A good starting point for the global growth outlook is the IMF’s April World Economic Outlook (WEO), which sees global GDP contracting by 3% in 2020, followed by a 4.8% bounce back in 2021 as its baseline. For this case, the IMF’s Global Financial Stability Report (GFSR) suggests that the scenario for banks in developed markets is considerably worse than the stress test the financial sector assessment program (FSAP) envisages. But encouragingly, this is only the case for year one of the shock. For year two, on a cumulative basis, the shock to banks is less than the FSAP assumptions, not only for the WEO’s baseline, but even for its adverse case. For emerging markets, the shock is slightly smaller than the FSAP stress test even in year one. However, there are downside risks to this assessment. Currently data is tracking more closely to the adverse WEO scenario than the baseline. Under the adverse scenario the shock for emerging market banks is also worse than the FSAP envisions, but again only in year one. Overall, solvency risks appear much reduced compared to 2008, but uncertainty is high.
However, the relevant question is not just the size of the shock, but whether the banks are positioned to cope with it. Luckily, the global banking system is in meaningfully better shape than it was going into the 2008 crisis. There are two important reasons for this. First, 2008 occurred after a long boom in lending. In the US, bank credit in the five years leading up to the crisis rose by more than 10% per annum. During the most recent five-year period, lending rose by only 7.6% per annum. The difference is even starker for emerging markets: loans in EM recently grew by less than 9% p.a., while into the 2008 crisis loans grew by more than 20%. The extent of the loan growth going into a crisis is typically a good predictor for how large the loan losses are going to be, as exuberance often leads to poor lending decisions. The lower vulnerability due to lower loan growth is an unexpected upside from the more subdued growth environment post GFC, especially in emerging markets. Second, and largely as a result of more stringent regulatory requirements post GFC, the banks are much better capitalized than they were in the 2008 crisis. The GSFR notes that for the 29 countries with the most important banking systems, 25 have higher tier 1 capital ratios in 2019 than in 2007, and most of them are significantly higher. For EM banks overall the Tier 1 ratio is now close to 14%, up from around 10% in 2007. This increased strength is also visible in reserves. For EM banks, loan loss reserves as a percentage of NPLs in 2019 were almost twice as large as in 2007. Furthermore, EM regulators made sure that open USD exposures by local corporates and banks were sharply curtailed given that FX weakness in 2008 created major credit problems in several emerging markets. Some of the homework doled out to the banks by their regulators post GFC is now coming in handy.
Given the high uncertainty about the size of the shock, it is interesting to see how markets assess the situation relative to 2008. Encouragingly, the market agrees that 2008 was the larger problem. Bank stocks in developed markets sold off on average by around 75% during the GFC, while this time it is closer to 55%. In EM, bank stocks were around 73% lower during the GFC, while this time around it is less than 50%, even though they were the second worst performing industry group (after energy). Maybe more importantly than the equity performance is the performance of bank credit, especially if we are worried about insolvency risk. Here the difference to 2008 is even clearer. Bank spreads of US IG issuers peaked at 680bp in early 2009 compared to a peak of 390bp in March of 2020. Lastly, the behavior of funding markets is also key, as funding stress accelerates any underlying problems. In this area, the current crisis is of a magnitude smaller for banks than it was during the GFC. The 3m Libor-OIS spread in the US peaked in October of 2008 at 360bp, while the COVID-19 peak was in March of 2020 at around 140bp – and is currently back to just 33bp, a level not reached in the GFC until July of 2009, well after the crisis was firmly in the rearview mirror. The spillover of US funding fears to other countries through the cross-currency basis was also more benign. For EUR, for example, stress levels in March 2020 were around 80bp compared to close to 200bp during the worst days of 2008.
Of course, financial markets do not only reflect the size of the shock (still uncertain) and the initial position of the banking system (much stronger), but also the extent of policy support.