Steve Williams is TCFP’s long standing investment consultant. He sits on our investment committee and helps inform our investment strategy. These are his thoughts on banks….
“To understand what has happened in the US, and to appreciate that some banks are better run than others, you need a basic-level understanding of how banks work. And to that end, what follows is a caricature description – a cartoon bank of sorts – comprising only those features required to illustrate the problems which led to the spectacular collapse of Silicon Valley Bank (SVB)…
A cartoon bank
At the broadest level, Cartoon Bank borrows money from you and me in the form of deposits and lends it to others in the form of loans (commercial loans and residential mortgages for example). The rate of interest the bank offers to depositors, say 1%, is lower than the rate it charges on loans, say 4%. In that case, Cartoon Bank is earning 3% on its operations. That 3%, in banking parlance, is known as the net interest margin. To maximise its net interest margin, the bank wants to borrow capital from us at the lowest viable rate and it wants to loan it out at the highest viable rate. Of course, running a bank is not without its costs and, absent other revenue-generating activities, Cartoon Bank’s profits will fall short of its net interest margin.
Meanwhile, you will notice that there is a problem. Setting aside any fixed-term savings products that the bank might offer, deposits can be withdrawn at any time and in any amount. And not all of that money is sat in the bank’s vault safe from Wile E Coyote, our bank has lent some of it to the Road Runner.
Happily, Cartoon Bank is prudent and it keeps on hand sufficient cash to cover the kind of withdrawals that are typical on any given day. And since anticipated withdrawals are a fraction of total deposits, there is room for some ‘investment’…
Aside from (1) cash held in the vault, the bank can hold (2) reserves at the central bank, make investments in (3) interest-bearing securities which are considered to be free from the risk of default and (4) provide loans. To further our understanding, it is worth looking at each category in a little more detail…
- Cash (notes and coins) held in the bank’s vaults yield our bank nothing at all. Indeed, vaults full of cash are expensive to maintain and insure, so the bank is making a loss on these holdings.
- Cash remitted to the central bank (reserves) is very safe indeed. If Cartoon Bank deposits £10 billion with the Bank of England for example, there is no question at all that the same £10 billion will be available tomorrow. In fact, the deposit will be worth a little more than it was today because the Bank of England pays interest on those reserves in line with Bank Rate.
- ‘Interest-bearing securities which are considered free from default risk’ play an important role for Cartoon Bank. They include government bonds (British gilts, US treasuries etc.) and government-guaranteed bonds (US-agency mortgage-backed securities etc.) which would normally yield a return a little higher than that offered on reserves at the central bank. They have the benefit of being reasonably liquid (they can be quickly and reliably sold, or they can be used as collateral to borrow cash). But while they are free from the risk of default, they are subject to something called ‘interest rate risk’ (more on this later).
- Loans are at the heart of Cartoon Bank’s business. They are the most profitable, since the rates the bank can charge on these are much higher than the yields it can secure on the other three asset types. Commercial loans and mortgages do not come without their risks though, in addition to some interest rate risk the dominant risk for the bank is that a proportion of those loans may not be repaid. In other words, Cartoon Bank will bear default (or credit) risk on its loans.
Note the following return profile for each of those assets…
- Cash (notes and coins) held in the bank’s vaults promises a low negative return for the bank.
- Cash remitted to the central bank (reserves) promises a low positive return.
- ‘Interest-bearing securities which are considered free from default risk’ promise a moderate positive return.
- Loans promise a high positive return.
…and the risk profile for each of those assets…
- Cash (notes and coins) held in the bank’s vaults is free of risk (barring unlawful activity).
- Cash remitted to the central bank (reserves) is entirely free of risk.
- ‘Interest-bearing securities which are considered free from default risk’ present moderate risks (interest rate risk).
- Loans present high risks (interest rate risk and default risk).
The return profile pulls Cartoon Bank away from maintaining a high cash/reserve balance and toward maintaining a high commercial loan/mortgage book, but the risk profile pulls in the opposite direction.
Each of our four categories of ‘investment’ represent assets on Cartoon Bank’s balance sheet while deposits represent liabilities of the bank. Ordinarily, all is well so long as Cartoon Bank’s assets are worth more than its liabilities and can be easily drawn upon in the event that deposit withdrawals exceed ‘expected’ levels.
