When we look at a simple bank balance sheet today, on one side we see deposits from customers. These form the core stable funding for the bank which allows loans to be made on the other side of the balance sheet. This, however, introduces one of the main issues for the system. Maturity Transformation. Suppose that the bank lends money to a customer for a 25 year mortgage. The bank isn’t able to ask the customer for the money back at short notice. If deposit customers come into the bank to take their money out, the bank could have a problem of not being able to give them their cash back as it has all been lent. This is the maturity transformation problem. Deposits are largely ‘on demand’ whereas loans most certainly are not.
What’s the solution?
The easiest solution comes from the observation that generally, deposits are quite ‘sticky’. Particularly deposits from the general public, known as retail customers. By sticky, we mean that despite technically being available for instant withdrawal, they tend to be left in the bank, on average, for several years. As a result, the amount of cash required to meet the demands of depositors is quite small relative to the overall size of the deposit base. So as long as there remains a float of cash in the bank, long term loans should be absolutely fine. But there’s a problem. If for some reason depositors en-masse lose confidence in the bank, fearing losses for example, the demand for repayment of deposits could exceed the available cash. This is known as ‘a run on the bank’ and we saw such a run back in 2008 with the collapse of Northern Rock, with queues of customers outside their branches.
Following the GFC the regulators across Europe greatly increased liquidity requirements. Cash can be raised in other ways to meet these new requirements. On top of deposits, banks also issue other forms of funding, that themselves are longer term in nature. For example, senior bonds have lifetimes of several years, raising stable cash for the bank. And of course, there is also the bank’s own equity (the money that belongs to shareholders) and also other forms of bonds.
These additional forms of funding can leave the banks with a large surplus of cash, well in excess of the demand for loans. What to do? By and large, European banks leave this cash on deposit at their local central banks. For many this has been a drag on profitability for many years, but of late, as base rates have risen, it has started to pay its way. The money is instantly available for depositors should they need it. The scale of the cash balances is colossal. Take Barclays as an example. Across the bank, in December 2022, Barclays had loans to customers of £398.8bn. It also had deposits from customers of £545.8bn, a surplus of £147bn. That surplus alone represents 26.9% of the entire deposit base. But on top, as a result of other funding put in place by Barclays, it had a liquidity pool of £318bn in total. Of this, its balances at central banks total £248bn. Its central bank cash balances, on their own, could repay 45% of the entire deposit base.
So this brings us on to Silicon Valley Bank (SVB), and what went wrong.
To state the very obvious, SVB isn’t a European bank. It isn’t regulated in the same way as European banks. It is classed as a small bank in the USA, which allows it lighter touch regulation. It is also a bank authorised by State Charter, which again means that its regulation is not performed by the main federal agencies. It is also almost a monoline bank, by which we mean its customer base is composed of pretty much one industry – tech.
By focusing on start-up and early stage tech businesses, SVB had an unusual balance sheet. Tech businesses tend to ‘burn’ cash, that is they spend more cash than they earn. So over time their deposit balances decline as the money is spent, and then climb again with each round of capital raising from tech investors. Their demand for loans is relatively low. On the SVB balance sheet deposits totalled $173.1bn at the end of December 2022 compared with loans of $73.6bn. Rather than holding the excess in cash, as Barclays does, SVB placed their bet and invested in long term US treasuries in order to earn as much interest as possible. Using a small bank exemption, available in the US, they avoided the need to mark their balance sheet to market. In normal times, this is almost irrelevant, and simply removes small volatility from the capital ratios. But, at the moment, with interest rates that have risen very sharply, the value of the long term bond portfolio has fallen sharply. Estimates are that the decline in value is of the order of $15bn comparable to the entire equity of the bank. In short, were the bonds to be marked to their market value, SVB would be bust. With an almost inevitability, word leaked and deposits started to run. Corporate deposits, using the term above, are much less ‘sticky’ than retail deposits.
With so little real cash available, SVB was pushed into the position of selling the bonds and crystallizing real losses in order to repay the depositors. At this stage regulators stepped in and seized the bank. It would appear that the resolution process will cost its shareholders their entire investment, and possibly some of the bank’s bondholders too.
The memories of the GFC linger in investors’ minds. The right thing to have done then was to sell first and ask questions later. And that appears to be happening now. Coupled with the very strong performance of the bank sector of late, there were still profits to be taken.
But from our viewpoint across the pond, this is an idiosyncratic failure. Silicon Valley Bank had a very undiversified customer base, with corporate depositors that had constant demand for cash. The decision to lock those deposits up in long duration bonds, totally unhedged against movements in interest rates, was dangerous and ultimately fatal.
But the read across to other banks, particularly in Europe, is not there. SVB failed, having placed a very large bet, all on long dated bonds. That bet lost. Our banks are regulated very tightly. They have huge cash balances, available on demand at their central banks. By contrast SVB held 57% of its entire balance sheet in bonds. SVB also had a depositor base of customers that needed regular cash flows. In contrast again, deposits have been growing in Europe rather than shrinking.
So while it isn’t possible to rule out other failures for similar reasons in the US, and indeed we have seen some other small banks being ‘resolved’ (the new term for liquidated) over the weekend, the good news is that the problem is not credit quality. Customers are not defaulting on their loans. This is a liquidity problem brought about by the very sharp rise in interest rates and the knock on effects on bond prices.
The implications are as yet unclear. European banks are in rude health. But it is entirely plausible that we could see a movement, at the margin, of deposits from smaller banks, perhaps perceived as less secure, towards the large incumbents, simply as a precautionary measure. This might have the effect of raising the cost of deposits for those smaller banks, as an incentive for their customers to stick with them, which in turn will reduce their profitability. But these are marginal effects on profitability, not on existence.
There has also been regulatory response in the US already to stem the contagion. The Fed has introduced a new term funding scheme which allows banks to borrow against their bond holdings at the par value of the bonds rather than the current, somewhat lower, value. The idea being that there will be no need to sell the bonds and crystallize losses.
So, once again, we live in interesting times. There is some churn in shareholders and some increase in yields in the bond markets for banks, but this is understandable after the period of optimism seen in the last few months. What is clear is that the European banks are in a position that is unrecognizable when compared to how they entered the GFC. Regulations have done their job. Across the pond, however, the lighter touch of regulators on smaller banks might need some remedial work done.