Ructions in the banking sector don’t appear to be the first tremors of a systemic financial crisis, but rather a delayed, jolting reaction to the steepest rate increases in 40 years
The US Federal Reserve (Fed) has published its balance sheet every week for almost 200 years. It usually passes unnoticed. This Thursday evening, we were furiously refreshing our screens, eager to read how much money had been drawn from a new borrowing window set up for banks who had run into severe problems. The relief was visceral: not much, in the grand scheme of things, which was what we had expected and hoped would be the case.
A rocky road to recession
Still, that does not mean that all is well. We’re writing this update after three US regional banks have failed, and the market tested the resilience of Credit Suisse, a so-called ‘global systemically important bank’ (or G-SIB), forcing central bankers and regulators to arrange a forced merger with UBS. And this week’s Fed balance sheet also informed us that borrowing at the central bank’s discount window, the traditional source of liquidity available without a punitive interest rate to less stressed banks, surged to an all-time high of £153 billion. As a fraction of commercial banks’ deposits, this is less than the usage we saw during the global financial crisis (and back then the facility came at a higher cost), but it is still a very arresting rise, indicative of many banks turning to the central bank as they continue to deal with deposit flight in the wake of those regional bank failures. The equity market was closed when the data were released, but investors were taking it in their stride the next morning.
With a modicum of calm restored, the key question for all investors is whether these are the first tremors of a systemic financial crisis, or what will prove to be an unfortunate but brief period of illiquidity and volatility that soon passes (akin to the one around the “Trussonomics” debacle) and therefore presents a buying opportunity. The hard work of our economists, strategists, credit, equity and structured product research analysts over the last week suggests that it is likely to be neither. Rather, it is a jolting step on the progressively rocky but ineluctable road to a global recession.
For when central banks raise interest rates so rapidly – and we’ve had the sharpest rise in over 40 years in the US – we expect things to break. For sure, we did not expect one of those things to be a G-SIB (more on this below), but regional bank failures happen in the US all the time – there were 43 of them between 2014 and 2019, when the economy was booming. As rapid as they’ve been, rate hikes take time to feed through into the real economy. This is why we have remained defensive and continued to forecast recession even as some of our peers were revising up their forecast at the start of the year. It takes time for things to break.
We expect more things to break, although we still don’t expect a G-SIB to fail. In fact, we expect them to be beneficiaries of the deposit migration that the Fed’s balance sheet suggests is ongoing. To be clear, there is more value evident in financial markets today than there has been for years. As investors, we’re excited. But current market pricing is still at odds with the probability we place on a global recession (as well as sticky inflation). And this downward skew to the risks keeps us defensive. As we wrote in our end of year update, we have a checklist for what we need to see to turn more bullish, and we’ll provide a fuller update in our soon-to-be-published end of quarter update. We hope we’ll be checking it off before the year end, but we’re not there yet, as the last week has reminded us.
“When central banks raise interest rates so rapidly – and we’ve had the sharpest rise in over 40 years in the US – we expect things to break.”
The banking system isn’t broken
Bank failures are usually the result of wayward lending to jeopardous borrowers. Broad-based financial crises are almost always preceded by a period of deregulation, or innovation without regulation that encourages such speculative lending. This results in very sharp increase in aggregate credit relative to the size of the overall economy. But that is not what these bank failures were about.
The most prominent, Silicon Valley Bank, was due to basic mismanagement of its assets relative to its liabilities. Its deposits had tripled in two years as a venture capital boom flooded America’s tech hub with money. These deposits were highly concentrated among like-minded corporations, prone to herd behaviour. It had a tiny percentage of retail deposits compared to your average bank. Instead of using these deposits to make loans, it largely bought long-term government bonds. But it didn’t hedge against the rising interest rate environment, like most banks would. The bonds decreased in value as interest rates rose, which meant it didn’t have enough assets to meet its liabilities when depositors wanted their money back. This they started doing en masse when the venture capital boom came to an end, especially when the close-knit community of depositors figured out what had gone on.
