What really went wrong at Silicon Valley Bank

Shed no tears for investors in Silicon Valley Bank (svb). On March 10th the bank, which had $212bn of assets, failed with spectacular speed, making it the biggest lender to collapse since the global financial crisis of 2007-09. Most of SVB’s depositors were Bay Area tech startups with accounts holding well in excess of the $250,000 that is insured by the federal government. They had fled and their panic was rational. By loading up on long-term bonds, SVB had taken an enormous unhedged bet on interest rates staying low. That bet went wrong, leaving the bank insolvent (or near enough). The fact that shareholders have been wiped out and bondholders will take big losses is not a failure of the financial system. A bad business has been allowed to go bust.

It is what happened next that reveals the flaws in America’s banking architecture. SVB probably had enough assets for depositors to have got all or almost all of their money back—but only after a long wait. This left many tech firms facing life in a financial deep-freeze; Roku, a streaming giant, had nearly $500m tied up in SVB. Across the technology sector, lay-offs and bankruptcies loomed. And America’s regulators and government seemed to fear that depositors were losing faith in other banks, too. On March 12th they judged SVB too big to fail and guaranteed all the bank’s deposits. If the sale of its assets does not cover the costs of the depositor bail-out, a fund that is financed by all banks will have to chip in, penalising the whole industry for the recklessness of a single institution.

At the same time regulators have had to contend with the threat that other banks might also face runs. At the end of 2022 there were $620bn of unrealised securities losses on banks’ books. On March 12th regulators also shut down Signature Bank, another midsized lender—the third bank to fail in a week, given that Silvergate, an institution heavily exposed to cryptocurrency, collapsed on March 8th. And the fallout in the markets continues. As we published this leader on March 13th, bank stocks were continuing to plunge. Those of First Republic, a bank of comparable size to svb, were down by more than 60% on the day.

To shore up other banks the Fed is offering them support on strikingly generous terms. A new programme stands ready to make loans secured against long-term Treasuries and mortgage-backed securities, like those on which SVB had gorged. Usually a central bank making loans would impose a haircut on the market value of the securities being offered as collateral. By contrast the Fed will offer loans up to the face value of the securities, which, for long-term bonds, can be more than 50% above the market value. The haircut-in-reverse guarantees that another bank with a bond portfolio like SVB’s would have ample access to cash to pay depositors.

The deposit guarantee was inevitable, given SVB’s size (and in any case may be fully covered by SVB’s assets). The same cannot be said for the generosity of the system-wide liquidity support, which is a dramatic expansion of the Fed’s toolkit. Banks’ falling share prices in part reflect investors waking up to the risks long-term bond holdings pose to profitability. But whereas svb’s unrealised losses were enough to roughly wipe out its capital, other banks look solvent with room to spare.

It is right that the Fed lends against good collateral to stop runs. But doing so on such benevolent terms is unnecessary, and subsidises banks’ shareholders. And though the Fed’s backstopping of the system will probably avert a banking meltdown, policymakers should never have got to a point where such extraordinary interventions were needed.

SVB’s failure was so chaotic in part because it was exempt from too many rules designed to avert improvised bank rescues of the sort that the Fed has just engineered. After the financial crisis, America’s Dodd-Frank Act required banks with more than $50bn in assets to follow a panoply of new rules, including creating a plan for their own orderly resolution if they fail. The hope was that a combination of thick capital buffers for banks and careful planning would protect deposits and payments systems while losses were passed on to investors in an orderly way. Regulators planned for a swift recapitalisation of the biggest banks via the conversion of some of their debt to equity—a “bail-in”, in the jargon.

In 2018 and 2019, however, Congress and bank regulators watered down both the resolution planning and liquidity rules, particularly for banks with $100bn-250bn of assets, many of which had lobbied for lighter regulation. There have never been bail-in plans for banks of SVB’s size. Instead, the bank briefly sought last week to recapitalise itself via a doomed issuance of new shares.

The lack of robust planning for failure has meant regulators have had to work on the fly. The problem was made worse by the speed with which SVB lost deposits as Bay Area executives swiped money away using their banking apps. Regulators typically try to resolve banks over a weekend. So ferocious was the run on SVB, however, that it had to be closed during the working day on March 10th. Even if SVB had been solvent and eligible for emergency funding from the Fed—it had plenty of assets to post as collateral—it is unclear whether there would have been time to arrange it.

Some will conclude from the ability of depositors to flee and the readiness of regulators to backstop them that it would be better to abolish limits on deposit insurance altogether—and charge banks up front for full protection. But with adequate capital buffers and resolution planning, depositors would not have been caught up so badly in the crisis. SVB’s failure would have posed less of a threat to the economy and the financial system. Full deposit insurance for the banking system could lead to further recklessness. It would encourage banks to take bigger risks to boost the rewards they could offer depositors, who could be attracted by higher returns but would never have reason to leave a bank on account of its imprudence.

This moral hazard is not the only danger. The other is that the Fed, having seen how SVB buckled as interest rates rose, now chooses to ease off tackling inflation for fear that monetary tightening will cause more failures. Having a week ago bet that rates would reach 5.5% this year, investors now expect barely any more tightening—and for interest-rate cuts to start within six months.

The Fed should not take its eye off inflation (though higher bond prices will ease the strain on banks’ balance-sheets). Now that deposits are safe and the banking system has massive liquidity support on offer, the crisis is unlikely to slow the American economy by much. Moreover, it is not the job of monetary policy to protect lenders’ profits. The right conclusion to draw from svb’s failure is that the regulation of banks which were large but not enormous has been inadequate given the threat they pose to the economy. The job of policymakers now is to remedy that oversight. ■