Risk & Regulatory Academy 2020 DAY 2 BASEL IV

© 2020 Deloitte Deloitte Central Europe & Regulatory Academy 1 Agenda

BASEL IV - presentation by Michael Cluse & Ewa Winiarska

• How is COVID affecting the implementation schedule of BASEL?

• How is BASEL IV affecting risk measurement and model development for the ?

• How is the comparison evolving between standardized and IRB approach?

• The role of RegTech

PANEL DISCUSSION

Michael Cluse, Przemyslaw Szczygielski, Hervé Phaure, Dimitrios Goranitis

Q&A SESSION – Please type your questions into the chat box

© 2020 Deloitte 2 Our subject matter experts

Michael Cluse Ewa Winiarska Herve Phaure Przemyslaw Szczygielski Risk Advisory Director Risk & Regulatory Senior Manager Leader Risk & Regulatory Leader at Deloitte Germany at Deloitte Poland at Deloitte France at Deloitte Poland

© 2020 Deloitte 3 Intro Basel IV Basel III final, Basel 3.1 or Basel IV? Introduction to Prudential Supervision Basel development over the years

Basel Process

Basel I (1988) • basis principal credit risk 8% weight

Basel II (2004 – 2008) • Credit risk: possibility for internal ratings based models (IRB) •

Basel III (2010 – 2019) • Stricter regulatory capital requirements (“loss absorption”) • Leverage ratio • Additional liquidity requirements (LCR/NSFR)

Basel IV (2023 – 2028; “finalisation of Basel III”) • Revision of Standardised approaches • Limitations for internal model usage • Output floor • Additional leverage ratio requirements for G-SIBs

© 2020 Deloitte 5 Regulatory Roadmap Implementation timeline

Due to the effects of the COVID-19 pandemic the Basel Comittee decided to delay the implementation of the final Basel III rules by one year

2019 2020 2021 2022 2023 … 2028 Covid 19 Response Credit Risk (CRSA) Phase of analysis and preparation January 1st

Credit Risk (IRBA) Phase of analysis and preparation January 1st

CVA Risk Phase of analysis and preparation January 1st

Counterparty Credit Risk Implementation

Securitisation January 1st

Market Risk Reporting requirements January 1st

OpRisk Phase of analysis and preparation January 1st

Capital Floors Phase of analysis and preparation Phase-in: annual increase each January 1st

Floor: 50 % Floor: 72,5 %

Draft CRR III? BCBS EU

© 2020 Deloitte 6 Brief Summary Basel IV

© 2020 Deloitte 7 CRSA – Essential Changes Increasing risk sensitivity

Reducing dependencies • Usage of external ratings for the purpose of weighing risk at national discretion on external ratings • Mandatory internal due diligence, potentially increasing risk weight

• Introducing new calculation practices for the purpose of deriving risk weights More differentiated methods to derive risk • LTV (Loan to Value) for real estate financing weights • Regulatory ratios for unrated institutions (SCRA) and elimination of home country rating

Recalibration of risk • Adjusted risk weights and modified conversion factors weights and CCFs (e.g. unconditionally cancellable credit lines)

• New exposure classes put stronger focus on the type of borrower and the purpose of financing Revised definition of • exposure classes Higher differentiation within certain exposure classes (i.e. specialised lending)

• The output floor and the related transfer of exposures from IRBA to CRSA highlights the importance of the increased risk sensitivities and the revised definition of exposure classes • CRSA as input for the output floor has to be implement also by using the IRBA www.deloitte.com/fluent

© 2020 Deloitte 8 Basel III reforms – New Standardised risk weights Revised Exposure Classes and Risk Weights in the Final Basel paper

Current Risk Weights under CRR Revised Risk Weights by BCBS

Institutions 20% – 150% Banks 20% – 150% Multilateral Development Banks 0% – 150% Multilateral Development Banks 0% – 150% Corporates – Senior Debt 20% – 150% Corporates 20% – 150% Corporates – Specialised Lending 20% − 150% 1.5 multiplier Equity Exposures 100% – 250% Subordinated Debt, Equity and Other 100% – 400%* where the Regulatory Retail 75%** lending Retail 75% currency differs Other Retail 100% from the currency of the Secured by mortgages on Residential 35% Residential*** 20% – 150% borrower’s immovable property Commercial 50% Secured by real estate Commercial*** RWoC**** source of income Exposures in Default 100% – 150% ADC***** 100% – 150% Other Items 0% – 250% Defaulted exposures 100% – 150% Other Assets 100%****** Current Conversion Factors under CRR Revised Conversion Factors by BCBS Off-balance sheet 0% – 100% Off-balance sheet 10% – 100%

