is its largest competitor. Group companies include Flipkart, Myntra, Jabong and PhonePe. Flip- India offers significant value potential for Naspers and it is not kart offers over 80m products across 80+ categories. priced in. Naspers is a South African based conglomerate that owns a 33% stake in Tencent as well as stakes in a number of other internet OLX: OLX is the leading generalist online classifieds platform in businesses and Pay TV. The current valuation implies a market cap of India. The app has already been downloaded 35m times in the last US$98bn versus its Tencent stake which is worth US$131bn. Naspers two years alone and we believe this could increase to 116m in the next invested early in China through Tencent, and its investments in India five years. Classifieds require building a strong network for monetiza- were a logical next step given the market potential. We currently tion and given the potential to add on more subscribers we believe value its Indian investments at >US$5bn and see potential for addi- the company will take a little longer to monetize, but we see it poten- tional upside as companies like Flipkart continue to compound tially generating US$100m+ of EBITDA on a five-year framework. Its growth and classifieds start to monetize. main competitor is currently Quikr. Amazon recently announced it is also launching a classified offering. For our investment case see Naspers: Can the discount get any wider? (12 Jul 2017) PayU: PayU is a regulated financial institution and holds licenses from several national banks and local regulators. Its products include Naspers should benefit from increasing eCommerce penetra- a digital consumer wallet, a PCI DSS certified payment gateway, anti- tion… Naspers has exposure to leading players in all the key verticals fraud systems and an online Visa/MasterCard acquirer. PayU including online travel (MakeMyTrip covered by Parag Gupta), online acquired e-wallet Citrus Pay for US$130m in September 2016. As dis- retail (Flipkart), classifieds (OLX) and food delivery (Swiggy & cussed in detail in this report, the growth prospects for online pay- Delivery Hero / foodpanda). ments in India are material.

...and from the shift towards online payments. Naspers owns and controls payments gateway company PayU, e-wallet Citrus Pay and a 37% stake in digital banking services provider Kreditech. Whilst we believe it is smaller in scale relative to Paytm, growth is so significant that we believe there is room for multiple players.

Exhibit 164: Naspers' Exposure to India Value to Business Estimated value Ownership Naspers % of NAV Notes U$m % U$m % MakeMyTrip 4,842 40% 1,937 1.4% Valued on our target price for MMYT Flipkart 11,600 17% 1,914 1.4% Valued on funding round in August 2017 OLX India 876 100% 876 0.7% We estimate that India could represent ~12% of the total OLX value PayU / Citrus Pay unknown unknown 660 0.5% Pay U acquired Citrus Pay for $130m in September 2016 Delivery Hero 4,842 11% NM NM $513m is full value of stake in DH. India would be a small portion of this Swiggy unknown unknown NM NM Assumes Naspers led 50% of the recent $80m funding round in Swiggy 5,387 4.0%

Source: Company data, Morgan Stanley Research estimates

104 M BLUEPAPER Swiggy / Delivery Hero: Naspers recently led a US$80m funding MakeMyTrip: Covered by Parag Gupta. After the merger of Ibibo and round in Swiggy. It also has an 11% stake in Delivery Hero, which owns MakeMyTrip, the company has become the undisputed leader in foodpanda in India. They are both leading players in the market com- Indian Travel. See Street Comparing Apples to Oranges; Opportunity peting with Zomato and UberEATS. Swiggy has attracted over to Buy (15 Aug 2017). 12,000 restaurants to its platform and has experienced a six-fold growth in revenue over the past 12 months. The company runs its Where we could be wrong: Slower growth in the Indian internet own delivery fleet and claims to have an industry-best average of 37 industry, higher burn rates or significant slowdown in investment minutes per order. For more on Delivery Hero see Guardians of the activity could hurt the value of the portfolio companies. ganoush (8 Aug 2017)

MORGAN STANLEY RESEARCH 105 M BLUEPAPER Softbank

High level of commitment to investing in platform busi- 3) Financial settlement: Softbank invested in India’s Paytm digital nesses in India. Softbank’s investments in India are payments company in May 2017. The company’s subsidiary Alibaba focused on platform business, undertaken with a had previously taken a stake as well. The company has not under- medium- to long-term view, we believe. A ~US$200bn taken significant investments in financial settlement services in (by F2027) Indian eCommerce market becomes an other regions, but it did invest in US online loan arranger Social attractive opportunity for the company. Softbank is Finance, Inc., in October 2015. keen to hone platformer capabilities in anticipation of IoT. Where we could be wrong: Softbank's global investment has a good track record as its aggregate performance, but it also recognised We estimate that Softbank has thus far invested a total of about some significant write-offs. If the major portion of the US$4bn US$4bn in India, starting with its application of a stake on online investment in India becomes at risk, then it might shift investment marketplace Snapdeal in 2014. In December 2016, Softbank targets to other regions. president Masayoshi Son indicated that the company plans to invest more than US$10bn in India over the next 10 years. In con- Exhibit 165: junction with other investors including Saudia Arabia’s Public Softbank’s Major Investment Returns in India Investment Fund (PIF), the company established the Softbank Company name Business category Vision Fund, through which it plans further aggressive investment activities. 2014/10 Snapdeal B2C EC platform 2014/10 Ola Transportation service 2014/10 Housing .com Real estate information site Softbank’s investments in India are focused on platform business and 2017/5 Paytm Payment service we recognize that they are undertaken with a medium- to long-term 2017/8 Flipkart B2C EC platform view. We also think that Softbank’s investment stance (not only in Source: Company data, Nikkei and Morgan Stanley Research India) is, at this point, guided less by the pursuit of clear synergies with existing business than by the incremental coalition-building through capital participation with the founding families of the com- Morgan Stanley is acting as financial advisor to Fortress Investment panies in which it is investing, with the aim of securing business Group LLC (“Fortress”), in connection with its definitive agreement opportunities at an increasing pace over the longer term. under which SoftBank Group Corp. will acquire Fortress, as announced on February 12, 2017. The transaction is subject to Softbank’s major investment returns in India approval by Fortress shareholders, certain regulatory approvals and other customary closing conditions. This report and the information 1) B2C e-commerce: Softbank has invested in the domestic e-com- provided herein is not intended to (i) provide voting advice, (ii) serve as merce leader Flipkart (August 2017) and the number three in the an endorsement of the proposed transaction, or (iii) result in the pro- online market Snapdeal (October 2014). It has also already invested curement, withholding or revocation of a proxy or any other action by in e-commerce platformers in other regions, and is generating a security holder. Fortress has agreed to pay fees to Morgan Stanley returns on its investments in Alibaba (China), Tokopedia (Indonesia), for its financial advice, including a transaction fee that is contingent and Coupang (Korea). upon the consummation of the proposed transaction. Please refer to the notes at the end of the report. 2) Transportation services: Softbank invested in ride-hailing company Ola in October 2014. The company has also invested in similar companies in China (Didi Chuxing) and South-east Asia (Grab), as it steadily builds a global structure for transportation services.

106 M BLUEPAPER TransUnion

Beneficiary of Credit Expansion in India at the end of 2010 to 0.57% in 1Q15. Due to additional data and analytics, we would also expect that over time credit will be further With the strong growth of the Indian economy and the emerging extended to lower score borrowers, with CIBIL indicating that as of middle class in the early stages of becoming credit active, we June 2015, 79% of all retail loans across India were for individuals with expect TRU to be a key beneficiary, as it owns 92% of TransUnion scores above 750. TRU's revenue model is volume-based, as financial CIBIL Ltd, India's leading credit information company. institutions purchase data and solutions to determine whether to lend money to a business or individual. Thus, an increase in credit Large and growing database: TransUnion CIBIL has over 2,400 application volumes drive additional growth within TRU's Interna- members including all leading banks, financial institutions, non- tional segment. CIBIL products sold include analytics and decision- banking financial companies and housing finance companies. As of making solutions for banks, telecommunication companies, and February 2007, its credit database includes information on over 270 insurance companies, as well as online credit reports and scores for million consumers and 13 million businesses, a ~30-35% increase consumers. from its disclosure of over 200 million consumers and 10 million business entities as of February 2016. To enhance its solutions, CIBIL India contribution currently small, but growing fast with high also pulls from non-credit data sources including the national voters potential. Though CIBIL was started in 2000, TRU has rapidly registry with 750 million recorders, the national ID database with increased its stake in the last few years, going from 27.5% ownership over 1.1 billion records, as well as the confirmed and suspected fraud in 2013 to 92% today. Back in 2013, TRU's equity stake in CIBIL dis- registry, property registry, and tax ID database. Using its extensive closed on its balance sheet implied a US$103mn book value, while its database and TransUnion's analytics capabilities, TransUnion CIBIL most recent 10% increase in ownership implied a ~US$587mn valua- developed and launched the first generic credit score for India, which tion. Though TRU does not explicitly disclose the revenue contribu- is the most widely adopted credit score across the country. tion, we estimate CIBIL generated ~US$45mn of revenue in 2015 (~3% of TRU revenue), but expect this unit to be growing faster than Data drives lending: Increased availability of data drives additional the overall business. In TRU's emerging market business, we forecast lending, as lenders can make a more educated decision on the risk of a ~9% CAGR between 2016-2021, with the India business being a the borrower. This has led to strong growth in retail loans, as well as large driver of this. With a large opportunity for growth in credit and a shortened time period for loan approvals. A 2016 CIBIL report CIBIL currently the only meaningful credit bureau, we see strong found that retail loans were growing at a 28% CAGR over the prior upside for TRU in the Indian market. three years, and the average time of a loan approval accelerated to 3-4 days from 7-9 days over the same time period. CIBIL also indicated that increasing data has improved the quality of loans, with delin- quency of repayments (90 days and more) coming down from 1.06%

