In recent days, we have been repeatedly assured by institutional figures that the banking system is much better positioned to weather this crisis than it was in 2008. Whilst that would be reassuring, it is unfortunately not true.
To be clear up front, we like banks. Not only have they been our main source of employment for most of our adult lives, but their function is a key driver of growth (compare global economic growth before 1900). As we emerge in some fashion from this crisis, we will need credit expansion to fuel recovery, and that means banks. But simply put, when someone has suffered a potentially traumatic injury and may be in shock, a good friend is not the one who pretends that nothing has happened.
It is of course possible that the reassuring messages we hear from institutional figures are intended purely to shore up public confidence in the system, and do not reflect the true assessment of the level of damage which may be suffered by the financial system. We hope so, and if that is the case, we hope we are not adding fuel to the fire that they are trying to battle. But experience teaches us that in the early stages of crises, denial kicks in, and this is what it currently feels like. Given the unprecedented nature of this crisis, the sooner we recognise reality and the massive level of intervention that is likely to be required to preserve the credit channel capacity for credit expansion and policy transmission, the more the damage will be limited.
An assessment that banks are much better positioned to weather this crisis than the last (at least in terms of credit losses) must rely on some combination of the following theses:
1) The economic crisis is not as bad as 2008.
2) Banks have better asset quality.
3) Banks have better capital buffers.
4) The system is better prepared to support banks.
5) The Covid19 problem will turn out to be brief, with the overall shock passing quickly.
There is a possibility that (5) turns out to be the case. Suggestions in a number of geographies that new cases have peaked, studies postulating that the prevalence of unreported and asymptomatic infection mean that the end of the crisis is closer than expected and reports of the potential availability of fast-tracked vaccinations all leave open this potential outcome. However, public health policy is clearly being formulated on the basis that this will not be the case; whilst we hope for a short, v-shaped recovery, prudent policy in relation to the impact of the crisis on the financial system should not rely upon this hope either. Accordingly, this analysis focuses on the first 4 theses.
1) Crisis Not As Bad
We will not dwell at length on this point, as it was covered in our note of 25 March. Suffice it to say that every indicator points to a far deeper and more extended economic crisis than in 2008.
From a credit viewpoint, this crisis seems likely (absent outcome (5) above) to be of a depth we have not experienced in our working lives, with all the accompanying stresses to asset values, income and borrower performance. Underlying borrower and collateral value performance seems likely to be materially worse than in the 2008 crisis; how much worse remains to be seen.
2) Asset Quality of Banks
In 2008, banks entered the crisis with historically low NPL ratios. As in every crisis, they had learned the lessons of the past and their loan books were well positioned to weather a downturn. At least that was the prevailing view. That same old song seems to be playing now.
The reality is that on the asset side, banking systems across our markets enter this crisis weakened by three factors:
• In many geographies, banks are still heavily burdened by legacy NPL books of the last crisis (€600 billion across EU markets alone), particularly in southern and south-eastern Europe (e.g. Italy at 10% officially acknowledged (1) average NPE ratio, Greece at 40% NPE ratio), although other jurisdictions are not immune. Outside of the EU, rapid NPL formation was already anecdotally apparent in Turkey in 2019 and collapsing real estate values in the Gulf suggested rapid but asyet unreported NPL formation across those markets.
• Many European banks also retain significant indirect exposure to legacy NPLs, whether through financing provided on NPL sales or retained risk on certain NPL securitisations (2).
• New, high-quality loan books: This could be a lengthy debate, but we limit ourselves to the observation that the covenant-lite leveraged loan market reached record volumes in 2019. Closer to our immediate sphere of competence, LTV ratios on real estate lending were driven by valuations at rent levels that were in many cases at or close to historical highs and yields which were at unprecedented lows. These are the same observations one can make about 2007.
On the asset quality front, rather than being well prepared, banks, for the first time in living memory, are entering a second severe crisis with the unresolved legacy of the last crisis still sitting on their books. If this seems unclear, a simple test is to contemplate why Europe’s banks have been consistently trading at an average 50% price-to-book ratio in recent times.
3) Capital Buffers
Of course, the genuinely good news is that capital buffers have increased. However, right now, with average Leverage Ratios at 4.5 – 5.0%, that means banks are 20-25x levered. This buffer does not represent adequate loss capacity in a potentially precipitately falling market.
Somewhat less simplistically, if we focus on CET1 Ratios, these have increased significantly since 2007, somewhere between a 66% and 100% improvement (depending on the impact of definitional changes). Right now, across most of our markets, reported CET1 Ratios are of the order of 15%. However, these reported levels need to be considered in light of the following:
• Absolute (not risk-weighted) asset inflation: Total loan assets across the banking system are up significantly; there is actually more leverage in the banking system than before the last crisis.
• Risk weightings: We should not forget that a residential mortgage at less than 80% LTV is a 35% risk-weighted asset. Many residential markets will already have suffered a hit sufficient to wipe out buyer equity. It seems that certain RWA definitions / calculations have not been built for this market. There is a real danger that RWAs are not currently a good measure of risk.
• Provisioning: Net new Loan Loss Provisions of Europe’s 20 largest banks have recently been below 2007 levels, both in absolute and percentage terms. The buffer represented by LLPs has been depleting, which means a more direct hit to capital from poor loan book performance.
1. We believe that this figure materially understates the true NPL ratio.
2. RWAs on some types of retained risk have increased, but these should be evaluated vis-à-vis probable total loss in a crisisstruck market.
• Asset Quality: Our cautions expressed in (2) Asset Quality of Banks (above), suggest a simple systemic underreporting of potential loan losses, which upon correction would represent a direct hit to reported capital buffers.