*400% introduced for the speculative part ** 45% for transactors (obligors in relation to facilities such as credit cards and charge cards where the balance has been repaid in full at each scheduled repayment date for the previous 12 months) ***Distinction for materially & non-materially dependent cash flows, see specification ****Risk weight of counterparty, check specification slide *****Land Acquisition, Development and Construction ******Two minor exceptions © 2020 Deloitte 9 Note: Exposures to Central Governments and Central Banks and Public Sector Entities dealt with in the forthcoming, more comprehensive review on sovereign risk. Risk weights for real estate exposure The determination of risk weight for real estate exposure is based on counterparty and purpose of real estate

Eligibility criteria for residential and commercial real estate Risk weight table for residential real estate exposures Repayment is not Repayment materially Additional eligibility criterion – the property securing the exposure must be fully materially dependent dependent on cash flows completed (with the exception of forest and agricultural land). Local supervisor may on cash flows generated generated by property eliminate this criterion for loans to individuals that are secured by residential by property property under construction provided that the property is a one-to-four family residential housing unit that will be the primary residence of the borrower or where LTV ≤ 50% 20% 30% Risk the sovereign or PSEs involved have the legal powers and ability to ensure that the 50% < LTV ≤ 60% 25% 35% property under construction will be finished. weight 60% < LTV ≤ 80% 30% 45% Risk weights 80% < LTV ≤ 90% 40% 60% Risk weights depend on the type of property (residential/commercial), LtV value and 90% < LTV ≤ 100% 50% 75% whether repayment is materially dependent on cash flows generated by the property. LTV > 100% 70% 105% Alternative method Risk weight table for commercial real estate exposures Local supervisor may choose an alternative method for risk weights assignment – method similar to the current approach but with different multiplier for determining Repayment is not Repayment materially fully and completely secured part of exposure and different risk weight for this part materially dependent on dependent on cash flows of exposure (multiplier of 55% and risk weight of 20% for residential real estate and cash flows generated by generated by property property 60% for commercial real estate). Risk Risk weight of 150% weight LTV ≤ 60% Min(60%;RWoC) 70% 60% < LTV ≤ 80% 90% Risk weight of 150% assigned to exposures secured on residential or commercial real LTV > 80% RWoC estate when the prospects for servicing the loan materially depend on cash flows 110% generated by this property and the property does not meet eligibility criteria. For the purpose of LtV calculation, the value of the property shall be maintained at the value measured at origination unless national supervisors elect to require banks to revise the property value downwards. © 2020 Deloitte Modifications made to the property that unequivocally increase its value could also be considered in the 10 LTV. IRBA – Essential Changes Extensive changes due to new regulations for the IRBA procedure

• Migration from the advanced approach (A-IRBA) to the foundation approach (F-IRBA): Banks and Large Corporates (Revenue > €500 Mio.) Reduced scope for IRBA • Discontinuation of the F-IRBA: Equity exposures (only CRSA permitted) • A-IRBA and slotting approach for specialised lending remain

Restrictions in the estimation of • PD: Increase of the input-floors for all risk exposure classes IRBA parameters • LGD and EAD: Introduction of the input parameter floors (input floors)

General: • Discontinuation of the Double-Default-Approach as a credit risk mitigation technique Credit risk mitigation techniques F-IRBA: • Generally lower LGDs under the F-IRB for collateralised and not collateralised positions are possible • New calculation methods for the LGD determination for (partly) collateralised exposures • Introduction of haircuts instead of current minimum collateralisation ratio

Scaling factor • Elimination of the scaling factor, which currently equals 1.06

© 2020 Deloitte 11 Output Floor Background, technical, implementation, issues How does the Output Floor work? Capital Floor Mechanism An aggregated Capital Floor of 72.5% limits the capital savings through the use of Internal Models

