The COVID-19 economic shock provided an opportunity to assess the effectiveness of the overhauled bank regulatory framework. In mid- to late March 2020, as the scope and scale of the potential disruption became apparent, a series of events led to the near-complete freezing of some of the core global financial markets. As asset prices suddenly dropped, financial institutions (including banks) pulled back significantly from the market, since the newer regulations had effectively raised the cost of market-making and repo activities. In response, banks maintained their liquidity levels both by deploying their ‘liquidity buffers’ – the holding of which was a key requirement of the updated regulations – and by going to central bank discount windows to exchange assets for cash. Central banks such as the US Federal Reserve and the ECB became buyers of last resort to avert a market meltdown.
If the objective was to avoid a repeat of the bank failures of 2008, then the BCBS regulatory framework can be considered to have worked well – at least, so far. Of course, no two crises are the same. From a high-level perspective, the global financial crisis is considered to have been a product of banks’ poor behaviour, in terms of both their risky activities and the excessive liabilities they held against their assets at the time. Moreover, the crisis mainly affected the US and Europe. This time the situation is very different. The economic shock originated in a global health crisis, rather than in poor practice on the part of banks. Furthermore, where banking systems entered the crisis in good shape, banks have managed to provide financial support to the real economy quickly, thereby also helping to transmit government policy. In other words, banks have been part of the solution, not the problem.
Since the start of the pandemic, there have been almost no bank failures. Even taking recent loan loss provisions into account, banks’ capital and liquidity levels remain well above the required minimums. Countercyclical capital buffers, designed to be built up in good times and drawn down in bad, have been loosened quickly in some countries (e.g. Sweden and the UK), exactly as intended. However, not all countries required local banks to hold positive countercyclical capital buffers going into this crisis. The US, for one, was a market where banks had not built up such buffers. This made it harder to free up capital to allow banks either to take losses without triggering regulatory thresholds or to increase lending to support the economy. Nonetheless, given the scale and scope of fiscal and monetary policy support in most countries, large credit losses have not (yet) materialized, and there is little evidence of unmet credit demand from firms or households, suggesting that banks are themselves not capital-constrained.
Of course, a financial health warning is necessary here: there could yet be substantial hits to banks’ balance sheets as government subsidies are withdrawn from various sectors and from labour markets, and as credit defaults impact the economy. In this scenario, bank profitability would likely be affected for years to come, exacerbated by now-ubiquitous low-interest-rate policies.
Unintended consequences of post-2008 reforms
In addition, not everything in the current regulatory framework has worked as desired. Indeed, there have been some perhaps unintended consequences, particularly in terms of banks’ market-making ability during periods of turbulence. This situation developed in the run-up to the COVID-19 crisis, but its full consequences only became apparent during the crisis itself. Specifically:
- For larger borrowers, there has been a shift in the type of financing preferred, from loans to bonds. On an average annual basis from 2008 to 2020, global corporate bond issuance grew at a rate of 8.3 per cent, whereas traded corporate loan issuance grew at a rate of just 1.9 per cent. Bond issues have shrunk in average size, with the result that smaller borrowers are now also more able to fund themselves via financial markets instead of solely with bank loans. But market-based financing is risky, particularly for borrowers in smaller, less liquid markets such as the European high-yield market. Such markets are prone to drying up when there is a shock, as they did in the spring of 2020. Fortunately, in that instance banks were able to act as liquidity providers of first resort to borrowers.
- Banks are less likely to provide market-making liquidity in the key financial markets such as those for bonds. Liquidity is now often provided by non-bank financial institutions (NBFIs), including asset managers, insurers, hedge funds and specialist liquidity providers. However, NBFIs have struggled to be reliable providers of liquidity in turbulent markets.
- At the worst point in the financial reaction to the pandemic, the US Federal Reserve and other central banks had to provide a backstop to their local markets across a range of assets. These included, in the US case, the markets for Treasuries and agency debt, but also those for corporate bonds (and commercial paper), municipal bonds and so on. Between March and June 2020, the Fed’s balance sheet increased by nearly 70 per cent in size.
