Silicon Valley’s lesson: How to change banking to prevent crises and bailouts

Silicon Valley Bank’s failure, and the ongoing turmoil in the banking system, provide a deep lesson. Our current bank regulation system, vastly expanded under the Dodd-Frank Act after the 2008 financial crisis, is fundamentally broken.

The cause of Silicon Valley Bank’s failure was simple: The bank got its money almost entirely from large business deposits. A tech company might get a huge amount of cash from its venture capital investors; the company parks the cash at SVB, and uses it over a few years to build the company. The bank turned around and invested this money in long-term assets, primarily U.S. Treasury bonds and government-guaranteed mortgage-backed securities. 

When the Federal Reserve started raising interest rates to combat inflation, the value of those securities plummeted. Ordinary depositors like you and me might not care, because the Federal Deposit Insurance Corporation (FDIC) insures our deposits, up to $250,000. Even if the bank goes under, our money is safe. But businesses with much bigger deposits do not have that reassurance. In the case of SVB, they ran to get their money out, forcing the bank to sell assets at big losses and prompting even more depositors to run. 

The FDIC stepped in and closed the bank. Over the weekend, the Fed announced that all deposits — even those with hundreds of millions of dollars — would be guaranteed. 

Since the 1930s, we have had deposit insurance to stop such runs. But deposit insurance begets risk-taking. Depositors have no incentive to worry if the bank is safe and can just chase the highest interest rate. Banks are gambling with the government’s money — heads, they win; tails, the FDIC loses — and thus have incentives to take risks. So, we have government regulators to try to stop banks from taking too much risk. 

Over the years, after each banking or other financial crisis, the U.S. has expanded the range of creditors — depositors — who are guaranteed to stop runs, and has vastly expanded the regulatory apparatus to try to stop risks. But SVB did not fail by fraud, by toxic derivates, by anything complex. The risk of its business model was utterly obvious. How could the immense bank regulatory apparatus have missed this elephant in the room? 

The answer is part inevitable human failure and part rules that are so overly complex that they hide elephants in plain sight. 

The Dodd-Frank Act already has spawned over 30,000 pages of regulations. Do we really need another 10,000 just to spot the oldest risk in the book? It merely needed someone to put two and two — uninsured deposits and long-term assets — together to come up with four, but that’s not advanced math. Rules reportedly were “loosened” so that SVB did not have to pass “stress tests.” But the Fed’s stress tests, astonishingly, did not even ask banks to consider the effect of rising interest rates.  

Long-term government bonds and mortgage-backed securities are the safest long-term investments on the planet. They are vastly safer than the cash flows of any stock. Yet we have an army of regulators trying and failing to monitor risks of the safest assets on the planet. Why? Only because we allow companies like SVB to borrow $90 of every $100 it gets to fund those investments in the form of run-prone deposits. 

The idea that “more regulation” together with a blanket guarantee of all deposits will stop future crises is obviously bust. We now have firmly enshrined that you can make money in good times, and the government will sop up the losses in bad times. And banks have become an oligopolistic, crony-capitalist business. SVB’s management has been criticized for doing little risk-management and pursuing political connections instead. Yet, in retrospect, they appear to have been right to do so! That strategy has now paid off handsomely. 

There is an alternative. If banks had to funnel all deposits and short-term debts into interest-paying reserves at the Fed and short-term Treasury investments, runs would end forever as there would always be money to satisfy any depositor. Banks should obtain funds for long-term bonds, loans and risky investments by borrowing long-term (certificates of deposit, long-term bonds) and issuing lots of equity. Investors in these mediums cannot demand the face value of their money instantly, and would bear losses without causing crises.

If we allowed such a system to emerge, we would eliminate private-sector financial crises and government bailouts forever. It would require almost no regulation at all. Whatever the quibbles and small costs of such a system, the only question is how many more crises and bailouts we will suffer through before adopting it — or whether we will converge to a sclerotic government-run system instead. 

John H. Cochrane is a senior fellow of the Hoover Institution and an adjunct scholar of the CATO Institute. He blogs as the “grumpy economist.”