Finance: Research, Policy and Anecdotes

As the COVID-19 pandemic continues its march across the globe, with the epicentre having moved from Asia to Europe (with another move towards the Americas possible in a few weeks), the socio-economic disruptions have increased the likelihood of a global recession and with that of problems in the financial system.  What can policymakers do now and what should they be prepared to do when the damage will become clearer when the virus has receded?   While there is a high degree of uncertainty and limited information, lessons from previous crises can come in useful to define possible policy actions and gauge policy actions that have been taken.  The following discussions will focus on three main policy steps:


(i)             Governments have to stand ready as the ultimate loss absorber, in the first instance for the real economy, but also for the banking system

(ii)           This has to happen on the EU or eurozone level, given limited fiscal space for some countries

(iii)         This has to be announced clearly and early on to create confidence


Proactive measures needed


COVID-19 and the disruption it has caused and continues to cause constitute an enormous shock to both real economies and financial sectors, reflected in financial market distortions, mis-aligned prices (which arbitrage should make unsustainable in normal times) and funding concerns for many market participants, including banks. The problems originate in the economic disruption: not earning money, households might not be able to repay mortgages and consumer credits; not having clients or not being able to produce goods/services results in lost revenues for firms, undermining their ability to repay loans. But it is more: where they have them available, firms are drawing down credit lines to have a sufficient cash buffer during times of economic disruption. This trend is exacerbated as access to market finance is drying for most firms that used to have access to it.  While banks can help overcome their clients’ liquidity constraints, they have a limited ability to do so.  Initiatives such as by the UK government and Bank of England as well as of the German and French governments to offer funding liquidity for banks and credit guarantees can be helpful in this context. Fiscal support measures can also help mitigate the negative effect of the crisis on banks’ asset quality by ultimately targeting the source of the losses.


One of the defining characteristics of bank lending over the business cycle is its procyclicality: in a recession banks reduce lending rapidly, especially to smaller enterprises and riskier borrowers. While some of this lending retrenchment is demand-driven, agency conflicts at the core of banking point to substantial supply-effects. Regulation can further exacerbate this procyclicality, requiring forward-looking loan classification (as under IFRS 9) and provisioning, as well as forcing an increase in risk weights and thus capital.   Anticipating negative effects from the disruption for real economy and financial sector and mitigating these negative effects is therefore critical and urgent. Steps taken by supervisors across Europe to (i) lower countercyclical buffers where currently above zero, (ii) reduce capital requirements, allowing banks  to operate temporarily below the level of capital defined by the Pillar 2 Guidance and the capital conservation buffer (CCB), and (iii) allowing banks to operate below 100% of the liquidity coverage ratio (LCR) are adequate and can help reduce risks of lending retrenchment. Capital forbearance seems the correct way to go; hiding losses less so.


In addition, it seems advisable to stop the clock on the timeline for implementing further capital increases under Basel III reforms as well as delay the EBA 2020 stress tests (as it has already decided to do). Building capital buffers is important, but timing is critical.


While all these measures can help mitigate credit retrenchment and thus avoid a deeper recession, losses for banks and reduced equity buffers will still be a concern for markets. While these losses might not show until later this year, financial markets will price them in as soon as more information become available, which will require further forward-looking actions, discussed below.



Instilling confidence


Government is the ultimate backstop for absorbing losses in crisis situations like this one. One of the critical tasks for policymakers is to provide a maximum degree of certainty and instil confidence. As much as this applies to the public policy response to the health crisis caused by the virus, it also applies to the financial sector.  Communication is critical in these circumstances. We saw both good and bad examples last week, with Christine Lagarde’s remarks during her 12 March press conference that the ECB was not in the business of reducing spreads causing immediate negative financial market reactions as well as positive reaction to the correction of these remarks by her subsequent interview and the blog by ECB’s chief economist Philipp Lane the following day.  Standing ready as lender and market-maker of last resort and providing clear signals that central banks will stand ready to avoid any price overshooting and market freezes is critical in such situations.


Planning ahead


Assuming projections on how the virus will play out turn out to be accurate, the attention will turn to economic recovery later this spring and early summer. This will also be the moment, when there will be a clearer picture of the losses in the financial system and policy actions will be needed. The regulatory reforms of the past ten years, including bank resolution frameworks, will then be put to their first massive test.  


