Interview: Sameer Sharma, Data Science Consultant & former Central Banker
* ‘unemployment, seasonality adjusted monthly Government tax revenues and near zero tourist arrival numbers point to a Mauritian L-shaped recovery’
* ‘If one were to assume that all will be over soon with a magical vaccine and that Mauritius will have a miraculous V-shaped recovery: Wishful thinking’
Sameer Sharma, Data Science Consultant currently based in the US and former Central Banker, argues in today’s interview for a “gradualist risk managed border opening strategy”. The collapse in aggregate demand in the economy, he says, is significant and “neither the MIC, interest rate cuts, moratoriums on interest payments nor the wage assistance scheme will be able to make up for the large slack if we do not adopt a gradualist risk managed border opening strategy and implement stimulus measures more efficiently and soon.” As regards what type of recovery we should expect post Covid, he says that real economic indicators such as unemployment, seasonality adjusted monthly Government tax revenues and near zero tourist arrival numbers point to a Mauritian L-shaped recovery.
Mauritius Times: How is the Mauritian economy performing relative to the rest of the world more than eight months since the start of the Covid-19 crisis?
Sameer Sharma: As at the beginning of September 2020, high frequency economic data of major developed economies in North America, Europe, Australia and emerging China continued to point to a still fragile but sustained recovery with GDP growth now casts and alternative data showcasing a recovery to levels that are now nearing activity levels last seen in the pre-Covid period.
While these growth indicators are positive, doubts persist. 70% of growth indicators are up compared to last month and are part of a trend seen since April but unemployment rates remain high: 8% in the United States, down from 14.7% in April. In the Eurozone, unemployment rates have moved little because of the social shock absorber system, but recently the unemployment rate has begun to rise from 7.3% to 7.9%: a limited variation but potentially the beginning of a trend that should be followed carefully.
The same applies to default rates especially for non-investment grade companies (Anything below this ‘BBB’ rating is considered non-investment grade. If the company or bond is rated ‘BB’ or lower it is known as junk grade, in which case the probability that the company will repay its issued debt is deemed to be speculative). Many other emerging economies which are less prepared to deal with the pandemic or like Mauritius which have chosen to close borders are lagging behind with Mauritius being one of the worst laggards right now.
Source: MSCI data, SEMTRI USD Index
* Have the unconventional monetary and fiscal measures taken by the Government worked so far?
Mauritian policy makers have announced more unconventional monetary policy (quasi helicopter money) measures than almost any other country on earth as a percentage of its GDP, but the SEMTRI1 which had already been lagging behind its emerging and developed market peers over the past 5 years has barely recovered and is now well behind even African equities. Other real economic indicators such as unemployment, seasonality adjusted monthly Government tax revenues and near zero tourist arrival numbers point to a Mauritian L-shaped recovery. In sum, Mauritius is stagnating and the outlook is not rosy.
* Why is Mauritius lagging behind despite more than 150 Billion Rupees of unconventional monetary policy, roughly 34.5% (150Bn/435Bn) of this year’s projected GDP?
The Mauritian economy was already out of breath before the Covid-19 crisis started because we have failed to tackle our structural ills, and secondly because the non-financial private sector itself which had unspectacular free cash flow to debt metrics and return on capital metrics coming into the crisis had limited investment capacity. The blacklisting of Mauritius has not helped in terms of global business fee income and the worse is to come if we remain there for long.
Foreign villa investment flows and tourism receipts which kept old skeletons in the closet for years have seen their taps shut off. The collapse in aggregate demand in the economy is significant and neither the MIC, interest rate cuts, moratoriums on interest payments nor the wage assistance scheme will be able to make up for the large slack if we do not adopt a gradualist risk managed border opening strategy and implement stimulus measures more efficiently and soon.
Mauritius hosted 1.3 million tourists last year, a figure roughly equal to its population and we have a large number of residents and local citizens stuck abroad. Wages have also been cut and all this is impacting demand.
Furthermore, when you speak about these unconventional monetary policy measures, the devil in Mauritius is not about ideas but about implementation. For example, the Bank of Mauritius unlike other central banks which are engaging in asset purchases to quickly throw money into the system decided to expand its liabilities by apparently issuing Bank of Mauritius paper in order to mop up excess liquidity and grant this money to Government.
As at June of 2020, the Bank of Mauritius had some MUR 79.6 Billion of monetary policy instruments as liabilities compared to more than MUR 128 Billion in January and as at last week had increased it back to MUR 110 Billion. For the BoM to make the grant, it must increase the amount outstanding by at least MUR 60 Billion. At the same time, the BoM must have enough of these monetary policy instruments so that it can maintain interest rates at a level that is consistent with its monetary policy objective.