Implementation Final Floor Regulation Credit risk CCR Introduction for Pillar I: Minimum requirements • Comparison of total RWA CVA based on new Full implementation regulations under consideration of Market risk approved Internal Models with the respective OpRisk Standardized Approaches  ( implicit compensation) 2023 2024 2025 2026 2027 2028

• Limitation through a Every January the 1st Capital Floor of 72.5% 72.5% based on total RWA using Capital Floor 70% the respective Standardized Approaches 65% 60% • Transitional arrangements 55% to mitigate the effects 50%

Transitional periods should give institutions time to adjust their capital base or portfolios to the new conditions

© 2020 Deloitte 14 Capital Floor Mechanism An aggregated Capital Floor of 72.5% limits the capital savings through the use of Internal Models

Implementation Final Floor Regulation 100 Floor: 72.5% • Comparison of total RWA Additional Capital under consideration of 50 Requirements approved Internal Models with the respective Standardized Approaches ( implicit compensation) 0

• Limitation through a RWA according to all Standardized RWA according to approved Internal Capital Floor of 72.5% Approaches Models based on total RWA using = Sum of RWA of = Sum of RWA of the respective Standardized . Credit risk Standardized Approach . The regulatory approved Internal Approaches (CRSA) Models . SA-CCR • Transitional arrangements . The Standardized Approaches for risk . BA-CVA/SA-CVA/100% RWA types that are not treated according to to mitigate the effects Counterparty risk* Internal Models . SEC-ERBA/SEC-SA/1250%* . Standardised Approach for Market risk . Standardised measurement approach for OpRisk

© 2020 Deloitte 15 Calculation example: Effect of the Capital Floor The effect of the Capital Floor Example 1

Assumptions 1st case Scenario

Credit risk: IRBA Credit risk: Unchanged Market risk: Standardized Approach Market risk: Increase Operational risk: Standardized Approach Operational risk: Unchanged

Impact Standardized Approach Internal Model Credit Credit Before Floor Eff. Δ Market OpRisk Total Delta Floor Market OpRisk Σ RWA CRSA IRBA Floor Effect RWA

t0 800 150 50 1000 725 t0 500 150 50 700 25 725

t1 800 300 50 1150 150 833.8 t1 500 300 50 850 0 850 125 72.5%

• In the initial situation (t0) the Floor regulation leads to additional/higher total capital requirements (+25) • There is no more effect from the Floor after the increase in Market risk (t1)

Floor allocation problem

• Which business area needs to bear the costs for the Floor effect in scenario t0? • To which business area will the difference between the increased market risk requirement and total be attributable?

© 2020 Deloitte 17 The effect of the Capital Floor Example 2

Assumptions 2nd case Scenario

Credit risk: IRBA Credit risk: Unchanged Market risk: IMM Market risk: Increase Operational risk: Standardized Approach Operational risk: Unchanged

Impact Standardized Approach Internal Model Credit Credit Before Floor Eff. Δ Market OpRisk Total Delta Floor Market OpRisk Σ RWA CRSA IRBA Floor Effect RWA

t0 800 150 50 1000 725 t0 500 100 50 650 75 725

t1 800 300 50 1150 150 833.8 t1 500 200 50 750 84 833.8 109 72.5%

• In this scenario, the total capital change of 108,8 is greater than the change in the Market IMM Internal Model (100) as there is a Market risk difference between the SA and IMM approaches. Nevertheless, the additional capital requirement is lower than Scenario 1 without IMM. The Floor acts as a "buffer"

Problem

• Which business area needs to bear the costs for the Floor effect in scenario t0? • How is the Floor effect distributed between business units?  A proportional distribution of the Floor effect is not causative  Methodological approaches to risk capital allocation and overall bank management need to be reviewed and adjusted

© 2020 Deloitte 18 Calculation example: Distribution of Floor Effects The effect of the Capital Floor Distribution of Floor effects - Calculation example (1/2)