What are the risks?
The risk, from our perspective as depositors is that Cartoon Bank is unable to one day return all of our money to us. For those Banks which are authorised by the Prudential Regulation Authority in the UK, that is a risk that is to some extent mitigated by a form of deposit insurance which guarantees the full return of up to £85,000 (or up to around £1,000,000 for a term of six months for certain ‘temporary high balances, such as the proceeds from private property sales’). In the main, balances in excess of £85,000 should be considered to be at risk. There is a similar scheme in place in the US, where SVB was based, with a limit of $250,000.
Depositors can further mitigate these risks by limiting deposit sizes to the insured amount and/or undergoing a sufficient level of due diligence.
The risks, from Cartoon Bank’s perspective is that deposit withdrawals exceed expectations and capital cannot be raised quickly enough to meet those withdrawals. To mitigate the risk of unexpected deposit flight, it is wise for the bank to try to attract a diversified deposit base – both in size (with as many deposits under the insured limit as possible) and in customer profile (an ice cream sales company will have a different deposit/withdrawal pattern to an umbrella sales company for example).
It is also wise for banks to protect against default risk on their loans. Default risk cannot be eliminated, but it can be limited with good underwriting practices and, again, some diversification in the profile of loanees. But what did it for SVB was its poor handling of ‘interest rate risk’.
Interest rate risk?
Interest rate risk (sometimes called ‘duration’ risk) is not immediately easy to grasp, but an example will help…
Suppose you have £100 to invest over the next 2 years and that you decide to purchase a 2-year British government-backed interest-bearing security, or what is more commonly known as a gilt. (The equivalent in the US is knowns as a ‘treasury note’ or just a ‘treasury’). In this example, you buy a gilt with an associated rate of interest of 1%, that is to say that you have a £100 claim that pays £1 each year.
Now fast forward one year. You still have a £100 claim paying £1 each year. The difference now though, is that the prevailing rate of interest is 2%. That is to say that investors are able to buy newly issued gilts which pay £2 each year. There is no question about how much your gilt is worth when it matures in 12 months’ time; it will be worth £100. But what if you wanted to sell it to someone else in today’s world where 2% is the prevailing rate. You will be competing with the newly issued gilts that pay 2%. Only a fool would buy your 1%-yielding gilt for £100 when he could buy a similar gilt paying a 2% rate.
There is a way to make your gilt almost as attractive as the new 2% issues though; you could drop your asking price.
In fact, if you drop your asking price to £99 you may well attract a buyer since the new buyer will receive £1 in interest and a further £1 more when the gilt matures for the full £100 in 12 month’s time. That’s the same return as that offered by newly issued 2% gilts.
The reverse is also true, by the way. If the prevailing rate of interest had fallen to 0.5% rather than increased to 2%, you might be able to sell your gilt at the higher price of £100.50. In that example, our new buyer would receive £1 in interest but make a £0.50 loss when the bond matures at £100, thus yielding the new buyer a return of 0.5% in line with the prevailing rate.
Put simply, ‘interest rate risk’ describes the risk that the value of your gilt (or US treasury etc.) will differ from its maturity value (or the price you paid for it) in line with movements in interest rates. That’s a problem for a bank if it has invested, say $120 billion of depositors money in US treasuries when the prevailing rate of interest was very low indeed.
Interest rate risk is one of the more controllable risks for the largest of banks. Whilst complex, and sometimes expensive, there are measures that can be taken to mitigate its ill-effects almost entirely.
How many banks fail?
Now that we understand, at least to some extent, what banks do and how they do it, we are in a position to better understand what went wrong at SVB. Before we take a closer look, I thought it useful to fit some context around what is a very well publicized banking collapse. Specifically, how many banks are there in the US, and how common is it for a bank to fail?
The Federal Deposit Insurance Corporation (or FDIC) is an ‘independent agency created by Congress to maintain stability and public confidence in’ the US banking system. It is the FDIC that provides the insurance guarantee for deposits up to $250,000 and it is also the lead agency, working alongside the Federal Reserve and the Department of the Treasury, tasked with resolving banks which go into (or are likely to go into) receivership. According to the FDIC, a total of 563 US banks have failed in the last 22 years or so. Naturally, the Great Financial Crisis is responsible for the bulk of those failures. I count 507 failed banks between 2008 and 2014, but there were multiple cases in 2002 (11), 2015 (8), 2017 (8), 2019 (4) and 2020 (4).
