"Silicon Valley Bank committed one of the most elementary errors in banking: borrowing money in the short term and investing in the long term." This tweet was written on March 13th by Larry Summers about the recent collapse of Silicon Valley Bank (SVB). However, as the former US treasury secretary understands as well as anyone, borrowing short and lending long—so-called maturity transformation—is both the big problem with banking and the whole point of it. It’s a balancing act that SVB got catastrophically wrong.
Banks match the complementary needs of two sets of people. They borrow money from depositors on a short-term basis and lend it to their borrowers for the long term. Banks profit by paying lower interest to their depositors than they charge borrowers. They just have to set enough of the depositors’ aside as cash or in a form they can turn into cash quickly, like short-term securities, in case it’s wanted. If the longer-term investments lose money, either because they default or have to be liquidated quickly for less than their face value, banks need a buffer of shareholder capital—money from people who don’t need to be paid back, essentially—to absorb the losses.
SVB made mistakes with both its assets and its liabilities after a growth spurt. Deposits tripled between the end of 2019 and the early part of 2022. The bank erred by relying too heavily on deposits from firms in the tech sector, and taking too many large deposits: many of its customers deposited sums far larger than the limit of federal insurance, $250,000. These sorts of customers were able to access their accounts at the touch of a smartphone screen, and also communicate with each other in groups on social media—lately about the health of SVB. This made it very easy for them to withdraw their money very quickly.
SVB’s greed on the asset side led it to over-indulge in long-term securities. Their market value fell as the Federal Reserve responded to inflation over the past year by raising interest rates, and markets expected rates to stay higher. This was a risk not all that hard to see, given how low rates had been until recently. The only large movements possible were in an upwards direction.
Had SVB been able to sit on the bonds and redeem them when they matured, pocketing the face value, everything could have been fine (assuming there were no defaults). Instead it faced a leakage of deposits over the past year—albeit one that began slowly—probably caused by a fall in revenues, funding and sentiment across the tech sector. This spelled trouble. It left SVB needing to either raise more capital or liquidate long-term bonds at a large loss. The tech elite’s Whatsapp chats were abuzz with talk of the dilemma SVB faced. An aborted share issue sparked a run on deposits last week, forcing regulators in California to close the bank on March 10th.
SVB might have survived with less avarice and more distance from the tech sector on the deposit side, or even with just a little more luck. It is puzzling that the high-value tech depositors and their venture-capital backers did so little homework on their cash management until the last minute. It is equally baffling that supervisors at the San Francisco Fed did not find the concentration of large deposits all from one sector alarming and act accordingly.
These failures have already prompted some changes to the way banks are policed. More substantial ones may emerge over time. The Fed and the Treasury announced that all deposits in SVB and Signature Bank (which also failed last week) would be protected, “insured” or not. They are so worried about the possibility of wider problems in American banking—a number of small and medium-sized banks have suffered scary falls in share prices while big banks have also taken a hit—that a generous one-year loan facility is now available. It is unlike anything ever offered before.
Banks needing funds can post eligible long-term securities against their face value, even though this may be far above the current market value. If a bank coming to the facility fails, the Fed takes the bonds as collateral. But unlike the bank that posted them, it can afford to hold the securities to maturity and ignore their depressed face value; it is not as if its depositors are going anywhere.
Each financial crisis fuels enthusiasm for tighter bank regulation: stiffer capital-adequacy requirements so banks can weather larger losses, more-invasive supervision and stress- testing. Each period of calm leads to a loosening as memories fade. The same cycle may now play out in banks far smaller than the giants which tottered in the financial crisis of 2007-09, even though SVB was in several respects—not least the homogeneity and perceived economic importance of its depositors—an outlier. Common sense, rather than reams of new rules, could probably prevent such a situation from occurring again.
There may also be a reassessment of how monetary and financial-stability policy interact. In an ideal world, bankers would do the right thing, or be forced to by supervisors and regulation. Monetary policy would involve using interest rates as vigorously as necessary to stabilise inflation and activity in the real economy. But if, owing to incompetence in finance, interest-rate volatility poses a threat to banks, and thus to real economic activity, some of the strain may be taken by either more active use of tax and spending instruments, or by greater tolerance of more volatile inflation.
Is quantitative easing (QE), or bond-buying, something that offers a way out of this trade-off? Not really. Employing QE before financial problems boil over is of dubious effectiveness. And in so far as it does work it does so by manipulating long-term interest rates. So it does not solve the problem of how to set an interest rate that will neither endanger the banks nor jeopardise price stability. Using QE to tackle financial problems that have already boiled over is fraught with risk. Central banks’ balance-sheets are arguably already strained. They have grown enormously during recent crises, when central banks bought assets as a way to ease monetary policy, because they had already cut interest rates as low as they could go.
QE has to be used sparingly to avoid creating political hostility and undermining the credibility of monetary policy. The Bank of England seemed to snuff out a crisis in Britain’s government bonds last autumn with QE. For now the Fed’s balance-sheet offer seems to have worked, too. Both institutions will hope that recourse to such actions is as infrequent as possible.
Tony Yates is a former senior adviser on monetary-policy strategy at the Bank of England and a former professor of economics. He now teaches, consults and writes.