Calculation example I A. Standardized Approach B. Internal Model Credit Market Credit Market Before Floor OpRisk Total Delta Floor OpRisk Σ RWA Δ RWA CRSA SA IRBA IMM Floor Effect

t0 800 150 50 1000 725 t0 500 100 50 650 75 725

t1 800 300 50 1150 150 833.8 t1 500 200 50 750 83.8 833.8 108.8 72.5%

Determination of a standard distribution for the Int. Model

Standardised distribution t0 Allocation ∆ Floor in t1 Credit Market Op Before Market ΔFloor Δ RWA Credit IRBA OpRisk Total IRBA IMM Risk Floor IMM ∆ B(t1) – B(t0) 0 100 0 100 8.8 108.8 ∆ A (t0) - B (t0) 300 50 0 350 Distribution I 0 8.8 0 8.8 Floor Effect weight* 86% 14% 0% 100% Distribution II 7.5 1.3 0 8.8 Weighted Floor** 564.3 110.7 50 725 Distribution before Weighted Floor Distribution after Weighted Floor

Open Q&A:

• Is the distribution over a weighting causative?

The IRBA Credit risk is unchanged in t1, but the credit risk still receives a part of the Floor Effect • Assumption in this calculation: There is an agreed distribution of capital requirements on business units made by management

* Floor Effect weight is weighted as the Risk type´s ´∆ A (t0) - B (t0)´ over the Total * * Weighted Floor t0 = Internal Model + (Floor Effect weight * Floor Effect) © 2020 Deloitte 20 The effect of the Capital Floor Distribution of Floor effects – Calculation example (2/2)

Calculation example I

The allocation of RWA is affected by the distribution assumptions management make, as seen in Calculation example (1/2) previously, as well as by the order

in which the calculations are made, as seen here by comparing the distribution outcomes from comparing t1 to t0, t2 to t1 and lastly t2 to t0

Calculation example II A. Standardized Approach B. Internal Model Credit Credit Before Floor Market OpRisk Total Delta Floor Market OpRisk Σ RWA Δ RWA CRSA IRBA Floor Effect

t0 800 150 50 1000 725 t0 500 150 50 700 25 725 t1 800 170 50 1020 20 739.5 t1 600 170 50 820 0 820 95 t 600 180 50 830 0 830 105 t2 800 180 50 1030 30 746.8 2 72.5%

Standardised distribution t0 Allocation ∆ Floor in t1

Credit Credit Op Before Market OpRisk Total Market ΔFloor Δ RWA IRBA IRBA Risk Floor ∆ B(t ) – B(t ) 100 20 0 120 -25 95 ∆ A (t ) - B (t ) 300 0 0 300 1 0 0 0 Distr. I (t ) -20.8 -4.2 0 -25 Floor Effect weight 100% 0% 0% 100% 1 Distr. II (t ) -25 0 0 -25 Weighted Floor* 525 150 50 725 1

∆ B(t ) – B(t ) 0 10 0 10 0 10 •! Allocation vs. t not appropriate 2 1 1 Distr. I (t ) 0 0 0 0 In t a change has occurred with respect to the t result. Allocation depends 2 1 0 Distr. II (t ) 0 0 0 0 on the calculation order, results are not necessarily additive 2 •1 Causation? ∆ B(t ) – B(t ) 100 30 0 130 -25 105 The Market risk area does not have an internal model 2 0 Distr. I (t2) -19.2 -5.8 0 -25 •2 Causation? Distr. II (t ) -25 0 0 -25 The capital requirements in the Market risk area are also rising 2 Distribution before Weighted Floor * Distribution after Weighted Floor Weighted Floor t0 = Internal Model + (Floor Effect weight * Floor Effect) © 2020 Deloitte 21 The effect of the Capital Floor Preliminary conclusion

Consideration in risk control, RoE, pricing etc. The changed capital allocation should be considered at all relevant levels – up to the individual transaction at pre- and post-calculation

The order of calculation steps is important Intermediate steps can affect the result. Therefore, a fixed reference point for the calculations must be agreed

Adjustment of total capital allocation The Floor changes the total capital backing. Therefore, agreed allocation criteria are to be worked out. The allocation of additional capital requires a cross-divisional vote

The Floor level effect is difficult to anticipate The aggregated Floor must be determined across risk types. However, changes in the overall RWA on CRSA portfolios can also have an impact on the benefits of IRBA models