MORGAN STANLEY RESEARCH 107 M BLUEPAPER

Exhibit 166: Emerging Markets, Which Incorporate India, Will Drive a Larger Portion of Growth in the Coming Years in $ millions 3-Year CAGR Contrib. to Growth TransUnion 2013 2016 2019 13-'16 16-'19 13-'16 16-'19 U.S. Information Services 749 1,045 1,325 12% 8% 52% 64%

International Developed Markets 89 109 139 7% 8% 3% 7% Emerging Markets 152 205 270 10% 10% 9% 15% Total International 241 314 409 9% 9% 13% 22%

Consumer Interactive 204 407 470 26% 5% 36% 14%

Total Revenue (Ex-Eliminations) 1,194 1,766 2,204 14% 8% 100% 100%

Emerging Market Contribution to Int'l Growth 73% 68%

Source: Company Reports, Morgan Stanley Research estimates

Equifax data breach implications: Over the long term, we think the Where we could be wrong: We see the following potential risks: fundamental strength of TransUnion's business is intact and believe India represents a compelling growth opportunity. However, with l An economic slowdown could drive lower demand for the disclosure that credit bureau competitor Equifax suffered a mas- credit information. sive data breach (with 143 million US consumers impacted), near- l Higher US interest rates could pressure mortgage and card term stock movements for TRU will likely center around the news applications. flow around the breach. We believe that TRU could benefit from l Contributing data to free or freemium consumer moni- share shift (as customers may decide to switch from EFX to one of the toring service providers could cannibalize TRU's existing other bureaus), and TRU's consumer business may see increased direct-to-consumer business. business from customers more worried about ID protection. On the l Leverage remains at the high-end of our Analytics cov- other hand, it is at risk of tighter industry regulation. As it relates to erage. cybersecurity, TRU management has indicated that it continues to l Increased regulation as a result of EFX cybersecurity focus on having the best processes, software, partners, and tech- breach could lead to industry changes or increased costs of nology to help mitigate their risk. compliance.

108 M BLUEPAPER Visa and Mastercard

Digitization Should Grow the Pie for Visa and Master- l Innovations such as BharatQR should allow rapid growth Card in card acceptance: Over 1 million new acceptance points have been added in India since demonetization (November Card payments represent only 5-7% of PCE (Personal Consumption 2016), relative to only 250,000 new acceptance points Expenditure) in India. As recent regulatory and technological devel- added in the 12 months prior to that. Of these, 200,000 opments steepen the digital adoption curve in India, we expect the are QR code based acceptance points. Innovations like absolute value of non-cash payments, including utilization of card BharatQR, which have been developed by collaborative networks, to grow rapidly. As the digital payments pie grows, both efforts among Visa, Mastercard, and the NPCI, allow mer- networks (Visa and Mastercard) should benefit, particularly as initia- chants in India to leapfrog the traditional (and more expen- tives like BharatQR allow merchants in India to leapfrog the tradi- sive) POS card acceptance terminal investments. Visa tional (and more expensive) POS card acceptance terminal believes that QR and contactless will comprise 25% of its investments. At the same time, the lack of a well-established card total transactions in the next 3-5 years. Additionally, acceptance infrastructure and the rapid pace of change should allow mobile point of sale or mPOS devices, which also provide a newer, non-card digital payment innovations to also gain traction. low cost acceptance solution to merchants, are starting to Credit and debit cards may see their mathematical grip on digital pay- gain traction and are anticipated to grow rapidly. ments loosen, but the pie should grow for everyone. l Strong growth in card issuance: There are approximately 850 million credit and debit cards outstanding in India, Can Visa and Mastercard remain dominant players in the India giving the networks a head start versus any new digital market amidst growth of new entrants? payment types that have yet to build scale. New players are also fast gaining traction –– e.g. Paytm exceeded A key question is whether Visa's and Mastercard’s differentiation in 200mn registered users earlier this year. But Visa and Mas- terms of global operability, superior security tools, and experience in tercard also plan to accelerate card issuance. Visa has innovation can enable them to remain among the dominant players stated that over the next 3-5 years, it expects to issue an in the market. We believe the answer is yes. incremental 200 million cards, both physical and digital. l Ability to leverage global innovations: Visa and Mastercard l Global scale and reach: Visa and MasterCard are interna- have a broad suite of product innovations globally that can tional schemes with global acceptance at ~40mn merchant be leveraged on to suit the unique needs of the Indian dig- locations worldwide. Many of the new entrants sprouting ital payments market. The Government of India's collabora- in India are still working to develop domestic acceptance tion with the networks for development of the BharatQR with no presence internationally. Meanwhile, thanks to a code is a reflection of that. combination of lower travel costs and a growing affluent/ l Marketing and advertising engine: Having done this in middle class in India –– affluent income categories in India many markets over and over, Visa and Mastercard have are growing the fastest and driving a disproportionate well developed tools in their standard playbook on share of increasing consumption –– India is rapidly building brand awareness and preference and improving becoming one of the fastest-growing outbound travel mar- card usage at the POS. Both are investing in marketing kets in the world, second only to China. This will likely activities, which are being amplified by issuing and mer- favor growth and usage of global payment schemes such chant partners. e.g. seven public sector banks in India that as Visa and Mastercard in the affluent income categories, account for ~20% of debit cards in-force leveraged Visa's in our view. Additionally, the underlying specifications for campaign and placed it within their marketing toolkit on BharatQR developed by Visa, Mastercard, and NPCI their bank web pages and digital channels. (National Payments Corporation of India) can be implemented in other countries to deliver a globally interoperable solution for credentials issued digitally in India.

MORGAN STANLEY RESEARCH 109 M BLUEPAPER Risks to Visa and Mastercard dominance: Sizing the Addressable Opportunity:

Fee regulation may curtail interest from card issuers: Reserve Growth Off a Negligible Base Bank of India has a new proposal for regulation of Merchant Discount Rates (MDRs) (See Exhibit 21 ). A significant reduction in MDRs Credit and debit card spend aggregated to ~US$88bn in total pur- could be a deterrent for issuing and acquiring banks (as all acquirers chase volume in India in CY2016. This is a rounding error for Visa/Mas- in India are issuers as well), although we have not found this to be the tercard payment volumes today at <1% of Visa's and Mastercard’s case in other markets where interchange fees have been regulated combined 2016 purchase volume and ~3% of Visa's and Mastercard’s (Australia, UK, Spain, US debit, and so on). While lower economics combined volume in the APAC/APMEA region. Credit and debit card may not be viable for banks in the near term given that the card busi- spend made up ~90% of total digital spend (credit + debit + Rupay nesses are subscale, this could change if digitization takes off materi- + m-wallets + UPI) in India during this period. ally as the increase in volumes could more than offset the loss in spread for the issuing/acquiring banks. Our base case assumption is for digital payments to grow from 8% of PCE in F2017 to 36% in F2027. If this comes true, ~US$1.2 trillion of Networks may have less flexibility on pricing versus new disrup- payments would be transacted digitally in F2027, which is a good tors: As new entrants try to build share, some may be aggressive on proxy for CY2026. If card networks can maintain 45% share (versus pricing in order to build usage. e.g. new disruptor Paytm waived off ~90% today), total purchase volumes on card would be ~US$600bn, the merchant fee to Zero temporarily. Although it may be a money- representing a CAGR of ~20%. The exhibit below shows some sce- losing proposition in the near term, Paytm has access to some fairly narios of what the network revenue contribution from India could deep pockets (Alibaba and Softbank) and is likely to view these potentially look like in 2026, assuming much lower than average net- losses as investment in acquiring customers at scale, which can be work yields given the potential for fee regulation in India. Our anal- monetized later. Visa and MasterCard may be restricted in their ysis assumes that Visa and Mastercard make up the majority of share ability to compete on price, despite their low cost structure, as they of the credit and non-Rupay debit card market. While American need buy-in from the issuing and acquiring banks to forego eco- Express and Discover cards may also have a presence in India, we nomics from merchants in exchange for volumes. And banks may be expect their share is fairly low. more averse to cutting their card revenue pools relative to tech dis- ruptors. Note that given the solid potential for growth in purchase volumes in developed parts of the world, India may remain a relatively small contributor to overall growth rates for Visa and Mastercard longer term, but it represents one of many growth drivers that could help sustain the double-digit growth rates for the companies for an extended period of time.