As many economists, I have pointed to missing elements in the banking union: a common deposit insurance scheme and a limited funding backstop for the SRF. While the former has still not been addressed, there is progress on the latter, but probably not in time for the current crisis.  Another misstep in my opinion was that a new regulatory framework was being implemented without the legacy of the crisis being addressed first. The Italian approach to bank failures in recent years can be easily explained with this legacy: bailinable debt that was politically not bailinable and legacy assets that had not been resolved yet resulting in taxpayer support where such support was – at least in principle – not to happen under the BRRD. A final item on the unfinished agenda has been to cut the link between banks and governments in the eurozone, which would require concentration limits for sovereign bonds and the creation of the safe assets. Obviously, these are not reforms that can be introduced in time for a possible crisis later this year, but one can hope that a crisis might be the trigger for finally completing the banking union.  


It is widely accepted that the bank resolution regimes established under the BRRD across the EU and on the euro-area level are not adequate for a systemic banking crisis.  If we see widespread bank undercapitalisation if not failures, a flexible and systematic approach is needed, and on the Eurozone rather than national level.  


We might easily come into such a situation later this year when authorities in the eurozone are faced with several mid-sized if not large banks showing significant undercapitalisation. At a time when the real economy will try to get back on its feet and will have to rely on bank lending for this, widespread recapitalisation efforts with taxpayer support might be the only option.  Applying an approach built for idiosyncratic bank failures – including bail-ins and liquidation -, on the other hand, might deepen the crisis further.


Some countries, most prominently Italy, might not have the fiscal policy space for such a recapitalisation of banks and might also face limits in direct support for the real sector. This is where a eurozone response is required. While the ESM has the option (so far not used) to directly recapitalise banks, the total amount is limited to 60 billion euros. An alternative option, more adequate for a systemic crisis situation and suggested by several economists, is to establish a eurozone-wide bank restructuring agency (e.g., Beck and Trebesch, 2013).  Such a temporary agency could be in charge of restructuring viable and non-viable banks throughout the eurozone, funded by the ESM and possibly leveraged with private sector funding. Such an ad-hoc approach would also signal that these are special circumstances and that the eurozone is not falling back into the bailout mode.



The moment for European solidarity


There has been less coordination across EU countries in public health responses than optimal.  Many countries have also taken decisive actions to address the economic disruptions caused by the virus-induced social shutdown, while there are tentative efforts at the EU level to follow suit. However, the fiscal firepower of such efforts on the EU level is simply not enough. A longer recession and increased fiscal policy demands might turn some member countries’ fiscal position unsustainable, further undermining their banking sector, and deepening the crisis further.


For the eurozone to survive such risks, a eurozone- or EU-level approach is needed. This is the moment of European solidarity, required on the highest political level, i.e., the European Council (EU heads of governments). As wrong as president Lagarde’s remarks were during her 12 March conference (“we are not in the business of reducing spreads”), as clear is it that it is fiscal policy that has to finally come to the table.  Monetary policy has carried for too long the burden. Yes, the ECB will again step forward when needed, but it is clear that this is politically not a long-term crisis resolution strategy.


It is time for a fiscal policy “whatever it takes” moment in the euro area. During the eurozone crisis, there was often talk of moral hazard risks coming with “bail-outs” of peripheral countries of the euro area. As wrong as these criticisms were then, there is no basis whatsoever for such concerns now, as this is an exogenous shock.  It is important to stress that it is not enough for the European Commission to loosen budget guidelines for national governments, as financial markets will still price in countries’ debt sustainability. Now is the time to put the money where the words are – a post-pandemic eurozone budget where fiscally stronger countries help fiscally weaker and more affected countries. Addressing a possible systemic banking crisis jointly can be part of such an effort.


And even though events are still unfolding, and the final tab has not been passed on, stating upfront and clearly that eurozone governments stand together and will fiscally do whatever it takes, can be incredibly useful!  


In the evening of Thursday, 27 February, I received an email from VoxEU editor-in-chief Richard Baldwin whether I could write a short piece on the possible effects of COVID-19 on the financial sector; deadline: the following Monday. This morning, the eBook was released – a collection of 14 chapters by economists around the globe, with chapter on macroeconomics, trade, policy cooperation and finance.  My chapter (Finance in times of COVID-19) is not trying to predict the impact of the virus on the financial sector, but rather offer some ideas on how to interpret what might happen during the next months. Obviously, the effect of the virus on the financial system will depend on (1) how much further the virus will spread across the globe and its effect on economic activity, (2) fiscal and monetary policy reactions to the shock, and (3) regulatory reactions to possible bank fragility. Current economic scenarios range from a small growth dip over a recession in several affected countries to a global recession as in 2008/9. While there is less monetary policy space today than during the Great Recession, bank regulatory and resolution frameworks certainly offer more policy options than 12 years ago, though the question is whether they are really fit to deal with a systemic crisis. I am writing all this, recognising that there are much more urgent and immediate public policy questions related to containing the spread of the virus and the associated socioeconomic damage.


