Unlike a commercial bank, central banks can technically continue to operate even with negative net worth. In effect, a central bank can print money/create liquidity to conduct its operations and cannot go insolvent. Laws which mandate a certain minimum level of economic capital can themselves be amended. However, there is general consensus across central banks that they need to hold an adequate level of economic capital and need a proper risk based economic capital framework to do so. The role of central banks’ financial resilience is to enable these institutions to focus on their primary function of fostering monetary, financial and external stability, even in the midst of crisis, without being diverted by balance sheet concerns. This is particularly important given that central bank capital generally represents public resources and the central bank’s management can be held accountable for its losses (at least globally). A well-capitalized central bank reduces the risk of having to issue money to finance itself amidst instability (e.g., to meet its obligations as lender of last resort). Thus, the country and the central bank are better prepared to deal with a range of critical situations. The Bank for International Settlements (BIS) defines economic capital as the methods or practices that allow banks to consistently assess risk and attribute capital to cover the economic effects of risk-taking activities.

The Bank of Mauritius (BoM), like other central banks, faces multiple risk factors where an appropriate level of economic capital must be set aside for. The BoM manages the country’s foreign exchange reserves and these investments are subject to market risks at varying degrees depending on the asset classes the central bank invests in. Gold for example is a highly volatile investment and has an expected shortfall (a stress risk measure) of above 30%.

The BoM like other central banks faces market risk given its current equity investment in the MIC, a structure which is itself now highly leveraged, credit risk be it internationally or domestically via the advances made to Government and to the MIC which remember invests in distressed (high default risk and with recovery rates which are highly dependent on asset values at tme of default and collateral rankings) companies over a long time period, foreign exchange risk given that the majority of its assets are held in foreign currency but its balance sheet is in Rupees, liquidity risk given its investments in highly illiquid and distressed “convertible” bonds which will likely be held to maturity and have been wrongly structured, operational risks based on the way it runs its central banking, foreign exchange reserves and MIC operations and last but not least contingency risks which may require additional capital to withstand periods of shocks such as what we have gone through recently with the pandemic. While many of the top emerging market and developed market central banks hold median economic capital ratios of close to 8%, this does not in anyway mean that any individual central bank should aim for such a target. Some may be over-capitalized while others may not be. The determination of capital adequacy should be risk based.

In terms of trends, central banks are increasingly adopting a model-based approach for assessing all these risks which allow them to estimate the right level of economic capital relative to their assets. The Reserve Bank of Australia, the Bank of Canada and the Bank of England for example estimate their capital adequacy by stress testing their balance sheets based on historical and forward-looking extreme event scenarios in order to determine an appropriate level of economic capital to asset ratio. The Reserve Bank of India, the European Central Bank, the Norges Bank and many others use an Expected Shortfall measure at the 97.5% to 99% percentile to measure extreme tail risk events of all of these risk factors in order to determine the appropriate level of economic capital. Many central banks which had been using Value at Risk to estimate the appropriate level of economic capital have moved to the more conservative Expected Shortfall measure since the last Great Financial Crisis.

Central banks of course being regulators of banks like to lead by example and are very transparent about the specifics of their Economic Capital Frameworks and continiously reassess their balance sheets before paying out any dividend on any profits to Government or before making any capital transfers to Government. Central banks only rarely transfer excess reserves to Government and clearly elaborate how they have excess reserves based on the balance sheet risks before ever doing so. When making such determinations, i.e. that the central bank has more economic capital than they really need based on current and stress scenario risks across all risk factors, Boards of central banks advised by professionally managed risk management departments also focus a lot on the key components of their economic capital before making any distributions. For example, unrealized gains driven by domestic currency depreciation which bloat foreign held assets and hence the level of special reserves are heavily discounted before making any payments.

In the case of the BoM, its Economic Capital Framework and the models it uses to continuously assess balance sheet risk is unknown. For example, in June 2020, the BoM had an Economic Capital of 14% which was largely driven by Rupee depreciation which goes to the Special Reserve Fund. Minimum capital had also been increased from MUR 2 Billion to MUR 10 Billion by July 2020. At least 80% of the BoM economic capital by July 2020 was still made up of revaluation reserves so the 14% level should be heavily discounted. Given the volatility of foreign assets held in Rupee terms on its balance sheet (currency risk) and all risks faced by the central bank and especially considering its liabilities with more than MUR 87 Billion in monetary policy instruments in July 2020, it will be interesting to understand how the Board of the BoM ever thought that they had excess reserves to distribute to Government in the form of a grant vs. an advance? What models did they use to estimate the appropriate level of economic capital and what buffers did they add to cater for model estimation limitations? Did the Board take external events and contingencies which are hard to model beyond stress tests into account? Did the Risk Management Department of the BoM provide a report to the Board with calculations and recommendations based on a Board approved ECF framework and were the numbers and methodologies audited by its internal audit department? Given human capital limitations, did the Board ask for an independent Asset Liablity Study to be conducted by external advisors who understand central banking and have credibility in this field for validation? What kind of stress scenarios were conducted and who signed off on their appropriateness? Why did the BoM decide to pay MUR 60 Billion in grant when the level of economic capital it had was much lower at MUR 45 Billion? Before disbursing these funds to Government, did the Board take into account the risks it would be facing via its MIC investments which are being made in distressed assets and did it plan to set more capital aside for this? How is giving more money than you have in equity (a grant cannot create assets as offset) consistent with a formal Economic Capital Framework and in line with international best practice?