Do these unintended consequences matter? And, if so, what should be done?
Capital and liquidity buffers and their use in the pandemic
The increases in capital and liquidity buffers, to protect banks from downside ‘tail risks’ (i.e. extremely low-probability events), were determined by applying hypothetical stresses to banks’ financial projections and calculating how large the buffers would need to be for each bank to survive the stresses. Banks had to consider scenarios based on stresses to the general financial markets and economic conditions. In these scenarios, banks had to be able to survive certain minimum stresses prescribed by regulatory requirements. They also had to determine if the results demonstrated adequate protection for their needs, or if they wanted the capacity to survive a worse stress (i.e. one that was deeper, longer or both).
The total set of buffers in the banking system had been built up through a series of discrete reforms, each of which had added to the buffers and protected against particular risks. These risks included: intra-day liquidity stresses; short-term (i.e. 30-day) and medium-term (i.e. one-year) liquidity stresses; financial market counterparty stresses; and the general recovery and resolution of each bank. Local banking regulators also added their own stress tests – including, as mentioned, the CCAR mechanism in the US – against which the banks had to hold yet more buffers. Today, as a result of some of these changes, G-SIBs must hold additional layers of protection or so-called ‘G-SIB buffers’; the bigger and more complex the bank, the bigger the buffer must be.
The reforms have made banks more resilient, not only through the addition to mandatory financial reserves but also through improved risk management and public disclosures. Banks’ understanding and capturing of the risks in their balance sheets have markedly improved since the 2008–09 financial crisis. Modelling of the impact of stresses is clearly better (interestingly, this has led to banks being more selective about the risks they run, with some institutions either reducing risky activities or hedging those risks). Banks have to publicly report key metrics, such as their liquidity coverage ratio (LCR). Regulators report other measures of bank health, including the size of G-SIB buffers and stress test results.
The core rationale for this system was that, in times of stress, the buffers would be drawn down, giving banks time to take mitigating actions to avoid running out of liquidity or becoming insolvent, and without having to rely on central bank or government support. Intermittent economic shutdowns and other mechanisms to restrict the spread of COVID-19 will continue to test the effectiveness of capital buffers, as the creditworthiness of borrowers is likely to deteriorate. However, a full reckoning of the effects on banks, as noted, will be delayed by the fact that governments have continued to provide fiscal support to households and firms.
Were banks adequately prepared for a financial market shock of such magnitude? As part of the updated regulations, banks are encouraged to ‘war-game’ liquidity and capital stresses, putting their plans to the test at least once a year. They are also required to provide resolution plans to their lead regulator to ensure an ‘orderly’ wind-up in the event of their failure. In this context, the financial market shock of March–April 2020 was not completely outside the scenarios envisioned in policy ‘playbooks’ – that said, the rapid and intense reaction of the real economy still took many banks, regulators and governments by surprise.
Since central banks around the world stepped in to provide a backstop to financial markets by providing as much market liquidity as needed, we will never know how well prepared banks really were. In some cases, banks did not use their buffers before turning to central bank lending facilities. Many explanations for this have been provided, including: (a) the fact that the duration of the financial market shock was (and remains) unknown, so banks wanted to hold buffers in reserve for the next phase of the unfolding crisis; (b) the fact that regulators instructed banks not to dip into certain parts of their buffers; and (c) banks’ own worries that, were they to draw down their capital buffers, it might become public knowledge that such buffers had run low, with corresponding implications for confidence in their soundness.
Did the system get the balance right between individual banks having sufficient buffers and central banks offering to turn almost any assets into cash? In effect, central banks provided a public service (averting the collapse or shutdown of financial markets) at a public cost (taxpayers’ money), which also resulted in private benefits (financial market players not experiencing extraordinary losses) at the risk of increasing moral hazard. The central banks provided coverage of the tail risk beyond what the individual banks were ready for. Getting this balance right – motivation for self-insurance to mitigate moral hazard, versus provision of insurance against extreme tail risks to prevent a systemic spiral – is a central problem for those tasked with maintaining financial stability. Was the boundary between public and private protection set appropriately?