© 2020 Deloitte 22 Capital Floor – Strategic Implications Capital locked by the Floor must be allocated to business units – requiring a revamp of the overall allocation processes

Allocation of Floor RWA

Floor RWABusiness Unit = Floor RWA * Floor Distribution Factor (FDF)

Whole … Additional Floor sale RWA …

Retail OpRisk Credit

Market Real Estate

Allocation of (additional) Floor RWA to business units is a strategic – and political – decision

• The Floor Distribution Factor determines which share various business units have to absorb, increasing their ROE capital base • The decisions regarding the Floor allocation must be taken by senior management due to the impact on profitability • Allocation should be consistent to both business and risk strategy • Business units only using Standardized Approaches may for example be credited for providing headroom for business units using Internal Models

© 2020 Deloitte 23 Capital Floor Distribution Examples Standardized Approach is the (only) relevant RWA driver once the Floor becomes applicable/binding

How does RWA change if/when the Floor is a binding parameter?

Change of marginal RWA “under Floor“

RW RWA Floor- Delta EAD CRSA IRBA CRSA IRBA Floor Δ RWA Check Impact eff Floor (Δ RWA Portfolio 1.000.000,00 100% 65% 1.000.000,00 650.000,00 725.000,00 75.000,00 eff – Δ Floor ) Case 1 1.000,00 100% 50% 1.000,00 500,00 Sum 1.001.000,00 1.001.000,00 650.500,00 725.725,00 75.225,00 725,00 225,00 500,00 Case 2 1.000,00 100% 150% 1.000,00 1.500,00 Sum 1.001.000,00 1.001.000,00 651.500,00 725.725,00 74.225,00 725,00 - 775,00 1.500,00 Case 3 - 1.000,00 100% 100% - 1.000,00 - 1.000,00 Sum 999.000,00 999.000,00 649.000,00 724.275,00 75.275,00 - 725,00 275,00 - 1.000,00 Case 4 - 1.000,00 150% 100% - 1.500,00 - 1.000,00 Sum 999.000,00 998.500,00 649.000,00 723.912,50 74.912,50 - 1.087,50 - 87,50 - 1.000,00

Key findings

• As long as the Floor is binding the effective RWA change is EAD*RWCRSA*Floorfactor • Only the Standardized Approach is relevant for controlling and management. IRBA serves as proof that a bank is eligible for a “rebate” equalling (1-Floorfactor)

on RWCRSA

• Check: ΔRWAIRBA = Δ RWAeff – Δ Floor or Δ RWAeff = ΔRWAIRBA – Δ Floor; Differences between RWCRSA and RWIRBA will be absorbed by the Floor as long as the Floor is applicable

Conclusion: Since Δ RWAeff is caused by the change in CRSA, regulatory capital management should be based on the standardised approach in the future

© 2020 Deloitte 24 Summary: The Capital Floor – More than „just a comparison“ Anticipating the Capital Floor The various strategic implications should be assessed early to adjust bank management processes

Dec. 2017 Approx. Q1 2021 From Jan. 2023 Jan. 2028 Final Basel III Reforms § Draft CRR III/ Planned start date End of transition phase (BCBS) CRD VI (EU) Capital Floor: 50 % Capital Floor: 72.5 %

2018 2020 2021 2022 2023 2028

Consideration of strategic implications

Allocation of Floor RWA Pricing increased cost of capital A causative and fully accepted allocation method To which extent can additional cost of capital be for additional cost of capital requires an included in pricing of products? Important question, assessment and adjustment of the capital especially for products with long maturity – to be allocation management. solved based on market potential, competition and client base.

Benefit of Internal Models Portfolio optimisation The introduction of the Capital Floor impacts the In order to manage portfolios, taking into account cost-benefit ratio as well as the chances and Floor effects (eg. CRSA vs. IRBA), requires transaction from using internal models instead of standardised specific data (and often modifications to the data approaches. The benefit of internal models should be pool). This information should be available (well) reassessed. before Basel IV enters into force, thus it should be implemented in time.