Leverage in the system is disturbingly high in the current circumstances. The repeated warnings of certain commentators (3) currently seem at risk of being realised.
4) System Capacity
Certain systemic changes that were implemented during and post the 2008 crisis (e.g. in Europe the SRM) clearly have improved the ability of the system to deal with the emergence of banking crises on a narrow basis (i.e. of single banks, groups of banks or on a single country basis). This is
unfortunately not the reality we currently face.
Rather, we are entering into this crisis, a systemic crisis, at a time when the capacity for critical fiscal and monetary intervention is already vastly depleted. The urgent interventions we have seen from central banks and governments are to be applauded. However, with interest rates already having been close to zero and vast liquidity already pumped into the system, monetary policy options are narrowing. Equally, deficits are not at historically comfortable levels, particularly when adjusted to reflect some of the more obvious impacts of the crisis (just for example, the Italian government is now the economic owner of approximately €50 billion of insured GACS NPL risk).
These systemic factors are compounded by the continued weight on the broader system of those legacy NPLs which no longer sit on the balance sheets of the banks themselves. Of the ca. €450 billion of NPLs which have left the balance sheets of European banks since the crisis, a very large proportion currently remain unresolved, sitting instead with alternative capital providers (fortunately net of significant write-downs). These unresolved loans represent a continued drag on general economic performance (as is perhaps best illustrated by the anaemic nature of recovery in Italy in recent years) and on the overall capacity of the economy to absorb further NPLs.
Perhaps the silver lining is that politicians and regulators seem to have learned from the 2008 crisis. The massive interventions already undertaken signal, for instance, that the deficit concerns which undermined action post-2008 have been re-evaluated. We do expect massive state intervention to continue, although the capacity to continue this will ultimately be constrained by the commencement of inflationary pressures.
Simply put, we see a willingness to pull out the stops on state intervention; unfortunately, monetary and fiscal authorities (across most of our markets) arrive at this juncture like a marathon runner who on crossing the finish line is asked to run 2 more marathons uphill.
Perhaps it is worth adding that since we started to write this note, we have seen UK mortgage lenders substantially close for business, with those still open increasing minimum buyer equity contributions to 40%. The ECB has ordered banks to cease share buybacks and dividend payments. Rating Agencies have begun to systematically cut bank ratings. Limitless in Dubai has announced that it will be entering into restructuring talks. We expect that this negative news is likely to snowball in coming days.
As we stated at the beginning of this note, we like banks. Our comments are not intended to undermine banks, but rather to highlight the issues the system faces and discourage complacency. It is critically important that steps are taken now to ensure the availability of credit at such times as a new normal is established. This will require dramatic decisions, which will only be able to be made if the true potential impact of this shock is accepted and internalised.
In the short term, this means that banks should:
1. Urgently triage loan books in order to clearly identify and categorise borrowers / loans as (a) still viable, (b) viable but requiring some flexibility on timing and interest amounts, (c) potentially viable but requiring material restructuring (including possible debt forgiveness), and (d) not viable. These judgements need to be made using realistic assessments of the potential duration and depth of the crisis and of the process of post-crisis economic recovery.
2. Apply triage findings firmly in determining where to deploy financial and managerial resources. Categories (b) and (c) merit time, effort, funding and flexibility. Category (d) does not.
3. Present to regulators and government a realistic picture of the damage they are suffering. Unlike 2008, the challenges currently faced, and the crisis itself, are not the fault of banks; the financial system is as much a victim of the crisis as airlines, retailers or the hospitality sector. Given the pivotal importance of credit channels and credit expansion to monetary policy transmission, the health of the sector is vital to prospects of recovery. Interventions such as the IFRS 9 delay are to be applauded, but should not be allowed to camouflage the potential need for much more significant, direct intervention.
4. Actively seek early, dramatic intervention. Banks must focus on ensuring their continued viability as providers of credit to the economy and need to fight to avoid 2008-style measures which left the system fundamentally handicapped by the continued prevalence of bad loans as a drag on performance and management. Whilst the full introduction of broad support measures for banks is premature, the potential for a deep, prolonged crisis is clear and the time to prepare the appropriate large-scale contingency mechanisms is now.
5. Push for the creation and implementation of rational NPL resolution structures and mechanisms. Bank balance sheets are not the logical place for holding and managing loans which need, for public policy and equity reasons, to be managed gently. Similarly, half-measures risk creating significant structural impediments to intelligent resolution (4). The most obvious model for impaired portfolios currently seems to be that of the RTC in the US Savings & Loan Crisis. Given the near certainty that public policy will dictate that many non-performing borrowers should be saved, with their loans managed for the social benefit rather than the lender’s benefit, the creation of the appropriate public NPL resolution structures should be accelerated.
Across the globe, there is no doubt that the wave of fresh NPEs is coming. Now is the time to put in place the frameworks which will allow these to be addressed with appropriate sensitivity at the same time as preserving the viability of the policy-critical banking sector. As green shoots emerge, the world will need banks that are banks, not asset management companies. The time to act is now, rather than when the full force of the storm on banks’ loan books is already apparent.
Our thoughts are with all our friends in the banking and regulatory systems in addressing the issues they face.
(1) We believe that this figure materially understates the true NPL ratio.
(2) RWAs on some types of retained risk have increased, but these should be evaluated vis-à-vis probable total loss in a crisisstruck
(3) One of our favourite’s is here: https://www.mdpi.com/1911-8074/12/3/152/htm
(4) See for instance the difficulty of implementing REOCO solutions within the Italian GACS framework.