One big factor will be whether virus-related disruptions will be temporary or persistent. As important as this is for the economic damage done by the virus shock (a V-shaped dip and recovery or a deeper U-shaped recession), this will have repercussions for the financial system. In the case of a temporary disruption to supply chains or a mild demand-side shock resulting in a delay in consumption, banks can serve as support for struggling firms, especially in the case of many European banking systems with close and long-running relationships between firms and banks. Recent research (including my own work with Hans, Ralph and Neeltje) has shown that relationship lenders can help firms during times of recession and economic crisis, and based on their extensive knowledge of firms and long-run relationships. A longer slowdown or even a recession, on the other hand, will put pressure on banks’ loan portfolios and solvency positions. Rather than the recent correction of stock markets across the globe, it will be non-performing loans (as well as a freezing of funding markets) that could be a direct source of bank fragility. Non-performing loans, however, will not show up immediately, but rather (in a negative to adverse scenario) in the second half of 2020.


One starting point to assess the impact of such a negative or adverse scenario are stress tests undertaken by regulators across the globe, including by the Single Supervisory Mechanism (SSM) and European Banking Authority (EBA) for the largest banks in the euro area and EU. The 2018 stress test modelled a cumulative fall of 8.3% over three years relative to the baseline projection in its adverse scenario and concluded that even after such a shock, the average CET1 ratio would still be 10.1%, though with a large variation across banks. Obviously, there might be quite some variation in such an adverse scenario across countries and banks, and there certainly could be bank failures, especially among banks whose loan portfolios are concentrated in the areas most affected.


Regulatory forbearance with respect to loan classification and thus loan loss provisions would be the wrong response. Letting markets guess what the true financial situation of banks is rather than providing such information can make things only worse. While there is an ongoing academic debate on whether more transparency is always better, experience from the early EBA stress tests in the EU – which turned out to be too lenient, with banks that passed the test failing shortly afterwards – suggests that pretending that things are just fine is not conducive to creating confidence. Rather than allowing forbearance on loan classification and thus loan provisions, regulators should instead allow banks to eat into their capital conservation and counter-cyclical buffers. Such loan losses would not show up that quickly anyway and consequent losses would not be expected before late 2020. At the same time, bank resolution frameworks might be put to the test, as will the political willingness to let supervisors and resolution authorities do their job.


Loan losses are only one source of fragility, though. Last October, the Hong Kong Monetary Authority ran a crisis simulation exercise with its major banks, which included the breakout of a disease like the Coronavirus, with the resulting operational challenges. Operational risks can loom large in scenarios with widespread socioeconomic disruption, and the better prepared central banks, regulators and financial market participants are, the more limited the damage to the financial system and the real economy will be.


A third challenge (though related to the previous two) would be the loss of confidence in banks, be it by depositors (resulting in bank runs) or by markets. Loss of access to funding markets can easily turn into systemic distress, and much earlier than non-performing assets will show up on banks’ balance sheets. Swift intervention by central banks as lenders and market-makers of last resort will be critical in such circumstances.


What will be the policy reaction of monetary and fiscal policy authorities? In the euro area, the ECB has all but run out of munition, unlike the Federal Reserve and the Bank of England – with the former having already taken action this week. While there might be still be options to influence the yield curve, large aggregate demand effects cannot be expected from such actions. Lowering already negative interest rates further might trade off aggregate demand effects with putting further pressure on banks’ balance sheets.


Fiscal policy, on the other hand, has quite some space, especially in some of the ‘frugal countries’ such as Germany. Italy has just announced temporary tax cuts and higher health spending, with an obvious negative effect on its fiscal position. This seems the most reasonable approach right now, though it certainly might lead to problems further down the road in terms of Italian debt sustainability. The Italian government has requested that the European Commission relax the fiscal policy targets for Italy in light of both expected growth and a higher deficit resulting from COVID-19.  However, it seems to matter little if the Commission loosens fiscal criteria for the Italian government, as it will ultimately be the market that will take a view on whether or not Italian sovereign debt is sustainable. In a perfect storm, an increase in Italian government bond yields together with rising loan losses could put Italian banks under pressure. While this might seem like a tail risk at this stage, it certainly should not be excluded. Policy responses to such an event would certainly fall outside regular frameworks. They might require a new ‘whatever it takes’, a restart of the Outright Monetary Transactions (OMT) programme (announced by Draghi in summer 2011, but never used) and a coordinated effort at the euro area level.