A fourth bank, First Republic, was under pressure, but was given an injection of deposits from a consortium of blue-chip Wall Street banks. A private resolution to a liquidity run is very rare (in fact it may be the first time this has happened since the First World War). We see this as a sign of confidence in the solvency of the system.
The large, systemically important banks run extremely different models. They must mark more of their assets to market (i.e. report what they are currently worth if they had to be sold today), so the potential losses become more rapidly apparent. They have much more diverse deposit bases, with a much greater percentage of stickier retail depositors (most of whom are insured against losses by the government, so less likely to suddenly demand their money back). They hedge their interest rate exposures to prevent a mismatch in their assets and liabilities. Finally, they do not have large exposures to the cryptocurrency community, which is what brought down the two other US banks we spoke about.
Moreover, post-2008 reforms mean that the G-SIBs have much stronger capital and liquidity buffers: capital adequacy ratios (a measure of available capital to absorb potential losses) are 50% higher on average and the range of banks' ratios is much smaller too, while in 2008 many had capital adequacy ratios well below the average. In fact, European banks have capital ratios significantly above even these stringent regulatory requirements. This means that their assets would need to suffer much, much larger losses in order to become insolvent. This seems unlikely, given the absence of a profligate lending boom. In fact, private-sector debt is a much smaller share of GDP compared to 2008 levels.
Banks’ liquidity positions (cash available) are stress-tested every year. In Europe, banks hold €3 trillion in central bank deposits, worth 22% of their customer deposits, which makes runs highly unlikely. More technical reforms have also seen a larger share of transactions between banks now run through centrally cleared exchanges (which prevents problems arising simply because of temporary panic), and there’s more robustness to critical market infrastructure like central counterparties.
But wasn’t SVB the second largest bank failure of all time? That sounds like it was a large bank. How did this happen if the regulatory regime was so robust, you may well ask. This statistic is based on deposits and ignores the fact that economies and therefore aggregate deposits grow exponentially. As a percentage of GDP, it was the US’s eighth largest failure, according to our analysis, and a fraction of the top three. Also, remember that SVB had made relatively few loans, which is more important in determining the economic impact of its demise. Still, in terms of deposits, it was the 16th largest US bank. The crucial point here is that because just two years ago these deposits were two-thirds lower, it was still regulated like a small bank. It was not of ‘systemic importance’.
The same cannot be said for Credit Suisse of course. Credit Suisse is one of the 33 G-SIBs. It is a poorly run bank. Compared to the 42 banks in the Stoxx Europe equity index and the 16 in America’s S&P 500, it was the least profitable, the only one where the average return on equity over the last five years – a key measure of success in the sector – was negative. This contrasts to a sector, and our preferred stocks in particular, that have demonstrated improving returns overall.
But in the run up to the last week, Credit Suisse was very solvent, and even very liquid, despite huge outflows of deposits in its private banking arm last year - which were well known, again because they were unique among the global banks. It raised more capital last year, not because it was insolvent, but because it needed to spend more money to invest in trying to restructure the business – to turn that ship around. As a result, Credit Suisse had liquidity and capital adequacy ratios well above regulatory requirements.
So what went wrong? It started with Bloomberg reporting a comment from the bank’s largest shareholder, effectively the Saudi Arabian state, that it would not increase its stake, as this would mean breaching a limit of 10% that would have regulatory implications. This was widely misreported, missing the crucial context that the Saudis couldn’t increase their stake for regulatory reasons and, in any case, Credit Suisse was not asking for more capital because it did not need it.