© 2020 Deloitte 26 Future of the IRBA – Is it still an attractive alternative? Choice of Approaches per Exposure Class Under Basel IV, not all approaches (CRSA, Foundation IRBA and Advanced IRBA) will remain available for all exposure classes

Most Important Exposure Classes CRR Basel IV

Banks and Financial Institutions CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA

Retail CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA

Corporates (revenue ≤ 500 Mio. EUR) CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA Corporates CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA (cons. revenue > 500 Mio. EUR) Corporates – Specialized Lending CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA

Equity CRSA F-IRBA A-IRBA CRSA F-IRBA A-IRBA

available prohibited

• With the exception of equity exposures, the foundation IRBA (F-IRBA) remains available for all relevant exposure classes • The advanced IRBA will be limited to those exposure classes where sufficient loss history allows the development of appropriate LGD rating systems

© 2020 Deloitte 28 Costs and Benefits of an IRBA Application – Overview The impact on RWA is not the only topic to be considered when evaluating a potential IRBA application

Potential IRBA Benefit Potential IRBA Costs

More risk sensitive RWA calculation for pillar I IRBA approval • Because of the more risk sensitive approach and better collateral recognition, RWA • Budgets required for achieving IRBA compliance can often be reduced by 20 to 40% by using the IRBA rather than CRSA o Data (Retail, Corporates, Real Estate) o Models • The removal of the 1.06 scaling factor and reduced LGDs in FIRB increase the o Processes (regulatory reporting and risk reporting) attractiveness of the IRBA (while the output floor limits the overall benefit of o IT systems internal approaches to 27.5% of SA-RWA) o Reporting o Improved and integration of pillars I and II Use test • Increased consistency for capital management an planning by using internal risk • Costs of regulatory approval process including on site inspection parameters in RWA calculation Running the IRBA models IRBA eligibility of existing rating models • Recurring costs to maintain IRBA compliance, including o Independent model validation and review (esp. internal audit) • Most banks have been using rating models for pillar II since many years, so several o Continuous model improvement and development incl. regulatory approval IRBA model requirements can be met without developing new models process for significant model changes • In some cases, very similar models are IRBA approved for use in other banks o Higher adherence to regulatory requirements for the entire credit process, (group or pool models) (most notable in front office)

The IRBA approach should not solely be understood as a means to save on regulatory capital requirements. It comes along with a more risk sensitive capital management and planning process that is closer aligned to pillar II. An individual assessment of the potential benefits is recommended for any bank considering a potential IRBA application to include the specific cost drivers like data availability and expected RWA reduction per portfolio/exposure class.

© 2020 Deloitte 29 Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited (“DTTL”), its global network of member firms, and their related entities (collectively, the “Deloitte organization”). DTTL (also referred to as “Deloitte Global”) and each of its member firms and related entities are legally separate and independent entities, which cannot obligate or bind each other in respect of third parties. DTTL and each DTTL member firm and related entity is liable only for its own acts and omissions, and not those of each other. DTTL does not provide services to clients. Please see www.deloitte.com/about to learn more. Deloitte Central Europe is a regional organization of entities organized under the umbrella of Deloitte Central Europe Holdings Limited, the member firm in Central Europe of Deloitte Touche Tohmatsu Limited. Services are provided by the subsidiaries and affiliates of, and firms associated with Deloitte Central Europe Holdings Limited, which are separate and independent legal entities. The subsidiaries and affiliates of, and firms associated with Deloitte Central Europe Holdings Limited are among the region’s leading professional services firms, providing services through nearly 7,000 people in 44 offices in 18 countries. This communication contains general information only, and none of Deloitte Touche Tohmatsu Limited („DTTL“), its global network of member firms or their related entities (collectively, the “Deloitte organization”) is, by means of this communication, rendering professional advice or services. Before making any decision or taking any action that may affect your finances or your business, you should consult a qualified professional adviser. No representations, warranties or undertakings (express or implied) are given as to the accuracy or completeness of the information in this communication, and none of DTTL, its member firms, related entities, employees or agents shall be liable or responsible for any loss or damage whatsoever arising directly or indirectly in connection with any person relying on this communication. DTTL and each of its member firms, and their related entities, are legally separate and independent entities. © 2020. For information, contact Deloitte Central Europe Risk & Regulatory Academy