Exhibit 167: Visa and Mastercard Aggregate Revenue Potential from India in 2026 2016 2026e India Digital Payment Volume ($bn) 98 1200 CAGR assumption 28.0%

Estimated Revenue Potential in $millions V/MA Market share 90% 25% 35% 45% 55% 65%

Network Fee yield 0.050% 150 210 270 330 390 0.055% 165 231 297 363 429 0.060% 180 252 324 396 468 0.065% 195 273 351 429 507 0.070% 210 294 378 462 546

Source: Company Data, RBI, Morgan Stanley Research. e = Morgan Stanley Research estimates.

110 M BLUEPAPER Current State of India's Card Payment Rupay domestic scheme market Card type: The Rupay card scheme was launched in 2012 by the National Payments Corporation of India (NPCI) as a India's card market is dominated by debit cards... debit only scheme. This domestic scheme was conceived The total number of credit and debit cards outstanding to promote financial inclusion in India, while offering has grown rapidly from 280mn cards at the end of 2011 banks a more affordable payment solution vis-à-vis what to 739mn cards at the end of August 2016, representing was being offered by international schemes like Visa and an increase from 0.23 cards per capita to 0.56 cards per MasterCard. Although NPCI had plans to eventually capita over this period. The growth in card issuance has launch Rupay credit cards as well, that could take some largely been driven by debit cards and is in part time. attributable to the introduction of India's domestic cards, Acceptance: Rupay started off with limited acceptance, Rupay. Debit cards represented ~96% of total cards but through partnership with acquiring banks, Rupay outstanding as recently as 1H2017. Rupay debit card cards are now accepted at all ATMs and POS terminals in issuance has grown rapidly from 17mn or less than 5% of India. Rupay expanded acceptance to online transactions all debit cards outstanding in India at the end of 2014 to in April 2013 and is now accepted at over 20,000 e- 321mn or nearly half of all debit cards, based on recent commerce merchants in India. Internationally, through an data from India's Department of Financial Services. agreement with Discover, Rupay cards can be used at all Exhibit 168: Discover and Diner POS and ATM networks. Visa and Mastercard Cards Dominate Transaction Transaction costs: Rupay was designed to be a low-cost Volume, Even Though Rupay Has a High Share of Total scheme for banks with a fixed fee of INR 0.45 per Cards transaction for ATM use and INR 0.90 per transaction for Debit cards1 - Rupay vs. V and MA Debit card POS transactions - Rupay vs. V and MA POS or e-commerce transactions. This compares to Rupay transaction 8% international schemes like Visa and MasterCard that Rupay All other debit cards debit cards 45% typically charge variable fees based on the transaction 55% All other debit transaction value. 92% Issuing banks: Rupay cards are offered by ~40 or so Source: National Payment Corporation of India, PMJDY (Department of Financial Services, Government of public, private and small cooperative banks in India. The India) international schemes typically work with the larger public and private banks and may not include the small … but credit cards dominate card spending at POS rural cooperative banks in their networks. Credit cards account for only 4% of total outstanding cards, but they constitute a majority of the card-based spend at the POS, although the mix of credit card transactions has been declining and more recently at an accelerating pace. In 2017, credit cards represented ~50% of total card-based POS spend, down from >60% in 2012. Debit cards have gained some share, largely after the demonetization initiative in November 2016. But use of Rupay cards for retail transactions still remains relatively low versus overall debit card utilization at the POS, at 8% during the month of August 2016. One of the reasons for this divergence could be that the growth of Rupay card issuance has been relatively high in remote/rural areas, which have limited POS acceptance infrastructure and low consumer education/awareness.

MORGAN STANLEY RESEARCH 111 M BLUEPAPER Appendix 1 - JAM: Catalyzing Digitization

Three major trends ("JAM") have emerged in India over the Jan Dhan Yojana - The government launched this plan in August 2014 past 6-7 years – creation of a universal identification / to ensure complete financial inclusion for the country's citizenry. authentication mechanism for the entire population; Since then, ~285mn bank accounts have been opened under this ensuring every household in India has access to a bank scheme. This is a no frills account offered to economically weaker account; and a surge in mobile penetration and a rapid sections of the society. Prior to the launch of JDY, ~35% of house- pickup in smartphone usage. Government and policy- holds in India did not have a bank account. But, since August 2014, makers realized that leveraging the combination of these most Indian households, we believe, have gained access to a bank trends would yield a powerful force for economic devel- account. opment, and this has given rise to the JAM initiative.

Exhibit 169: ~285mn Bank Accounts Have Been Opened Under the PMJDY Scheme (~23% of India's Population, and Much Greater Proportion of Households)

% OF ZERO- Particlualrs BALANCE IN Balance Per RURAL URBAN TOTAL BALANCE- (Data in Mn) ACCOUNTS A/C (Rs) ACCOUNTS F1Q15 0.0 0.0 0.0 0 NA NA F2Q15 31.7 22.1 53.8 42,732 76.8 795 F3Q15 62.3 42.1 104.5 83,534 73.3 800 F4Q15 87.8 59.3 147.2 156,703 57.9 1,065 F1Q16 99.0 65.3 164.3 190,154 51.8 1,158 F2Q16 112.6 72.8 185.4 249,392 40.3 1,345 F3Q16 121.0 77.3 198.4 292,256 31.8 1,473 F4Q16 131.7 82.6 214.3 356,720 27.4 1,665 F1Q17 136.9 86.0 222.9 392,516 25.3 1,761 F2Q17 151.8 95.5 247.4 435,327 24.1 1,760 F3Q17 159.6 102.4 262.0 710,366 24.1 2,711 F4Q17 168.7 113.0 281.7 629,724 2,236 F1Q18 171.0 115.3 286.3 643,649 2,248

Source: PMJDY website

Exhibit 170: Percentage of Population Over the Age of 15 With Accounts at a Financial Institution, Across Countries (2014) 100

50 99 99 99 98 98 97 97 96

88 88 87

82 82

81 79

70 69 68

63 53 25 50

36 31

25 13 0 Italy India Chile Brazil Spain China Japan Jordan France Mexico Greece Greece Pakistan Australia Malaysia Mauritius Germany Argentina Indonesia Singapore Philippines Switzerland South Africa South Saudi Arabia Saudi United States United Czech Republic Czech United Kingdom United

Source: World Bank Findex Database, Morgan Stanley Research 112 M BLUEPAPER Aadhaar - The intent of this programme (launched in 2010) was to Mobile phone penetration - There has also been sharp surge in ensure that every Indian has an officially recorded identity. Over 6-7 mobile phone usage in India, with the current number of mobile con- years, this has created a biometric digital database of 1.2bn people, nections at ~90% of the population. Fairly cheap telecom tariffs and ~92% of the country's population. Prior to this, nearly half of the the launch of no-frills low-priced phones has aided penetration. country didn’t have any form of identification. Individuals in the Aad- There has also been a sharp increase in smartphone penetration, with haar database have a 12-digit digital identity, which can be authenti- ~25% of mobile phones in the country today being smartphones. cated by fingerprints and retina scans. Since 2012, the share of smartphones in total shipments has increased from 7% then to 50%, and this could reach 100% by 2020. We expect smartphone penetration to surge further, reaching over 50% of total mobile penetration by F2021.