All this puts a lot of pressure on its liabilities. The Bank of Mauritius had MUR45 Billion in capital and reserves (equity) in June which is lower than the projected increase in its liabilities in order to fund the one-off grant to Government. Unless there is continued depreciation of the Rupee which will bloat the Rupee value of its foreign assets as has been happening so far, the BoM will have a tough time financing this grant and maintaining an appropriate level of economic capital.
You can understand that the way the stimulus has been designed can slow down the implementation and create long lasting balance sheet complications and monetary policy implementation credibility concerns, which is why the BoM must change its strategy. It would have been better for the Government to issue a 100-year perpetual bond bought by the central bank which would have increased the latter’s assets rather than the current strategy which pushes up its liabilities. Central banks for this reason expand their balance sheets (purchase bonds to increase assets) not their liabilities. The Zimbabweans had done a similar move with their central bank a decade ago. This is not the example to emulate.
The bulk of Government expenses is for now being financed with debt issuance and by the time this crisis is over our debt levels will be unsustainable even with BoM grants and we will not be able to reconcile this debt burden with that of a low tax regime anymore. This analysis ignores unfunded pension liabilities which will come home to roost soon enough. The risk of a credit downgrade will be much higher post crisis and this is not good for global business companies’ deposits held by domestic commercial banks and beyond.
* Countries like the Maldives, Greece, Israel, just to name a few, opened their borders and saw an increase in Covid cases. Wouldn’t opening up our borders not hurt the economy even more if we need to impose another lockdown subsequently?
We are talking about 60% to 70% effective rate for a Covid-19 vaccine. The current Russian and Chinese vaccines have an effective rate of less than 50%. Even with a vaccine, we will need to live with this virus and that assumes that everyone in Mauritius gets vaccinated. There is little clarity from Mauritian policy makers in terms of what will trigger a more pronounced opening of the border.
So what will we do: shut down or have limited border openings for months or years to come and take zero risk? Are we even pre-ordering vaccines? A zero risk strategy is not a realistic strategy.
I fear that the current political landscape is influencing the “it is complicated to open the borders” strategy. The more you suppress aggregate demand and bleed corporate balance sheets, the more the Zombie companies you shall create. You will suffer the consequences for years to come in the form of structural unemployment and subdued investment levels. Jobless people suffer too and a balance must be struck in order to avoid rising default risks.
The lack of job creation in Mauritius, and falling tax revenues and increasing debt is not sustainable. Mauritius needs a gradualist risk managed strategy, and it needs competent people who are willing to take accountability and manage the re-opening of the borders so that we minimize the risk of a second lockdown. Beyond hotels, think about what is happening to the bloated foreign villa real estate market right now!
* Are you advocating for the complete re-opening of our borders?
No, I am not for a complete opening of borders at this stage, but we must start in phases and do it soon as a zero-risk strategy is a negative return strategy for sure.
If you just take Beachcomber, Lux Island Resorts, Sun Resorts and Constance Hotels Services Ltd and just focus on their Mauritius sourced revenues and adjust for the first quarter of 2020, we are talking about more than MUR 14 Billion in foregone revenues for this year alone on aggregate. It will take years for revenues to come back which means that the bleed in revenues will be much more than this just for these companies and I am ignoring Air Mauritius, restaurants, smaller hotels, sugar companies turned smart city developers and so on.
The rest of the world is not crazy to open up. They have realized that maintaining closures can do more harm than good especially when death rates due to Covid-19 are manageable. Mistakes are being made but they are learning from them. We must learn to live with the virus but with strict sanitary norms and do risk management. Sure, we will not be flooded with tourists at first, but it will help us get ready and be well-prepared and also it will help us not lose more market share.
* Won’t the Mauritius Investment Corporation Ltd help make up for the slack?
If one were to assume that all will be over soon with a magical vaccine and that Mauritius will have a miraculous V-shaped recovery, then surely that would amount to wishful thinking.
When I wrote about the need to engage in unconventional monetary policy measures way back in early March, I did it because I was familiar with the realities of local corporate balance sheets and also because I think this virus and its consequences on the global economy will be with us for much longer than we could just borrow for. This crisis is a multi-round game under Game Theory and the recovery will be gradual and fragile for Mauritius. There was also a need to act quickly because Mauritius will not benefit from low levels of inflation forever and will at best only have one shot to get things right.