But investors were already jittery after the US failures and had long grown frustrated by Credit Suisse’s poor performance. They revolted anyway. But because Credit Suisse was meeting its regulatory requirements, the Swiss National Bank was able to step in swiftly to shore up confidence by making 50 billion Swiss francs of funding available. This brought Credit Suisse’s much-watched liquidity coverage ratio to a huge 190%, although we expected further deposit outflows to erode this somewhat. Investors were still fearful, however, and continued to sell Credit Suisse’s equity and debt. The Swiss government therefore forced UBS, another large Swiss bank, to purchase Credit Suisse, after offering some indemnity against any losses that may come to light subsequently and extending a very large line of low-cost funding.
Again, this relatively orderly merger (these things are always messy) was possible because Credit Suisse was meeting its regulatory requirements. Regulation can feel cumbersome and inhibiting in the good times, but it comes with huge benefits in the bad.
Nevertheless, this development has clearly challenged our view that a global financial crisis is very unlikely. We still believe that, but we must acknowledge that it is a possibility. Bank runs are, to use the academic term, coordination problems. And because it is human beings that are trying to coordinate, irrational fears can take hold. As investors we can become so fixated on analysing quantifiable risks that we can forget that we are also, on occasion, at the mercy of unpredictable psychological phenomena.
We are monitoring closely for signs that fear is taking further hold. We’re looking at indicators that can give us clues that deposit outflows might be becoming more widespread. We’re looking at various measures of financial stress for signs of contagion into the global system. There are some incipient strains, of course, but across the board, these indicators do not look like they did in 2008. And importantly, they are falling back at the time of writing.
“This relatively orderly merger (these things are always messy) was possible because Credit Suisse was meeting its regulatory requirements. Regulation can feel cumbersome and inhibiting in the good times, but it comes with huge benefits in the bad.”
Another reason to be cautious
As we said at the start, the developments of the past week confirm our cautious outlook. Despite the resolution to Credit Suisse’s woes, we expect all banks to become more cautious and for lending standards across the largest advanced economies to tighten further, raising the (already high) risk of recession. Banks in both the US and Europe have already been tightening their lending standards since late last year. With investor scrutiny on them intensifying given the failure of SVB and the pressure on Credit Suisse, it seems highly likely that they’ll become even more cautious in their lending. Already, before this episode, we saw a sharp drop last week in the Euro Area Sentix investor confidence index. Next week, the ZEW survey of confidence will likely follow suit. Such business and investor confidence surveys are usually a good leading indicator for turning points in the economic trajectory and could suggest an end to the nascent recovery in confidence.
Investors have aggressively changed their expectations for future interest rate increases in Europe and the US. They now expect significant rate cuts to start imminently. We're wary of this assumption, given the background of persistent inflationary pressure. We’ll discuss this more in our end of quarter update. Bank lending is of course one way that monetary policy makes itself felt in the economy, and if bank lending standards are likely to be even tighter, then monetary policy won’t need to work so hard. It’s right that rate expectations have come down, but we think the large cuts priced in are an overreaction, assuming that a banking crisis is averted. The European Central Bank (ECB) met on Thursday and did a good job of threading the communications needle: it acknowledged that it was taking the liquidity problems seriously and had many tools to deal with it (tools that were not there initially in 2008), without suggesting that the problems were systemically serious. The ECB went ahead with a 0.5% rate increase. ECB President Lagarde was adamant that there is not now a trade-off between financial stability and fighting inflation. As we’ve argued before, markets are not priced for even mild recessions. That’s clear from consensus earnings forecasts. You therefore don’t have to believe that globally systemic banking problems are imminent to be positioned defensively.
Hope for the future
There’s opportunity, for sure. European and US banks are trading on valuations close to where they traded during the pandemic and the financial crisis. The European bank dividend yield is over 7%, again at the top end of its historical range. Of course, this sector does not perform just because it is cheap. In fact, as we go into a recession it’s difficult to see valuations in the sector reverting back up to what the underlying profitability deserves. But if Europe’s banks can demonstrate that they can manage through a downturn successfully, we believe they are likely to be a key area to add to as the cyclical risks start to subside.