Exhibit 171: Number of Aadhaar Accounts Exhibit 173: 1400 India: Smartphone Penetration Has Surged

1200 900 (mn) 60% 764 800 49% 1000 673 50% 700 685 570 800 600 40% 490 31% 546 600 500 401 347 426 30% 400 322 22% 400 283 300 223 292 20% 140 200 200 11% 10% 100 0 - 0%

2010 2011 2012 2013 2014 2015 2016 CY14 CY15 CY16 CY17E CY18E CY19E CY20E

Aug-2017 Wireless Data Subs Net Smartphone Penetration

Source: UIDAI, Morgan Stanley Research 4G phones (Penetration) % Smartphone penetration Source: Statista, TRAI, Morgan Stanley Research, E= Morgan Stanley Research Estimates

Exhibit 172: Authentications Using Aadhaar Exhibit 174: Total AuthenticationTransactions (Rs Mn) India: Sharp Pickup in Data Subscribers 1200 Total Data Subscribers ('000)

450,000 1000 400,000 800 350,000 300,000 600 250,000 400 200,000 150,000 200 100,000 0 50,000 - Jul-16 Jul-17 Oct-16 Apr-17 Jun-16 Jan-17 Jun-17 Feb-17 Mar-17 Aug-16 Sep-16 Nov-16 Dec-16 Aug-17 May-17

Source: UIDAI, Morgan Stanley Research F1Q13 F2Q13 F3Q13 F4Q13 F1Q14 F2Q14 F3Q14 F4Q14 F1Q15 F2Q15 F3Q15 F4Q15 F1Q16 F2Q16 F3Q16 F4Q16 F1Q17 F2Q17 F3Q17 F4Q17 Source: TRAI, Morgan Stanley Research

MORGAN STANLEY RESEARCH 113 M BLUEPAPER A challenge in an emerging economy such as India had been that a Exhibit 175: large proportion of the population was undocumented; they did not India: Wireless Data Cost have access to financial services, and there was no proper way to con- 450 Airtel blended data price (Rs/GB) nect them with the rest of the world. This has been solved through 400 RJio data price (post promo) Rs/GB the implementation of JAM. 350 300 250 India Stack - Bringing this all together 200 150 The success of the above three trends (JAM) has ensured that mecha- 100 nisms to solve identity issues, lack of banking access, and need for 50 - physical infrastructure have been set up. The subsequent leap over

the past year or so has been bringing these together into single eco- F4Q12 F1Q13 F2Q13 F3Q13 F4Q13 F1Q14 F2Q14 F3Q14 F4Q14 F1Q15 F2Q15 F3Q15 F4Q15 F1Q16 F2Q16 F3Q16 F4Q16 F1Q17 F2Q17 F3Q17 F4Q17 F1Q18 system. This is being driven by India Stack. India stack is a set of appli- Source: Company data, Morgan Stanley Research cation programming interfaces (APIs) that leverage this infrastructure, moving towards presence-less, paperless, and cash- Exhibit 176: less service delivery. India: The JAM Trinity

A number of APIs have been launched, with full regulatory backing, which utilise this infrastructure. Five of the key APIs are e-KYC (which Jan Dhan Yojana – allows electronic know-your-customer verifications, making cus- Banking for all. ~285 mn J new accounts in last three tomer acquisition quick and very low cost); UPI (which allows mobile- years based payments on a real-time basis across individuals and businesses); Bharat QR code (an interoperable QR code system with Aadhaar – Unique identifier A for all Indians. About 1.2 bn all the key payment gateways and banks on board); Bharat Bill Pay people enrolled System (moving all bill payments in India - telecom, utilities, government, and so on - to one platform); and Aadhaar Enabled Pay System (enables customers to conduct transactions using Aadhaar Card and Mobile – Increased usage fingerprints for biometric authentication). M lowers the need for bank branches

This may appear to be a long list of initiatives. But they essentially cover the entire range of payments in the country. All one needs to Source: Morgan Stanley Research make these payments is a mobile phone and bank account. These are real time and, given that these systems are also regulator backed, they are interoperable - which should help kickstart digital payments Exhibit 177: in the country. We explain these five in some detail. Various APIs of India Stack

1. e-KYC - Electronic know-your-customer systems enable individuals to grant service providers access to personal identification information. This makes KYC almost instantaneous – and secure and paperless – as compared with traditional KYC practices (e.g., opening a bank account, which generally can take a few days, making such a process relatively costly). Under e-KYC, UIDAI only shares demographic information (name, address, date of birth, gender, photo- graph, mobile number) that is collected during Aadhaar enrollment with the consent of the customer.

Source: India Software Product Industry Round Table (iSPIRT)

114 M BLUEPAPER RBI has also allowed banks to open some new accounts using one- Exhibit 178: time PINs (OTPs) on mobile phones. This is likely to quicken the Monthly Transactions Using UPI process of opening such accounts. These are relatively small value Unified Payment Interface, Transaction Values (Rs Bn) 42.0 accounts, with a maximum deposit of Rs100,000. Still, banks are 35.0 required to conduct proper due diligence on the customer within a 28.0 year of opening these accounts, after which they can become stan- 21.0 dard customer accounts; such due diligence can be achieved just by taking the fingerprint of a customer to authenticate the Aadhar 14.0 details. 7.0

0.0 Jul-17 Jul-16 Apr-17 Apr-16 Oct-16 Jun-17 Jun-16 Jan-17 Feb-17 Mar-17 Nov-16 Dec-16 Aug-16 Sep-16 May-17 This is enabling customer acquisition at banks – quickly and at a very May-16 Source: NPCI, Morgan Stanley Research low cost. e-KYC reduces the number of people needed to acquire new customers, frees up physical space, and reduces requirements for back-office personnel for customer acquisition. This also applies to There have been other mobile phone apps that allow easy transfer of other services in which customer onboarding requires KYC – for funds, for example digital wallets. However, these are not inter-oper- instance, selling mutual funds to new customers, and targeting new able, i.e., there's no common system. The attraction of UPI lies in the customers for mobile phone subscriptions. fact that a consumer can link any number of bank accounts to a single UPI app. So, if an individual's primary account is, for example, at HDFC Bank, and that person has other accounts, for instance at ICICI and An example of banks using this feature is the launch of HSBC, the customer can use a single app (i.e., any bank or payment '811' accounts by Kotak Bank. In April 2017, Kotak Bank service provider's UPI-based app) and link all the accounts to that one announced the launch of 811 bank accounts. These app. When the account owner transfers money, they choose among accounts are set up using e-KYC (Aadhaar number, tax their linked accounts. The cost of a UPI transfer on P2P is also fairly number, and a onetime password). Depositors can use low, at ~Rs.0.5 per transaction, which will likely aid its growth. such accounts for basic banking services, and maintain balances of up to Rs100,000. No human interface is The other significant advantage of UPI is that it is a push (payer initi- needed to open these accounts; customers download the ates transaction) as well as a pull product. While push is useful for 811 app and open the bank account on their own. Kotak P2P transactions, pull will be important for merchant transactions. has ambitious growth plans through 811 – targeting a This allows merchants to ask for payment from a customer as long doubling of its total customer base over 18-24 months, as both parties have mobile phones and a bank account. This obviates from 8 million to 16 million. the need for payment infrastructure.

However, banks are still charging MDR (merchant discount rate) on these transactions in line with debit card MDRs. Over time, we expect 2. Unified Payment Interface (UPI) - This is the killer app for pay- these MDRs to decline sharply, causing a sharp increase in merchant ments in India, in our view. This protocol enables instant funds acceptance of payments through UPI. Also, given the lack of required transfer (24x7, both payment and collection) via a smartphone (only) infrastructure, even small transactions (e.g., paying cab fares) can be with the use of a virtual address (e.g., Aadhaar card, mobile number), made through UPI. just like a text message (avoiding multiple card/bank account details/ OTP, and so on). The only thing needed to transfer money is the virtual payment address (for instance [email protected]); there is no need to share bank account details.