According to the FDIC’s latest quarterly report (March 2023) there were 4,127 commercial banks and 579 separate ‘savings institutions’ operating under their auspices. Taken together these organisations have total assets worth $23.6 trillion. The total for deposits amounts to $19.2 trillion.
Silicon Valley Bank is widely reported to have been among the largest banks in the US. The Federal Reserve’s most recent ‘Large Commercial Banks’ release (December 2022) supports that assertion, ranking its $209 billion of consolidated assets in 16th place. The full list reveals a somewhat lopsided distribution though. The largest bank is JPMorgan Chase with assets valued at $3.2 trillion, meaning that the largest bank is more than 15 times larger than that ranked 16th. The top 10 banks in the US account for a little over half of the $23.6 trillion market. There is a very large pool of small banks. Indeed, if I’m not mistaken there are something like 2,000 FDIC-registered commercial banks each with consolidated assets of less than $300 million.
Actually, I noticed that the FDIC makes public the number of ‘problem banks’ it has in its sights. The names of those banks are kept secret but currently, there are 39 problem banks – something of a low I understand – with total assets between them amounting to $47.5 billion.
Failures among the US’s large pool of small banks are not common, but neither are they rare.
What actually happened at SVB
I recall reading in the Wall St Journal that SVB was home to around half of all the venture-capital-backed technology and life sciences firms in the US. I took that to mean that there were no ice cream sales companies, and no umbrella sales companies among its customer base; its deposit base was close to being homogenous.
During 2021 and 2022 those deposits increased from around $50 billion to $189 billion, making 2021 the bank’s most profitable year in all of its 40-year existence. I calculate the average deposit size at SVB to be around $5 million, frighteningly higher than the $250,000 insured limit. In other words, given the risk of near total loss, the vast majority of SVB’s depositors were incentivised to remove their capital at the merest hint of trouble.
Worse still, what did the bank do with a large proportion of those deposits? It bought around $120 billion of interest-bearing securities (US treasuries and agency-guaranteed mortgage-backed securities) when interest rates were at historical lows. And it chose not to mitigate the interest rate risk that it was taking on those securities. The subsequent increase in interest rates meant that the market value of those securities fell to around $100 billion.
The rise in interest rates also saw costs increase for the bank’s customers and venture-capital firms began to draw down on those deposits at a rapid pace, forcing SVB to sell some of its securities at a loss of $1.8 billion. It was this loss that came to light in a regulatory filing on 8 March, prompting a noticeable fall in the stock price of SVB’s NASDAQ-listed parent. Shares in SVB Financial Group fell from $268 on the day of the filing to $106 after, prompting the exchange to suspend trading in its stock.
A shift of that magnitude in my bank’s share price would very likely get my attention. And so it was with SVB. It is reported in the Wall St Journal that a staggering $42 billion of deposits were withdrawn on 9 March.
The end came on 10th March, less than 48 hours from the fateful filing, when the FDIC, Federal Reserve and Treasury resolved to close the bank and take control of the bank’s deposits and assets. The speed of SVBs collapse was breathtaking.
What will be the fallout?
You may have heard comparisons drawn between the collapse of SVB and w
hat occurred in 2008, comparisons which seemed to gain momentum during a climactic week of travails at Credit Suisse (now taken over by UBS). I’m not among those drawing a connection, not right now at least. I do expect further failures (or forced consolidations) but I expect they will be limited to the smaller banks. Indeed, the large banks are beneficiaries, hoovering up a significant proportion of the deposits now fleeing smaller banks.
Of course, the difficulties faced by the community banks will have an impact more broadly. Small companies borrow from small banks and the lending facilities offered by those banks add up. The total availability of credit across the US will likely fall as the small banks look to strengthen their balance sheet and avoid anything that might spook their deposit base (rising defaults on their loan book for example). At the same time the smaller banks will have to increase the rate of interest they pay on their deposits and the result will be to narrow their net interest margins.
Less lending from less profitable banks very likely means lower growth. The prospects for a near term recession in the US are increased.”