“The IMF ‘recommendations’ begin to make more sense”

When we think about all these questions which observers and politicians seldom ask in Mauritius, the IMF “recommendations” begin to make more sense. The grant being larger than the level of equity of the central bank itself, the actual grant amount had to be reduced significantly by almost half while the remainder would now have to be treated as an advance and would thus form part of domestic assets. While the IMF did save the BoM from having negative equity, the level of economic capital of the Bank of Mauriitus is now very low. Today, the BoM has MUR 13.5 Billion in capital and reserves plus some MUR 10.2 Billion in comprehensive income which together leads to an estimated economic capital of 5.4%. Even a 7% appreciation of the Mauritian Rupee (this is ironically bad for the BoM) vs its foreign assets, global market volatility on its foreign investments or any losses from its MIC investments (which would make the MIC unable to service the loans) could all wipe out this level of capital very quickly. The BoM also needs to estimate expected credit losses and lead by example on the loans being made to the MIC which is not easy to do with distressed assets.

Without needing to conduct any modelling exercise on the assets and liabilities of the BoM taking into account current, expected risks and contingencies, the Board of the BoM would be hard pressed to come up with an economic capital framework and rationale which could possibly justify 5.4% in economic capital as being adequate given the volatility we have seen on the balance sheet just over the last 2 years. Only a consistent and gradual depreciation of the Rupee which may not be consistent with its monetary policy objectives given the rise in commodity prices globally, continued diversification of its investments and a much more modern reserve management framework could allow the BoM to even hope to better match its liabilities and maintain let alone grow its level of economic capital over time. We may even see the BoM now inject more liquidity via the purchase of Government bonds and other instruments in the secondary market in order to expand its balance sheet.

As I have preciously and consistently highlighted, the way the MIC has been structuring these convertible bonds is very far from international best practice and has significant mark to market implications for an MIC which is itself increasingly being financed with debt. Too many do not understand the differences and nuances between convertible bonds and bank loans including politicians. The likes of the ECB and the Federal Reserve can hire the best and brightest in the world and even they go through experienced asset managers such as Blackrock to manage their bailout funds. It is nowhere acceptable to lose money when bailing out the private sector especially when some companies had very weak balance sheets well before the COVID-19 virus ever left Wuhan.

You need a very sophisticated team to professionally manage distressed investments and price the risks properly. The MIC is currently funded with MUR 1 Billion in equity and more than MUR 79 Billion in debt as per the latest BoM balance sheet. This is one of the most highly levered structures on the planet today and MIC investments are mostly distressed. This is quite the risk scaling. Even if the BoM allows the Government to buyout its 1 Billion of equity investments which will allow it to take the MIC off its balance sheet, its 79 Billion in loans will certainly still face MIC investment exposures within a highly geared entity. The likely story is that the Government lacks cash and will not be able to inject a lot of equity when it takes over the MIC. The Government can buy out 1 Billion of equity from the BoM but not a lot more. Even injecting MUR 10 Billion more will still lead to quite the leverage and one must really hope that the MIC team are investment gurus will get it all right. Given what we have seen on the bond structuring side so far, this is not likely. The concept of the central bank lending to the MIC is fine but the level of leverage is just too high. It seems like the Government is getting ready to inject billions in strategic investments into the country mainly financed by a highly geared SPV but the track record of Government making good risk adjusted returns in the past for 50 years has been poor.

In order to even start to offset its risk, the MIC will need to over collateralize its investments and ensure that it is also first ranked when it comes to its distressed investments. Convertible bonds are normally ranked after all other charges or have no collateral at all with near zero recovery rates. It will need to deeply discount the values of these hotels and land values and ensure that it is first rank which will impact commercial bank rankings and their own provisions on troubled companies. Such a framework may allow the Government as sole equity holder to justify a low level of equity investments in quite the sophisticated Special Purpose Vehicle. If the MIC over estimates its recovery rates and in general if the investments are not asset backed enough, it will invariably be the BoM as lender which will carry most of that risk which will of course require a lot more capital than it even needs or has already. The sad story is that a lot of this could have been avoided last year at the design of the MIC and at the design phase of these unconventional monetary policy measures. The Government should be careful to remember that leverage and not having the right people at the right places can make for bad bed fellows.