MORGAN STANLEY RESEARCH 115 M BLUEPAPER

Exhibit 179: UPI Merchant Transaction - Schematic Representation of a Merchant Transaction

1 Customer's Bank 3 Customer's virtual address Send notification for 6 7 2 payment Debit Successful Identifies customer's PSP ABC bank debit basis PSP APP (Merchant) Request and routes to ABC bank PSP App (Customer) Merchant's Merchant vpa@abcbank NPCI (UPI) Payer's PSP MPIN Customer PSP 5 PSP validates the customer details Submit 9 Submit Send the financial details to UPI for debit 8 Successful Credit to customer account Request credit 4 MPIN entered by Customer Merchant's Bank 10 Transaction Confirmation

Source: Morgan Stanley Research

Exhibit 180: Bharat QR Code: An Introduction

Source: Morgan Stanley Research

Customers can use their bank apps to scan a merchant's QR code, and Bharat QR (Quick Response) System: The lack of payment infra- pay using any of the payment networks (Visa, Master, Amex or structure has been one of the significant challenges in making elec- Rupay). Alternatively, the customer can also use the UPI app to scan tronic payments in India. To circumvent this problem, various the QR code and make payment from their bank accounts (similar to payment gateways (Visa, Master, Amex) had their own QR-code- a debit card transaction). based systems. However, these were not interoperable. In 1Q17, NPCI collaborated with the three payment networks to launch the Bharat This is in addition to UPI as a payment app. In our view, customers may QR-code-based payment system. In this payment method, the mer- still want to pay using cards – as they get a credit period on credit chant has to download the QR code - either static or dynamic - and cards, collect points, and so on. For those customers, Bharat QR code this can be scanned by the customer on his mobile, enabling him to removes the hinderance of lack of payment infrastructure available make a payment. at merchant establishments.

116 M BLUEPAPER Bharat Bill Pay System (BBPS) - A major component of any retail There will be multiple operational entities to facilitate bill payments. payment transaction is bill payment. An RBI committee studying the The operating units will on-board billers, aggregators and payment feasibility of giro-based payments in India estimated that in the top gateways, and will set up the agent network and customer touch 20 cities, almost 31 billion bills are generated annually, with Rs6 tril- points to handle bill payments via various delivery channels (self- ser- lion in payments every year. Over 70% of these transactions are still vice, assisted, electronic and manual modes). So far, >60 approved predominantly being carried out by cash or cheques. operating units (called BBPOUs) have been identified, of which ~52 are banking entities. Additionally, the present bill payment infrastructure in India is largely biller-specific in terms of modes of payment accepted and Aadhaar-enabled Payment System (AEPS) - This is primarily channels (payment gateways) supported. Its not interoperable and geared toward servicing economically weaker sections of the society. does not offer and 'anytime, anywhere' payment option to cus- It facilitates inclusion in online interoperable financial transactions at tomers. PoS (MicroATMs) through the business correspondent (BC) of any bank using Aadhaar authentication. AEPS enables payments at a PoS The Bharat Bill Payment System (BBPS) has been launched to cir- (MicroATM) without requiring a mobile phone or a debit / credit card. cumvent these constraints. It is an integrated bill payment system The only inputs required for a customer transaction under this sce- that offers interoperable bill payment services to customers through nario are Aadhaar number, fingerprint, and bank account number. a network of agents, enabling multiple payment modes, and pro- Regular banking services (e.g., balance enquiry, cash withdrawal / viding instant confirmation of payment. Initially, BBPS covers regular deposit, funds transfer to other Aadhaar-based accounts) can be payments, for instance at various utilities. The intent is to gradually made through this route. This should help accelerate financial inclu- increase coverage to other types of regular payments, such as munic- sion, as BCs can use microATMs to travel from village to village and ipal taxes, educational fees, and so on. help individuals conduct simple transactions.

Exhibit 181: BBPS Overview

Source: Morgan Stanley Research

MORGAN STANLEY RESEARCH 117 M BLUEPAPER Appendix 2 - Paytm - The Disruptor That Is Becoming a Big Player

Every business needs a disruptor that jolts incumbents Exhibit 182: into changing the business-as-usual nature of going about PAYTM Customer Base (mn) their activities. We believe that in India, Paytm may well 300 be such a disruptor. Paytm (which stands for Pay Through 240 Mobile) started as a mobile wallet company that was initially focussed on prepaid mobile recharging (the bulk of 180 Indian mobile subscribers have prepaid connections). It 120 subsequently started delving into other forms of small payment activities such as movie tickets, bus tickets and 60 Uber. Over time it has become India's largest mobile payments platform. It has also secured a payments bank 0 Feb-17 Feb-16 Feb-15 Aug-17 Aug-16 Nov-16 Aug-15 Nov-15 Nov-14 May-17 May-16 licence from the RBI. May-15 Source: Media including Times of India, Morgan Stanley Research It now has almost 250 million customers and aims to increase this to ~500 million by 2020. It has been offering multiple products to Exhibit 183: these customers, ranging from simple payments to micro insurance. India: Digital Payments (Rs Bn) Digital Payments Values, Rs Bn However, although it was becoming a big mobile wallet company, it 10,000 was not impacting the formal banking payment transactions in a very Credit Cards POS Debit Cards POS Ex Rupay meaningful manner. But this is now changing. If we look at total 8,000 M Wallet Rupay Cards POS mobile wallet transactions in India (not just Paytm), the value has UPI been increasing rapidly in the country, increasing from Rs10bn in 6,000 F2013 to Rs530bn in F2017. In fact the value is now equivalent to 4,000 about ~10% of the total merchant transactions on debit and credit cards. 2,000

0 F2004 F2005 F2006 F2007 F2008 F2009 F2010 F2011 F2012 F2013 F2014 F2015 F2016 F2017 F1Q18

Source: RBI, NPCI, Morgan Stanley Research

118 M BLUEPAPER We expect the total value of mobile wallet transactions to surge Exhibit 184: from F2018 onwards. Historically, Paytm was not big in merchant Digital Payment Options payments. But since the Currency Replacement Programme in Since less than a year ago Since a year or more ago November 2016, it has started focusing on this part of the business 32% 68% in a much more active manner. It has been on an aggressive drive to Debit Card acquire new merchants on its platform and now has ~5 million mer- Credit Card 31% 69% chants accepting payments through Paytm (QR based). This compares with ~2 million merchants who accept credit and debit cards Direct Transfer 24% 76% through the POS network.

PayTM 68% 32% AlphaWise - Our survey also shows the significantly higher awareness and usage of Paytm since the end of 2016 (post CRP); 68% of our BHIM / UPI 94% 6% surveyed customers said they have been using Paytm since the CRP and 77% said they have increased usage since then. Historically Base: Those using these options. Source: AlphaWise, Morgan Stanley Research Paytm was primarily used as a P2P payment option - but clearly usage with merchants is increasing. Exhibit 185: Change in Usage of Digital Payment Options in Recent Months Increase No change Decrease The reasons for the surge in usage of Paytm and other mobile wallets Cash are as follows: 25% 10% 64%

Cheque 58% 19% 23% 1. Increased awareness and ease of use - Paytm has been fairly Debit Card 71% 13% 16% aggressive in ensuring the increase in awareness of its product. It has also been on a merchant acquisition spree which has helped add ~ 5 Credit Card 64% 15% 21% million merchants who can now accept payments through Paytm. Direct Transfer 72% 15% 12%

2. Low transaction cost - One of the key reasons why merchants PayTM 77% 13% 10% have historically avoided digital payments was the cost - including PayTM QR code 63% 17% 19% the cost of POS machines and MDR (Merchant Discount Rate). Paytm BHIM 56% 28% 16% has brought down the cost for merchants significantly. Since payments are based on QR code technology, the incremental cost of the Base: Those using these options. Source: AlphaWise, Morgan Stanley Research payment infrastructure is very low. And Paytm is charging no MDR to the merchants. Paytm does charge an inter-change fee during the cash-out process, though.

Is Paytm losing money on this? Yes. It has invested more than US$600 million in creating and driving adoption of digital payment network on the platform. As the number of customers on the platform increases, Paytm can start offering multiple products to these customers - with almost zero incremental cost - potentially driving a parabolic increase in profitability. Do we know if and when will this happen? No. But given fairly deep access to funding (Alibaba and Softbank are amongst its biggest investors), it can afford to keep losing money for a few years and acquire more customers. Paytm is essentially hoovering up the 500 million people who constitute the under-banked market.