Which brings me to a final point. COVID-19 is a typical example of a shock that is hard for each country to handle separately. It is a challenge for which ‘Europe’ seems an appropriate level to coordinate action (notwithstanding urgently needed global coordination). Beyond handling the challenges for the health system, economic and financial policy coordination is critical for the EU and the euro area in the case of an adverse scenario. It might very well turn into another historic test for the EU and the euro area, in terms of economic policy response but also in terms of political significance and sending a signal of relevance and strength to its citizens.


In summary, in the most adverse scenario, COVID-19 could have quite important repercussions for the financial system. Immediate attention should obviously be focused on the public health aspects of the virus and on avoiding a global pandemic, if still possible. Adverse solvency effects in the financial system will most likely not be immediate, so appropriate responses can be prepared. Panic and spillover effects in markets, on the other hand, might come much more rapidly, so complacency over short-term effects might be mistaken as well.  The experiences and regulatory upgrades of the past decade will come in useful for regulatory and monetary policy authorities; however, there seems to have been as little preparation as before for a tail-risk event. If I had to make one humble recommendation to regulatory authorities, it would be to (1) focus on possible operational disruptions in the financial system, (2) strengthen confidence in financial markets by clearly signalling that they stand ready to intervene, and (3) prepare for possible interventions in and resolution of failing banks, without ignoring tail-risk events.

I am just returning from a technical workshop in Nairobi, where central banks staff from across Africa assembled to discuss the second data collection round for the long-term finance scoreboard that we launched last year.  While many of the discussions were technical, there were also more fundamental conversations on the role of data in the policymaking and, ultimately, development process. In my presentation, I pointed to the important opportunities that better data can afford research and ultimately policy making, but also certain pitfalls.  The empirical finance and growth literature (as started by Ross Levine and Bob King 30 years ago) relied initially on broad cross-country data from the International Financial Statistics, but quickly moved to industry and firm-level data to explore not only the question of causality, but also the channels and mechanisms through which finance affects growth. But it is important to note that measures of financial development (such as Private Credit to GDP) are not policy variables.  


A more recent example for the power of data has been the field of financial inclusion. In 2005, we had all but anecdotal evidence on the limited access to formal financial services across the globe. Initial data collection efforts on branch/ATM penetration and the number of deposit/loan accounts gave first insights and managed to raise the issue of financial inclusion, ultimately resulting in a permanent collection effort of these data by the IMF (Financial Access Survey). But it was not until the first wave of the Global Findex survey in 2011 that we got detailed data on the share of adult population with access to a formal bank account. Having better data allowing to compare financial inclusion across countries and within countries over time can provide an important reform impetus and allows to establish targets for policy reforms, such as the Universal Financial Access Goal, formulated by the World Bank Group in 2013, that “by 2020, adults, who currently aren't part of the formal financial system, have access to a transaction account to store money, send and receive payments as the basic building block to manage their financial lives.” We have seen many country-specific reform efforts and inclusion targets, based on the now-available detailed data.


However, there is also an important pitfall that I would like to stress, which links to Goodhart’s Law: when an indicator becomes a policy target, it ceases to be a good measure. Providing everyone with an account is one challenge, having people use these accounts for their benefit is another. There are many examples of dormant accounts, which sheds doubt on the usefulness of account ownership as headline indicator. On the upside, this insight has led to a shift in attention from ownership to use of an account and exploring the constraints in the use of financial services.


One critical lesson can be learned in this context from the “success” of the Doing Business data collection exercise. Providing data on critical aspect of the business environment faced by enterprises has given impetus to reforms in these areas across the globe. Ranking countries according to their business environment has fuelled a certain degree of competition, but the pitfalls of such rankings have also become clear. Governments might want to become first in their regional class, but rankings are always relative. Will a government go for low-hanging fruits but less consequential reforms simply to increase their ranking? Should laws and regulations really be changed primarily with indicator definitions and rankings in mind? Goodhart’s law might again be cited here – as doing business indicators and country rankings become policy targets, they might become less relevant for measuring actual constraints in the business environment.


