Martin Wolf: Banks are designed to fail – and they do

This is a system that is essential to the functioning of the market economy but doesn’t operate by its rules

Banks fail. When they do, those who stand to lose scream for a state rescue. If the threatened costs are big enough, they will succeed. This is how, crisis by crisis, we have created a banking sector that is in theory private, but in practice a ward of the state. The latter in turn attempts to curb the desire of shareholders and management to exploit the safety nets they enjoy. The result is a system that is essential to the functioning of the market economy but does not operate in accordance with its rules. This is a mess.

Money is the stuff one must have if one is to buy the things one needs. This is true for households and businesses, which need to pay suppliers and workers. That is why bank failures are calamities. But banks are not designed to be secure. While their deposit liabilities are supposed to be perfectly safe and liquid, their assets are subject to maturity, credit, interest rate and liquidity risks. They are fair weather institutions. In bad times, they fail, as depositors run for the door.

Over time, state institutions have responded to the inability of banks to provide the safe money their depositors expect. In the 19th century central banks became lenders of last resort, though supposedly at a penalty rate. In the early 20th, governments guaranteed smaller deposits. Then, in the financial crisis of 2007-09, they in effect put their entire balance sheets behind the banks. The banking system as a whole became, unambiguously, a part of the state. In return, capital requirements were raised, liquidity rules were tightened and stress tests were introduced. All then would be well. Or not.

The failure of Silicon Valley Bank shows there are holes in the US regulatory dike. That is no accident. It is what lobbyists called for: get rid of onerous regulations, they cried, and we will deliver miracles of growth. In the case of this bank, what stands out is its reliance on uninsured deposits and its bet on supposedly safe long-duration bonds. At the end of 2022, it had $151.6 billion (€142.5 billion) in uninsured domestic deposits against about $20 billion in insured deposits. It also had substantial unrealised losses on its bond portfolio, as interest rates rose. Put these two things together and a run became likely: rats will always abandon sinking financial ships.


Those who fail to escape in time will scream for a bailout. It may be amusing that those shrieking for a rescue this time have been the libertarians of Silicon Valley. But few people are capitalists when threatened by losing money they regarded as safe and nobody is better than a capitalist at explaining how essential their wealth is to the health of the economy. Uninsured depositors have duly been rescued at SVB and elsewhere. This removes yet another source of private sector discipline on banks.

One point is that it is a good thing if fear has reignited in the financial system. The anxiety created by small shocks makes big crises somewhat less likely

Yet SVB was only the 16th-largest bank in the US. This is, after all, why it had been left out of the regulatory net applied to the most systemically significant banks. It was conveniently non-significant in life, but became systemically significant in death. The Federal Reserve has also offered to lend at par value to banks that need liquidity. These are negative “haircuts” – call them “hair-grafts” – to banks who need emergency loans. Beyond this, president Joe Biden has asserted that “we’ll do whatever is needed”. True, this time shareholders and bondholders are not being bailed out. Moreover, losses will supposedly be borne by the banking industry as a whole. Yet the losses are again partially socialised. Does anybody doubt that socialisation will become deeper if the crisis also does?

Naturally, people wonder what this new shock means. Some analysts believe that the Fed will no longer tighten monetary policy this month. What is clear is that there is much uncertainty, which can justify delay of further tightening. But lowering inflation remains essential: the US consumer price index rose 6 per cent year on year in February.

At present, however, the big issue is not what is going to happen to the economy, but what is going to happen to finance. One point is that it is a good thing if fear has reignited in the financial system. The anxiety created by small shocks makes big crises somewhat less likely. There are additional lessons: banks remain as vulnerable to runs as ever and, like it or not, uninsured depositors will not be wiped out in a failure. Confidence that deposits are safe is just too important, economically and politically.

So, how is this new evidence of the extent to which the state backs the banks, even in relatively normal times, to be reflected in policy? One simple answer is that regulation of systemically significant banks must be extended throughout the system. Another is that deposits must be put above all other debt in an insolvency, to reflect their social and economic importance. Yet another is that balance sheets should always reflect market realities. Finally, capital requirements should be adjusted accordingly. If banks’ capital falls too low, at market valuations, it needs to be increased, promptly.

The fundamental lesson we have to relearn is that even in a modest crisis deposits cannot be sacrificed, and rules on haircuts for provision of liquidity will go out of the window. Banks are wards of the state partly because they are at the heart of the credit system, but even more because their deposit liabilities are so politically important. The marriage of risky and often illiquid assets with liabilities that have to be safe and liquid within undercapitalised, profit-seeking and bonus-paying institutions regulated by politically subservient and often incompetent public sectors is a calamity waiting to happen.

Banking needs radical change. – Copyright The Financial Times Limited 2023