MORGAN STANLEY RESEARCH 119 M BLUEPAPER Disclosure Section

The information and opinions in Morgan Stanley Research were prepared or are disseminated by Morgan Stanley Asia Limited (which accepts the responsibility for its contents) and/or Morgan Stanley Asia (Singapore) Pte. (Registration number 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H), regulated by the Monetary Authority of Singapore (which accepts legal responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research), and/or Morgan Stanley Taiwan Limited and/or Morgan Stanley & Co International plc, Seoul Branch, and/or Morgan Stanley Australia Limited (A.B.N. 67 003 734 576, holder of Australian financial services license No. 233742, which accepts responsibility for its contents), and/or Morgan Stanley Wealth Management Australia Pty Ltd (A.B.N. 19 009 145 555, holder of Australian financial services license No. 240813, which accepts responsibility for its contents), and/or Morgan Stanley India Company Private Limited, regulated by the Securities and Exchange Board of India (“SEBI”) and holder of licenses as a Research Analyst (SEBI Registration No. INH000001105); Stock Broker (BSE Registration No. INB011054237 and NSE Registration No. INB/INF231054231), Merchant Banker (SEBI Registration No. INM000011203), and depository participant with National Securities Depository Limited (SEBI Registration No. IN-DP-NSDL-372-2014) which accepts the responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research, and/or PT. Morgan Stanley Sekuritas Indonesia and their affiliates (collectively, "Morgan Stanley").

For important disclosures, stock price charts and equity rating histories regarding companies that are the subject of this report, please see the Morgan Stanley Research Disclosure Website at www.morganstanley.com/researchdisclosures, or contact your investment representative or Morgan Stanley Research at 1585 Broadway, (Attention: Research Management), New York, NY, 10036 USA.

For valuation methodology and risks associated with any recommendation, rating or price target referenced in this research report, please contact the Client Support Team as follows: US/Canada +1 800 303-2495; Hong Kong +852 2848-5999; Latin America +1 718 754-5444 (U.S.); London +44 (0)20-7425-8169; Singapore +65 6834-6860; Sydney +61 (0)2-9770-1505; Tokyo +81 (0)3-6836-9000. Alternatively you may contact your investment representative or Morgan Stanley Research at 1585 Broadway, (Attention: Research Management), New York, NY 10036 USA. Analyst Certification

The following analysts hereby certify that their views about the companies and their securities discussed in this report are accurately expressed and that they have not received and will not receive direct or indirect compensation in exchange for expressing specific recommendations or views in this report: Anil Agarwal; Grace Chen; Ridham Desai; James E Faucette; Andrea Ferraz, CFA; Vasundhara Govil; Parag Gupta; Rahul Gupta; Subramanian Iyer; Toni Kaplan; Sumeet Kariwala; Nick Lord; Brian Nowak, CFA; Gaurav Rateria; Sheela Rathi; Tetsuro Tsusaka, CFA.

Unless otherwise stated, the individuals listed on the cover page of this report are research analysts. Global Research Conflict Management Policy

Morgan Stanley Research has been published in accordance with our conflict management policy, which is available at www.morganstanley.com/institutional/research/conflictpolicies. Important US Regulatory Disclosures on Subject Companies

The following analyst or strategist (or a household member) owns securities (or related derivatives) in a company that he or she covers or recommends in Morgan Stanley Research: Ridham Desai - Edelweiss Financial Services Ltd.(common or preferred stock), Kotak Mahindra Bank(common or preferred stock); James E Faucette - MasterCard Inc(common or preferred stock); Gaurav Rateria - Multi Commodity Exchange of India Ltd(common or preferred stock); Sheela Rathi - Edelweiss Financial Services Ltd.(common or preferred stock), Kotak Mahindra Bank(common or preferred stock), Multi Commodity Exchange of India Ltd(common or preferred stock).

As of August 31, 2017, Morgan Stanley beneficially owned 1% or more of a class of common equity securities of the following companies covered in Morgan Stanley Research: Amazon.com Inc, Axis Bank, Bank of Baroda, Bharat Financial Inclusion Ltd, Edelweiss Financial Services Ltd., Federal Bank, HDFC, HDFC Bank, ICICI Bank, Indiabulls Housing Finance, IndusInd Bank, LIC Housing Finance Ltd., MakeMyTrip Limited, MasterCard Inc, Multi Commodity Exchange of India Ltd, Naspers, Shriram City Union Finance Ltd, Shriram Transport Finance Co. Ltd., Visa Inc., Yes Bank.

Within the last 12 months, Morgan Stanley managed or co-managed a public offering (or 144A offering) of securities of Indiabulls Housing Finance, Kotak Mahindra Bank, MasterCard Inc, PNB Housing Finance Ltd, Punjab National Bank, Softbank Group, TransUnion.

Within the last 12 months, Morgan Stanley has received compensation for investment banking services from Alibaba Group Holding, IndusInd Bank, Kotak Mahindra Bank, MasterCard Inc, Naspers, PNB Housing Finance Ltd, Punjab National Bank, Softbank Group, TransUnion.

In the next 3 months, Morgan Stanley expects to receive or intends to seek compensation for investment banking services from Alibaba Group Holding, Amazon.com Inc, Axis Bank, Bank of India, DBS Group Holdings, EICHER MOTORS, Eicher Motors Ltd., Federal Bank, HDFC, HDFC Bank, ICICI Bank, ICICI Prudential Life Insurance, IndusInd Bank, Kotak Mahindra Bank, LIC Housing Finance Ltd., Mahindra and Mahindra Financial Services, MakeMyTrip Limited, MasterCard Inc, Multi Commodity Exchange of India Ltd, Naspers, PNB Housing Finance Ltd, Punjab National Bank, RBL Bank Limited, Shriram City Union Finance Ltd, Shriram Transport Finance Co. Ltd., Softbank Group, State Bank of India, TransUnion, Visa Inc., Yes Bank.

Within the last 12 months, Morgan Stanley has received compensation for products and services other than investment banking services from Alibaba Group Holding, Axis Bank, Bank of Baroda, Bank of India, DBS Group Holdings, HDFC, HDFC Bank, ICICI Bank, ICICI Prudential Life Insurance, IDFC Bank, Indiabulls Housing Finance, IndusInd Bank, Kotak Mahindra Bank, MasterCard Inc, Naspers, Softbank Group, State Bank of India, Yes Bank.

Within the last 12 months, Morgan Stanley has provided or is providing investment banking services to, or has an investment banking client relationship with, the following company: Alibaba Group Holding, Amazon.com Inc, Axis Bank, Bank of India, DBS Group Holdings, EICHER MOTORS, Eicher Motors Ltd., Federal Bank, HDFC, HDFC Bank, ICICI Bank, ICICI Prudential Life Insurance, Indiabulls Housing Finance, IndusInd Bank, Kotak Mahindra Bank, LIC Housing Finance Ltd., Mahindra and Mahindra Financial Services, MakeMyTrip Limited, MasterCard Inc, Multi Commodity Exchange of India Ltd, Naspers, PNB Housing Finance Ltd, Punjab National Bank, RBL Bank Limited, Shriram City Union Finance Ltd, Shriram Transport Finance Co. Ltd., Softbank Group, State Bank of India, TransUnion, Visa Inc., Yes Bank.

Within the last 12 months, Morgan Stanley has either provided or is providing non-investment banking, securities-related services to and/or in the past has entered into an agreement to provide services or has a client relationship with the following company: Alibaba Group Holding, Amazon.com Inc, Axis Bank, Bank of Baroda, Bank of India, Canara Bank, DBS Group Holdings, Edelweiss Financial Services Ltd., HDFC, HDFC Bank, ICICI Bank, ICICI Prudential Life Insurance, IDFC Bank, Indiabulls Housing Finance, IndusInd Bank, Kotak Mahindra Bank, MasterCard Inc, Naspers, Punjab National Bank, Softbank Group, State Bank of India, Visa Inc., Yes Bank.

Morgan Stanley & Co. LLC makes a market in the securities of Alibaba Group Holding, Amazon.com Inc, HDFC Bank, ICICI Bank, MakeMyTrip Limited, MasterCard Inc, Softbank Group, TransUnion, Visa Inc.. 120 M BLUEPAPER The equity research analysts or strategists principally responsible for the preparation of Morgan Stanley Research have received compensation based upon various factors, including quality of research, investor client feedback, stock picking, competitive factors, firm revenues and overall investment banking revenues. Equity Research analysts' or strategists' compensation is not linked to investment banking or capital markets transactions performed by Morgan Stanley or the profitability or revenues of particular trading desks.

Morgan Stanley and its affiliates do business that relates to companies/instruments covered in Morgan Stanley Research, including market making, providing liquidity, fund management, commercial banking, extension of credit, investment services and investment banking. Morgan Stanley sells to and buys from customers the securities/instruments of companies covered in Morgan Stanley Research on a principal basis. Morgan Stanley may have a position in the debt of the Company or instruments discussed in this report. Morgan Stanley trades or may trade as principal in the debt securities (or in related derivatives) that are the subject of the debt research report.