Many of these companies operating in the tourism and manufacturing sectors did not need more debt but a more optimal capital structure that would correspond to the realities of a post-Covid world. This was why I wrote extensively about the need for convertible preferential shares v/s further debt. I also spoke about the need to have a modern private equity type fund which would strike the right balance between taxpayers getting a decent risk-adjusted return for the bailout and for these companies to get the breathing room they needed in the form of capital that was needed. A gradual opening up of borders with strict rules will also help generate some revenues.
If taxpayers who own the central bank do not get a decent return on investment for the risk they will be taking, then the populist wrath will be terrible. Think about the losses and opportunity costs that can come from a badly managed MIC which loses money and is unable to save jobs. Think about smaller business owners and the average person who sees large firms getting bailouts while he is left with peanuts.
A suboptimal outcome for the MIC is worse than no MIC. The idea of what is now known as the MIC was to create an off-balance sheet and bankruptcy remote special purpose vehicle with good governance like that of western peers. It would itself have an optimal funding structure financed by the state and the central bank in order to optimize its returns, and then invest in distressed albeit viable companies accompanying them in this long journey as a partner as they restructure and change business models.
In order to avoid any perception of conflict of interest and get the Government out of managing private companies, I wrote about the need to have a qualified board with relevant experience in global or private markets to issue an investment policy statement. It would not just focus on returns but on impact investing criteria especially on the jobs front, and outsource the investment job to investment professionals who have proven experience and track record in distressed investing or private equity. This is not a field where one can be penny wise and pound foolish, nor can anyone google for this kind of knowledge.
I still believe that we need to move in this direction and fast because the structural ills which plague the private sector will not magically disappear post Covid-19; nor will the state be able to remain the sole sheriff in town when it comes to investing in the economy for much longer. You need a healthy private sector to do that.
* You spoke about structural ills which plague the Mauritian economy and you paint a gloomy outlook. Lord Desai recently spoke about Mauritius needing to follow the Estonia model, but that was before the onset of the Covid pandemic. As someone who works in the Artificial Intelligence (AI) space in the United States, what are your thoughts about Mauritius following that?
Mauritius is an easy fix if you put the right people in the right places. I am concerned about the future of my country, but I am hopeful that at least one of the unintended consequences of this crisis is a less benign population who will demand better from the Government.
Regarding Estonia, we need to be realistic. The level of Mathematics and Science is declining even at the local public university level; do an internet speed test on Mauritian connections and compare it to the rest of the world and see how we rank; look at the ranking of Mauritius when it comes to African Fintech startups which are receiving funding; look at the publicly available AI strategy paper for Mauritius; search for when we plan to have 5G in the smart cities and on the island… do all these and you’ll end up with the conclusion that we are perhaps aiming for Zimbabwe but certainly not Estonia. Google, Huawei and Microsoft are setting up AI research hubs in the continent, not in Mauritius.
* How do we correct that perception?
Rather than building a high-tech hub with bricks and mortar, we should be investing in our data infrastructure and we should be busy revamping our education system. The University of Mauritius is free yes, but how about quality?
In the United States more than one third of companies struggle to implement a profitable AI strategy because AI is not easy. It is about data quality, it is about having the right information architecture with scalability in place, it is about having the best and brightest (think openness to immigrants), it is about having good AI governance to avoid unintended consequences of AI, etc. From watching parliamentary debates with the current Speaker, do we seriously believe that a sophisticated debate can be held about AI regulations? As one former Governor of the Bank of Mauritius told me last year, Mauritius is already full of artificially intelligent political nominees. We have a different kind of AI over here.
Take the Safe City project which apparently has facial recognition capabilities. Typically, one uses Convolutional Neural Networks with transfer learning in order to obtain high detection rates, but what I can tell you is that such models have been found to be as racist as the data which feeds them especially against African Americans having high false positive rates (rejections and security alerts). Did we even think about such unintended consequences in Mauritius? Will we conduct bias testing? Do we know how to?
Natural language processing models like BERT which leverage transformers and which you have all used when you google has been found to be racist and sexist. Credit underwriting models using AI models, which are not well tuned, discriminate too. AI can be good in that it allows you to replace humans who do mundane repetitive tasks and, if well-tuned, it can be a force for good and increased productivity given its ability to scale up. But if we think that building a technology park on the island will bring AI, we are indeed living in Dodo Land.
- The Total Return Index, the SEMTRI, is an index, which tracks the price performances of the constituents of the SEMDEX and ensures that the dividends paid by these constituents are reinvested. (Stock Exchange of Mauritius)
* Published in print edition on 11 September 2020