The most important thing to know about the recent banking problems in the United States is that they are not at all like the problems we saw in 2008-09. The crisis in the banking system during the Great Recession was a result of the fact that hundreds of banks across the country, including several of the largest, were insolvent. The current situation, with some banks facing runs from uninsured depositors, is not at all comparable. Media reporting, which has frequently made this comparison, has played a major role in exacerbating the recent crisis.
The banks’ problems in 2008-09 stemmed primarily from the collapse of a massive housing bubble. Real house prices in the United States had increased by more than 70% between 1996 and the peak of the bubble in 2006, creating close to $8 trillion in housing bubble wealth, the equivalent of more than $15 trillion in the current economy.
This run-up in house prices clearly was a bubble in that it was not being driven by the fundamentals of the housing market. Unlike the rise in house prices during the pandemic, the run-up in the years 1996-2006 was not accompanied by any remotely corresponding increase in rents.
Furthermore, as prices went ever higher, the vacancy rate on housing units was increasing. By contrast, the recent rise in house prices and rents has been associated with historically low vacancy rates. In short, while it is likely that we are now seeing some decline in house prices across most of the country, we are not going to see anything remotely comparable to what we saw when the housing bubble deflated between 2006 and 2010.
In those years, nationwide house prices fell by more than 30%. In many of the hottest markets, the decline was over 50%, and in some areas close to 70%. The extremes matter in this case, because that is where most of the recent mortgages had been issued.
The situation was worsened further by the fact that lenders had abandoned traditional caution in the issuance of mortgages. While the standard mortgage in prior decades required down payments of at least 10%, many of the mortgages issued in the bubble years had zero money down, and in some cases, people borrowed more than the value of the home to cover moving expenses or other costs.
This meant that many of the mortgages held by banks lost virtually all their value when house prices collapsed, since the costs of foreclosing could exceed what banks could expect to get from selling the house. The same was true of mortgage-backed securities and more complicated derivative instruments. In aggregate, banks were looking at trillions of dollars of losses, and some of the biggest losses were at giants like Citigroup and Bank of America, two of the three largest banks in the country at the time.
Making matters worse, the collapse of the housing bubble threw the economy into a recession, as millions of workers lost their jobs. This meant that even loans unconnected to housing, like car loans and credit card debt, also went bad. There is little doubt that without massive government intervention we would have seen most of the country’s banks go under.
The situation is very different today. Silicon Valley Bank (SVB), the poster child of the current crisis, was very much an outlier for several reasons. First, uninsured deposits were an extraordinary share of its liabilities. These are deposits that exceed the $250,000 limit that is covered by the Federal Deposit Insurance Corporation (FDIC). These deposits accounted for well over 90% of the bank’s liabilities. At a typical bank, the share would be in the range of 20-30%.
Second, the bank was closely tied to the tech sector, which is notorious as a boom-bust industry. When the sector boomed in the pandemic, it meant that the bank’s clients had lots of money to place with it. That has reversed in the past year as the economy has now largely returned to normal, and tech companies are laying off workers and making other cutbacks.
Due to the tech boom, SVB experienced extraordinary growth in recent years, with its assets quadrupling in size between 2018-22. It clearly did not have a plan for dealing with the flood of deposits received in these years. It chose to put them largely in longer-term US Treasury bonds.
While this is a very safe asset from the standpoint of default risk, long-term bonds do have interest rate risk. A ten-year Treasury bond that was bought in 2020 or 2021, when the interest rate on these bonds was close to 1%, may lose 10%-15% of its value, now that the interest rate on these bonds is close to 4%.
Since the increase in interest rates was widely advertised by the Federal Reserve Board, it is striking that SVB did not take steps to hedge itself against interest rate risk, which meant that it faced substantial losses as rates rose.
The concentration of the bank’s business in the tech sector meant that it was especially vulnerable to a run. When some highly visible people in this world withdrew their money, it led others to follow. The bank was forced to sell some of its assets at a loss. When this information became known, it accelerated the run and forced the FDIC to take over the bank.
While most banks are not remotely similar to SVB in their business model, any bank can be susceptible to a run if there is enough fear in the environment. It was this threat of contagion that led the Biden administration and the Fed to take measures to both fully insure all the deposits at SVB and to make a commitment to protecting the depositors at other banks as well.
The reporting on SVB played a large role in producing this contagion. Major news outlets, like The New York Times, highlighted the plight of business owners who they said would be unable to meet their payrolls because their money was tied up with SVB.
This was at best a gross exaggeration. At the time of the seizure, the FDIC announced that it would issue an advance payment the following week. This was likely to be worth well over half of the money in the accounts, and quite possibly as much as 70% to 80%.
It would also issue a certificate for the remaining funds. While the value of the certificate would not be known until the resolution was completed, and could be insufficient to replace the full amount in the account, it would likely make up most of the remaining gap.
Furthermore, these certificates are marketable. Any depositor who needed their money likely could have sold their certificate the same day, albeit at some discount relative to what it was expected to eventually pay. This reality was rarely conveyed in news stories, most of which implied that depositors stood to lose the full amount over the $250,000 insurance cap. This was a huge factor in feeding bank runs across the country.
For now, it appears that the panic has been contained. There are important issues left to be resolved in the US banking system, most notably whether all deposits will be insured. However, it seems the crisis is over. With luck, the damage to the economy will be limited, and we can go back to worrying about the direct impact of the Fed’s rate hikes.