Certain disclosures listed above are also for compliance with applicable regulations in non-US jurisdictions. STOCK RATINGS

Morgan Stanley uses a relative rating system using terms such as Overweight, Equal-weight, Not-Rated or Underweight (see definitions below). Morgan Stanley does not assign ratings of Buy, Hold or Sell to the stocks we cover. Overweight, Equal-weight, Not-Rated and Underweight are not the equivalent of buy, hold and sell. Investors should carefully read the definitions of all ratings used in Morgan Stanley Research. In addition, since Morgan Stanley Research contains more complete information concerning the analyst's views, investors should carefully read Morgan Stanley Research, in its entirety, and not infer the contents from the rating alone. In any case, ratings (or research) should not be used or relied upon as investment advice. An investor's decision to buy or sell a stock should depend on individual circumstances (such as the investor's existing holdings) and other considerations. Global Stock Ratings Distribution

(as of August 31, 2017)

The Stock Ratings described below apply to Morgan Stanley's Fundamental Equity Research and do not apply to Debt Research produced by the Firm.

For disclosure purposes only (in accordance with NASD and NYSE requirements), we include the category headings of Buy, Hold, and Sell alongside our ratings of Overweight, Equal-weight, Not-Rated and Underweight. Morgan Stanley does not assign ratings of Buy, Hold or Sell to the stocks we cover. Overweight, Equal-weight, Not-Rated and Underweight are not the equivalent of buy, hold, and sell but represent recommended relative weightings (see definitions below). To satisfy regulatory requirements, we correspond Overweight, our most positive stock rating, with a buy recommendation; we correspond Equal-weight and Not-Rated to hold and Underweight to sell recommendations, respectively.

Other Material Investment Services Clients Coverage Universe Investment Banking Clients (IBC) (MISC) Stock Rating Cate- Count % of Total Count % of Total IBC % of Rating Category Count % of Total Other MISC gory Overweight/Buy 1164 36% 306 41% 26% 555 37% Equal-weight/Hold 1425 44% 349 46% 24% 701 46% Not-Rated/Hold 61 2% 6 1% 10% 10 1% Underweight/Sell 606 19% 91 12% 15% 242 16% Total 3,256 752 1508

Data include common stock and ADRs currently assigned ratings. Investment Banking Clients are companies from whom Morgan Stanley received investment banking compensation in the last 12 months. Due to rounding off of decimals, the percentages provided in the "% of total" column may not add up to exactly 100 percent. Analyst Stock Ratings

Overweight (O). The stock's total return is expected to exceed the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months.

Equal-weight (E). The stock's total return is expected to be in line with the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months.

Not-Rated (NR). Currently the analyst does not have adequate conviction about the stock's total return relative to the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months.

Underweight (U). The stock's total return is expected to be below the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months.

Unless otherwise specified, the time frame for price targets included in Morgan Stanley Research is 12 to 18 months. Analyst Industry Views

Attractive (A): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be attractive vs. the relevant broad market benchmark, as indicated below.

In-Line (I): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be in line with the relevant broad market benchmark, as indicated below.

Cautious (C): The analyst views the performance of his or her industry coverage universe over the next 12-18 months with caution vs. the relevant broad market benchmark, as indicated below.

Benchmarks for each region are as follows: North America - S&P 500; Latin America - relevant MSCI country index or MSCI Latin America Index; Europe - MSCI Europe; Japan - TOPIX; Asia - relevant MSCI country index or MSCI sub-regional index or MSCI AC Asia Pacific ex Japan Index.

MORGAN STANLEY RESEARCH 121 M BLUEPAPER Important Disclosures for Morgan Stanley Smith Barney LLC Customers

Important disclosures regarding the relationship between the companies that are the subject of Morgan Stanley Research and Morgan Stanley Smith Barney LLC or Morgan Stanley or any of their affiliates, are available on the Morgan Stanley Wealth Management disclosure website at www.morganstanley.com/online/researchdisclosures. For Morgan Stanley specific disclosures, you may refer to www.morganstanley.com/researchdisclosures.

Each Morgan Stanley Equity Research report is reviewed and approved on behalf of Morgan Stanley Smith Barney LLC. This review and approval is conducted by the same person who reviews the Equity Research report on behalf of Morgan Stanley. This could create a conflict of interest. Other Important Disclosures

Morgan Stanley & Co. International PLC and its affiliates have a significant financial interest in the debt securities of Alibaba Group Holding, Amazon.com Inc, Axis Bank, Bank of India, HDFC, ICICI Bank, MasterCard Inc, Softbank Group, State Bank of India, Visa Inc..

A member of Research who had or could have had access to the research prior to completion owns securities (or related derivatives) in the State Bank of India. This person is not a research analyst or a member of research analyst's household.

Morgan Stanley Research policy is to update research reports as and when the Research Analyst and Research Management deem appropriate, based on developments with the issuer, the sector, or the market that may have a material impact on the research views or opinions stated therein. In addition, certain Research publications are intended to be updated on a regular periodic basis (weekly/monthly/quarterly/annual) and will ordinarily be updated with that frequency, unless the Research Analyst and Research Management determine that a different publication schedule is appropriate based on current conditions.

Morgan Stanley is not acting as a municipal advisor and the opinions or views contained herein are not intended to be, and do not constitute, advice within the meaning of Section 975 of the Dodd-Frank Wall Street Reform and Consumer Protection Act.

Morgan Stanley produces an equity research product called a "Tactical Idea." Views contained in a "Tactical Idea" on a particular stock may be contrary to the recommendations or views expressed in research on the same stock. This may be the result of differing time horizons, methodologies, market events, or other factors. For all research available on a particular stock, please contact your sales representative or go to Matrix at http://www.morganstanley.com/matrix.

Morgan Stanley Research is provided to our clients through our proprietary research portal on Matrix and also distributed electronically by Morgan Stanley to clients. Certain, but not all, Morgan Stanley Research products are also made available to clients through third-party vendors or redistributed to clients through alternate electronic means as a convenience. For access to all available Morgan Stanley Research, please contact your sales representative or go to Matrix at http://www.morganstanley.com/matrix.

Any access and/or use of Morgan Stanley Research is subject to Morgan Stanley's Terms of Use (http://www.morganstanley.com/terms.html). By accessing and/or using Morgan Stanley Research, you are indicating that you have read and agree to be bound by our Terms of Use (http://www.morganstanley.com/terms.html). In addition you consent to Morgan Stanley processing your personal data and using cookies in accordance with our Privacy Policy and our Global Cookies Policy (http://www.morganstanley.com/privacy_pledge.html), including for the purposes of setting your preferences and to collect readership data so that we can deliver better and more personalized service and products to you. To find out more information about how Morgan Stanley processes personal data, how we use cookies and how to reject cookies see our Privacy Policy and our Global Cookies Policy (http://www.morganstanley.com/privacy_pledge.html).

If you do not agree to our Terms of Use and/or if you do not wish to provide your consent to Morgan Stanley processing your personal data or using cookies please do not access our research.

Morgan Stanley Research does not provide individually tailored investment advice. Morgan Stanley Research has been prepared without regard to the circumstances and objectives of those who receive it. Morgan Stanley recommends that investors independently evaluate particular investments and strategies, and encourages investors to seek the advice of a financial adviser. The appropriateness of an investment or strategy will depend on an investor's circumstances and objectives. The securities, instruments, or strategies discussed in Morgan Stanley Research may not be suitable for all investors, and certain investors may not be eligible to purchase or participate in some or all of them. Morgan Stanley Research is not an offer to buy or sell or the solicitation of an offer to buy or sell any security/instrument or to participate in any particular trading strategy. The value of and income from your investments may vary because of changes in interest rates, foreign exchange rates, default rates, prepayment rates, securities/instruments prices, market indexes, operational or financial conditions of companies or other factors. There may be time limitations on the exercise of options or other rights in securities/instruments transactions. Past performance is not necessarily a guide to future performance. Estimates of future performance are based on assumptions that may not be realized. If provided, and unless otherwise stated, the closing price on the cover page is that of the primary exchange for the subject company's securities/instruments.

The fixed income research analysts, strategists or economists principally responsible for the preparation of Morgan Stanley Research have received compensation based upon various factors, including quality, accuracy and value of research, firm profitability or revenues (which include fixed income trading and capital markets profitability or revenues), client feedback and competitive factors. Fixed Income Research analysts', strategists' or economists' compensation is not linked to investment banking or capital markets transactions performed by Morgan Stanley or the profitability or revenues of particular trading desks.

