Analyst comments related to COVID-19 in this file include:

Monday, 6 July 2020 Mail launches a Hybrid Cloud solution based on Web Services – Kirill Panarin US election, the virus gets less important as it spreads faster, SA and its IMF deal, Ethiopia, Nigeria currency – Charles Robertson

Mail launches a Hybrid Cloud solution based on

Ticker: MAIL LI Rating: BUY Target price: $25.5 Current price: $24.4

Mail.ru Cloud Solutions (MCS) has launched a solution for automating Hybrid Cloud deployments based on Amazon Web Services’ (AWS) technology. This solution will automate Kubernetes cluster management in MCS and AWS using Kubernetes Federation, which allows clients to deploy clusters simultaneously on both. According to Kommersant, revenue will be shared between partners according to the consumption of solutions of each platform. In other words, Mail won’t get any commission for promoting AWS in , but the partnership should make its own products more appealing and drive the customer base expansion.

Among the main reasons for cooperation between MCS and AWS are geographical and legal issues. AWS is the largest cloud infrastructure company in the world, however its services offered in Russia are limited due to the personal data regulation and the fact that their nearest data centre is located in Germany. The partnership with MCS should result in a solution for running apps in a Hybrid Cloud environment of MCS and AWS in compliance with Russian legislation.

The Russian cloud market is nascent, with penetration of ~10% vs 60% in some international markets. It amounted to RUB86bn in 2019 and could grow to ~200bn by 2023, according to iKS-Consulting. Mail and Amazon are both small, with insignificant single-digit market share each (Mail generated less than RUB500mn cloud revenue in 1Q20, on our estimates), and the partnership should strengthen their competitive position. That said, Platform-as-a-Service (PaaS), which is the most profitable segment, is still small (~5% of total market), so material scaling for MCS and AWS remains a medium-term opportunity, rather than short-term.

Overall, we think for Mail there is unlikely to be a significant near-term financial impact from the announced partnership. However, longer-term the company’s new initiatives, such as cloud or online education, have material growth potential, and today’s deal is positive given it should strengthen Mail’s position in cloud. Mail remains our top pick in Russian as 1) we expect its well-diversified revenue structure to make it relatively resilient to COVID-related headwinds; 2) we see scope for games to surprise on the upside in the near-term; 3) MoEx listing should help improve liquidity, with a potential MSCI Russia inclusion later in the year; and 4) Mail trades on 11.7x EV/EBITDA 2021E, a 34% discount to , on our estimates.

Panarin Kirill Telephone: +7 (499) 956-4216 Maryana Lazaricheva Telephone: +7 (499) 956-4217

US election, the virus gets less important as it spreads faster, SA and its IMF deal, Ethiopia, Nigeria currency

Just reading a US bank’s analysis about a Biden victory. It was less than two months ago that betting markets (and some clients) thought we were wrong to say Trump will lose the 2020 election. Since then polls have massively swung to posting the result that always comes when the incumbent presides over a recession. We cared about this because a Trump defeat will encourage net FDI to flow to the rest of the world, which should weaken the dollar and is EM bullish for the next 4-8 years. With China representing 41.5% of MSCI EM – and China’s stock market jumping - we might be seeing Chinese retail investors front-running the global financial community on this trade of the decade, encouraged by local newspapers 