Data are thus critical for moving a policy area up in policymakers’ agenda and that is what we are currently trying to do with the long-term finance scoreboard in Africa. The need for more long-term finance (for infrastructure, firms and housing) is recognised but limited information is available on how much is out there and in what form.  Having a firm quantitative basis can serve as basis not just for rigorous analysis and evidence-based policymaking, but also provide impetus for reforms. The pitfall we should avoid is to focus on one headline indicator or a country ranking, which might simply divert attention from the problems at hand.

My paper “Sharing the Pain? Credit Supply and Real Effects of Bank Bail-ins”, with Andre Silva and Samuel Da-Rocha-Lopes, has been accepted for publication in the Review of Financial Studies. This paper gauge the credit supply and real effects of the resolution of Banco Espirito Santo in Portugal in August 2014. The bank was split into a bad bank, sent into liquidation with equity holders and junior debtholders effectively bailed in, and a good bridge bank (Novo Banco) that continued operating. We find that firms exposed to the failed and resolved bank see a reduction in lending by Novo Banco, but can make up for it by borrowing more from other banks they already have relationships with. However, SMEs exposed to the failed and resolved bank experience a reduction in credit lines and those SMEs with low liquidity reserves before the shock reduced investment and employment, while increasing their liquid assets.    In summary, the resolution was not a panacea as there were some real sector effects, but the negative effects were contained. The pain was shared across bailed-in debtholders and some of the banks’ borrowers, but the taxpayer did not have to shoulder the burden, unlike in the case of bailouts.


I have discussed this paper before on my blog and am happy to report that the main message has not changed. But the reviewers and editor pushed us hard to think more carefully about what drives the result that lending by Novo Banco (the “good-bank” remainder of Banco Espirito Santo) went down. Were it the losses in BES? No, as the bridge bank was recapitalised during the resolution to its original level, well above the minimum.  Rather, it was the bail-in as part of a broader restructuring process, replacing management, extensive reorganisation and changes in risk management. This is different from bailouts, where management is not necessarily replaced and there is not necessarily a broader reorganisation.  


A second interesting result from the revised paper is that on the substitutability of cash reserves and credit lines. There is a clear differential effect between firms with high and low cash reserves before the shock. Those with high reserves can draw down their cash reserves thus making up for the loss in credit lines, while those with low reserves have to refill them after suffering a reduction in credit lines.


While this paper is a case study of one specific bank, in a new paper with Deyan Radev and Isabel Schnabel, we look at the effect of bank resolution frameworks across banks in 22 countries. Rather than specific bank failures, we focus on the reaction of banks’ systemic risk contribution after system-wide shocks across countries with different bank resolution frameworks.  This paper is still very much work in progress so I will leave the discussion for another day.

I am on my way back from Barbados – not for an early spring break, but rather an IDB workshop on economic inclusion. It was my first time in this beautiful region in the world (but certainly won’t be the last time). Opening remarks by politicians are typically not the highlights of any workshop, but it was refreshing to listen to the Prime Minister of Barbados delivering competent and data-based remarks on the situation in her country, without the bluster or empty words that we have had to get used to in the US, UK and some European countries.


Not surprising, my presentation was focused on financial sector development and inclusion in the Caribbean – more specifically, on the six countries of the IDB Caribbean country department Bahamas, Barbados, Guyana, Jamaica, Suriname and Trinidad and Tobago.  It is quite a mixed group of high- and middle-income countries, two of which have off-shore financial systems and some of which have natural resource-based economies.  However, there are several common themes: One is the need of an efficient financial system to provide risk mitigation instruments due to natural disasters.  More than in other regions of the world and increasingly so in times of climate change, this function of the financial system is critical and might also explain why in many countries of this region the insurance sectors are relatively well developed.


Another common theme is that of the small size of all six economies and, hence, their financial systems.  Their banking systems are concentrated with few banks, thus reducing competition. Stock markets are large compared to the size and income level of the host economies, but exhibit a very low liquidity – a phenomenon that can be directly linked back to the small scale, as it is especially public capital markets that rely on network externalities. The small scale makes a focus on competition and contestability in the financial system even more important, which includes tapping the potential of non-bank players such as fintech.


The dominance of banking within the broader financial systems in the six countries also limits financing options for SMEs. As I have repeatedly argued, the segment of micro-, small and mid-sized enterprises is a very diverse one; and rather than focusing on the size criterion, the age of enterprises and the character of entrepreneurs as lifestyle or transformational entrepreneurs might be more important. This also implies a diversity of different financial products and players and looking beyond banks, which might not be best positioned to finance, e.g., start-up companies.  It also implies the need to look beyond the borders of each economy and draw on resources and expertise from the global financial system.