The "Important US Regulatory Disclosures on Subject Companies" section in Morgan Stanley Research lists all companies mentioned where Morgan Stanley owns 1% or more of a class of common equity securities of the companies. For all other companies mentioned in Morgan Stanley Research, Morgan Stanley may have an investment of less than 1% in securities/instruments or derivatives of securities/instruments of companies and may trade them in ways different from those discussed in Morgan Stanley Research. Employees of Morgan Stanley not involved in the preparation of Morgan Stanley Research may have investments in securities/instruments or derivatives of securities/instruments of companies mentioned and may trade them in ways different from those discussed in Morgan Stanley Research. Derivatives may be issued by Morgan Stanley or associated persons.

With the exception of information regarding Morgan Stanley, Morgan Stanley Research is based on public information. Morgan Stanley makes every effort to use reliable, comprehensive information, but we make no representation that it is accurate or complete. We have no obligation to tell you when opinions or information in Morgan Stanley Research change apart from when we intend to discontinue equity research coverage of a subject company. Facts and views presented in Morgan Stanley Research have not been reviewed by, and may not reflect information known to, professionals in other Morgan Stanley business areas, including investment banking personnel.

Morgan Stanley Research personnel may participate in company events such as site visits and are generally prohibited from accepting payment by the company of associated expenses unless pre-approved by authorized members of Research management.

Morgan Stanley may make investment decisions that are inconsistent with the recommendations or views in this report.

122 M BLUEPAPER To our readers based in Taiwan or trading in Taiwan securities/instruments: Information on securities/instruments that trade in Taiwan is distributed by Morgan Stanley Taiwan Limited ("MSTL"). Such information is for your reference only. The reader should independently evaluate the investment risks and is solely responsible for their investment decisions. Morgan Stanley Research may not be distributed to the public media or quoted or used by the public media without the express written consent of Morgan Stanley. Any non-customer reader within the scope of Article 7-1 of the Taiwan Stock Exchange Recommendation Regulations accessing and/or receiving Morgan Stanley Research is not permitted to provide Morgan Stanley Research to any third party (including but not limited to related parties, affiliated companies and any other third parties) or engage in any activities regarding Morgan Stanley Research which may create or give the appearance of creating a conflict of interest. Information on securities/instruments that do not trade in Taiwan is for informational purposes only and is not to be construed as a recommendation or a solicitation to trade in such securities/instruments. MSTL may not execute transactions for clients in these securities/instruments.

Certain information in Morgan Stanley Research was sourced by employees of the Shanghai Representative Office of Morgan Stanley Asia Limited for the use of Morgan Stanley Asia Limited.

Morgan Stanley is not incorporated under PRC law and the research in relation to this report is conducted outside the PRC. Morgan Stanley Research does not constitute an offer to sell or the solicitation of an offer to buy any securities in the PRC. PRC investors shall have the relevant qualifications to invest in such securities and shall be responsible for obtaining all relevant approvals, licenses, verifications and/or registrations from the relevant governmental authorities themselves. Neither this report nor any part of it is intended as, or shall constitute, provision of any consultancy or advisory service of securities investment as defined under PRC law. Such information is provided for your reference only.

Morgan Stanley Research is disseminated in Brazil by Morgan Stanley C.T.V.M. S.A.; in Mexico by Morgan Stanley México, Casa de Bolsa, S.A. de C.V which is regulated by Comision Nacional Bancaria y de Valores. Paseo de los Tamarindos 90, Torre 1, Col. Bosques de las Lomas Floor 29, 05120 Mexico City; in Japan by Morgan Stanley MUFG Securities Co., Ltd. and, for Commodities related research reports only, Morgan Stanley Capital Group Japan Co., Ltd; in Hong Kong by Morgan Stanley Asia Limited (which accepts responsibility for its contents) and by Morgan Stanley Asia International Limited, Hong Kong Branch; in Singapore by Morgan Stanley Asia (Singapore) Pte. (Registration number 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H), regulated by the Monetary Authority of Singapore (which accepts legal responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research) and by Morgan Stanley Asia International Limited, Singapore Branch (Registration number T11FC0207F); in Australia to "wholesale clients" within the meaning of the Australian Corporations Act by Morgan Stanley Australia Limited A.B.N. 67 003 734 576, holder of Australian financial services license No. 233742, which accepts responsibility for its contents; in Australia to "wholesale clients" and "retail clients" within the meaning of the Australian Corporations Act by Morgan Stanley Wealth Management Australia Pty Ltd (A.B.N. 19 009 145 555, holder of Australian financial services license No. 240813, which accepts responsibility for its contents; in Korea by Morgan Stanley & Co International plc, Seoul Branch; in India by Morgan Stanley India Company Private Limited; in Indonesia by PT. Morgan Stanley Sekuritas Indonesia; in Canada by Morgan Stanley Canada Limited, which has approved of and takes responsibility for its contents in Canada; in Germany by Morgan Stanley Bank AG, Frankfurt am Main and Morgan Stanley Private Wealth Management Limited, Niederlas- sung Deutschland, regulated by Bundesanstalt fuer Finanzdienstleistungsaufsicht (BaFin); in Spain by Morgan Stanley, S.V., S.A., a Morgan Stanley group company, which is supervised by the Spanish Securities Markets Commission (CNMV) and states that Morgan Stanley Research has been written and distributed in accordance with the rules of conduct applicable to financial research as established under Spanish regulations; in the US by Morgan Stanley & Co. LLC, which accepts responsibility for its contents. Morgan Stanley & Co. International plc, authorized by the Prudential Regulatory Authority and regulated by the Financial Conduct Authority and the Prudential Regulatory Authority, disseminates in the UK research that it has prepared, and approves solely for the purposes of section 21 of the Financial Services and Markets Act 2000, research which has been prepared by any of its affiliates. RMB Morgan Stanley Proprietary Limited is a member of the JSE Limited and regulated by the Financial Services Board in South Africa. RMB Morgan Stanley Proprietary Limited is a joint venture owned equally by Morgan Stanley Interna- tional Holdings Inc. and RMB Investment Advisory (Proprietary) Limited, which is wholly owned by FirstRand Limited. The information in Morgan Stanley Research is being disseminated by Morgan Stanley Saudi Arabia, regulated by the Capital Market Authority in the Kingdom of Saudi Arabia , and is directed at Sophisticated investors only.

The information in Morgan Stanley Research is being communicated by Morgan Stanley & Co. International plc (DIFC Branch), regulated by the Dubai Financial Services Authority (the DFSA), and is directed at Professional Clients only, as defined by the DFSA. The financial products or financial services to which this research relates will only be made available to a customer who we are satisfied meets the regulatory criteria to be a Professional Client.

The information in Morgan Stanley Research is being communicated by Morgan Stanley & Co. International plc (QFC Branch), regulated by the Qatar Financial Centre Regulatory Authority (the QFCRA), and is directed at business customers and market counterparties only and is not intended for Retail Customers as defined by the QFCRA.

As required by the Capital Markets Board of Turkey, investment information, comments and recommendations stated here, are not within the scope of investment advisory activity. Investment advisory service is provided exclusively to persons based on their risk and income preferences by the authorized firms. Comments and recommendations stated here are general in nature. These opinions may not fit to your financial status, risk and return preferences. For this reason, to make an investment decision by relying solely to this information stated here may not bring about outcomes that fit your expectations.

The following companies do business in countries which are generally subject to comprehensive sanctions programs administered or enforced by the U.S. Department of the Treasury's Office of Foreign Assets Control ("OFAC") and by other countries and multi-national bodies: MasterCard Inc.

The trademarks and service marks contained in Morgan Stanley Research are the property of their respective owners. Third-party data providers make no warranties or representations relating to the accuracy, completeness, or timeliness of the data they provide and shall not have liability for any damages relating to such data. The Global Industry Classification Standard (GICS) was developed by and is the exclusive property of MSCI and S&P.

Morgan Stanley Research, or any portion thereof may not be reprinted, sold or redistributed without the written consent of Morgan Stanley.

The Americas Europe Japan Asia/Pacific 1585 Broadway 20 Bank Street, Canary Wharf 1-9-7 Otemachi, Chiyoda-ku 1 Austin Road West New York, NY 10036-8293 London E14 4AD Tokyo 100-8104 Kowloon United States United Kingdom Japan Hong Kong Tel: +1 (1) 212 761 4000 Tel: +44 (0) 20 7 425 8000 Tel: +81 (0) 3 6836 5000 Tel: +852 2848 5200