Given humanity’s desire to hang a story around a market move, there is a decent chance that potential stock market weakness in coming months will be attributed to the likely change of president (the US bank says that’s not justified). This chart of US stock market moves in big recessions suggests we have about 3 weeks ahead of us before markets soften – if this is like 1937 or 2000/01. This tiny sample suggests the best case is the markets trend sideways if this is 1973 or the 2007/08. We find the historical comparison interesting because it implies that the flow of news and market psychology is similar in each crisis and runs something like this … (1) this is bad, (2) it’s not that bad, (3) oh, it is worse than we thought. We still hope it’s different this time, due to unprecedented stimulus and the recession being triggered by a virus not a structural economic problem like a tech bubble (2000/01) or massive spike in oil prices (1973) or excess financial sector lending for years (GFC). There is a good case that DM equities need to be re-priced upwards by the low US bond yields. This is the bull case. The caveat is worth repeating that I hoped it would be different this time in early 2008 as well. The bear case might be justified by the job loss announcements coming thick and fast at present. Another week has passed with southern US states sticking to our base case view that they will not re-impose lockdown. Mandatory mask wearing and closure of restaurants maybe, but no lockdown. So far, it is still only which has re-imposed a hard lockdown. Azerbaijan is at 2.5 deaths per 100,000 people. If every US state with a higher death toll followed the Azeri model, 99.3% of US GDP would be shut down (only Montana and Hawaii have less deaths). Most US states have more than 10 deaths per 100k (see graph at the end). As virus cases rise, they are becoming less important to markets. India is not yet recording the highest number of cases daily in the world but it will and will add tens of millions of confirmed cases (probably hundreds of millions unconfirmed cases) to the current global total of 11m cases. While we expect India’s coronavirus figures will become front page news, we doubt they will stay there very long. SA is on course to be the first African country to see deaths top 10 cases per 100,000 people. I’d guess that happens in August when the country’s coronavirus death toll touches 6,000 (just over 3k today). To put that into perspective, UN estimates imply South Africa will suffer 545,000 deaths in 2020. A breakdown of deaths for 2016 suggests SA had more fatalities due to assault than is likely from coronavirus deaths in Jan-August 2020. It is because of the relative unimportance of the coronavirus in raising emerging market mortality that we assume no re-imposition of lockdown is coming in EM either (Azerbaijan was a surprise). DM should get hit harder as it is older, but even here deaths don’t seem high enough for local politicians to justify renewed lockdown. Curiously, we discovered that Arkansas had more deaths from traffic accidents in 1H20 (282) than from the coronavirus (270). This happened even as accidents fell, presumably because cars were going faster. The lag effect from new cases suggests Arkansas coronavirus deaths will be much higher than car fatalities in all of 2020, but the main reason we mention this is because the Financial Times and others keep citing “excess deaths” as the best metric to see how deadly the coronavirus really is. But in Arkansas, some of the excess deaths will have been caused by higher car fatalities. Presumably if car accident fatalities had been 270 less than usual, the implication would be that the coronavirus caused no deaths in Arkansas at all in 1H20. Which would obviously be incorrect. There is no perfect metric. With regards to markets – every currency proxy I look at (yes I mean the ZAR really) is telling me we are in wait and see mode – and unlike in April (the bullish EM FX report) or May (the bullish EM report) I don’t have a view that differs significantly from market pricing. Except perhaps in long-end ZAR bonds – because I do expect SA to get a $4.2bn IMF loan this month, as Tito Mboweni suggested the other day. As in Nigeria, there is political resistance to accepting IMF support, but given few conditions and how cheap it is, and the ANC’s majority, it is looking probable. Egypt – the currency has not moved marginally weaker with the new fiscal year – it’s actually strengthened. Not sure what to make of that, unless it is global money picking up double digit yields. Pakistan has also seen the currency strengthen a little. The EGP is 15% overvalued based on our London REER model, while the PKR is the cheapest in south Asia and 7% undervalued relative to its own history. Ethiopia is looking a little messy – with no growth assumed by the IMF for 2021, and protests in recent days prompting an internet shutdown. No elections until the pandemic is over, Ethiopia has declared. Easier to like Kenya sovereign today. Nigeria – the CBN has moved the official rate which started the year at NGN305/$, and then shifted to NGN360/$ in 2Q20, to around NGN380/$ – which is close enough to the I and E window to count as pretty well converged. That means another 5% in the coffers of the government every time a barrel of oil is sold, or a dollar of IMF or World Bank money arrives in the country. It helps. It’s still considerably overvalued though Yvonne is looking for NGN449/$ for year-end, her model implies fair value is NGN427/$, our London REER model says NGN491/$ is the average rate for 1995-2020 in today’s money. CONCLUSION: Global investors are now assuming a Trump defeat in November, as we forecast in May, but don’t yet seem to have priced in the likely consequence of that, which is the bullishness this implies for EM. As our bullishness rests on dollar weakness driven by FDI flows – and as there won’t be many FDI flows in 2021 given this year’s global recession – perhaps the markets have decent reason to be cautious on this. We don’t see much market pricing that we disagree with at present.

Charles Robertson Telephone: +44 (207) 005 7835

© 2020 Renaissance Securities (Cyprus) Limited. All rights reserved. Regulated by the Cyprus Securities and Exchange Commission (Licence No: KEPEY 053/04). Hyperlinks to important information accessible at www.rencap.com: Disclosures and Privacy Policy, Terms & Conditions, Disclaimers