You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. Economic Forces arrives weekly in the inboxes of over 7,000 subscribers. You can support our newsletter by sharing this free post or becoming a paid subscriber:
Some people choose to consume less today in exchange for the ability to consume more later. We call those people savers. It is important to note that the decision to save is costly. By saving, one is giving up current consumption in exchange for future consumption. If people are indifferent between when they consume, this isn’t a big deal. However, all else equal, people tend to prefer having something in the present to having the same thing in the future. Thus, people who save will typically want to be compensated in some way for giving up current consumption.
A saver can receive compensation for saving by purchasing a financial asset or lending that savings to someone else. However, this creates possible risks for the saver. First, the individual saver might not have enough savings to diversify his or her savings. Second, the individual might experience something unexpected and need to get their money back right away.
One way to think about a fractional reserve bank is as a coalition of depositors. The depositors pool all of their money together, they set some of that money aside for when savers unexpectedly need to come and get their money back, and they lend out the rest of the money to people who need to borrow money. Depositors are therefore able to earn a rate of return on their savings while also having the ability to show up any time they want and get their money, if needed.
For this sort of thing to work, two things have to be true. The credit risk of the borrowers and the liquidity risk of the depositors should each be uncorrelated, idiosyncratic risks. In other words, there is always the chance that a borrower defaults. However, you don’t want a scenario in which a large number of borrowers all simultaneously default. Similarly, you don't want a scenario in which a large fraction of depositors all show up and want their money back immediately.
The bank needs to have diversification across both borrowers and depositors.
Typically, when we think about bank failures, the cause is a lack of diversification among borrowers. Maybe the bank fails to recognize the correlated risks of their borrowers. Or perhaps, as was true in the United States for a long period of time, there are restrictions on branch banking. In that case, it is difficult for banks to lend to people at distances far beyond their physical location. Borrowers therefore have some correlated risk associated with the performance of the local economy or the concentration of one particular industry in that area. Regardless, when risks are correlated across borrowers, one unexpected economic shock can lead to a wave of defaults, which leaves the bank insolvent and therefore unable to pay back all of its depositors in full.
But bank failures can also be precipitated by a lack of diversification among depositors as well. If the likelihood of showing up early to get one’s deposits is correlated across depositors, it is possible to have a bank run. People all show up at once and want their money back. The bank, of course, only has a fraction of the money necessary to allow for these withdrawals. To meet these requests, the bank will have to sell assets. Doing this in a hurry tends to cause the bank to take losses on these assets. After a sufficient amount of losses the bank could become insolvent.
The problem with Silicon Valley Bank seemed to be some combination of the two problems. The bank was holding a lot of long-dated bonds on the asset side of its balance sheet. The bank’s depositors seemed to be mostly tech companies. When the Federal Reserve started raising its target interest rate, this had two effects. First, this reduced the valuation of tech companies since these companies tend to be spending a lot of money now in the hopes of (high) profits in the future. When interest rates rise, it becomes more expensive to wait for those future profits. This meant that funding for these companies started to dry up and the companies started drawing down their deposits.
Second, the Fed’s tightening of policy led to higher yields on bonds. These higher yields come from lower bond prices. This meant that SVB was experiencing net withdrawals while the market price of its assets were declining. Selling these bonds meant realizing losses. When the magnitude of the potential unrealized losses became apparent to depositors, many of these companies started running to get their money out. This meant that SVB had to realize losses and was at risk of insolvency. Ultimately the Federal Deposit Insurance Corporation (FDIC) stepped in.
The view of banks as a coalition of depositors helps us to understand stories like this. The lack of diversification can lead to a bank’s insolvency. However, viewing banks as a coalition of depositors is only part of the story. For example, if banks are just coalitions of depositors, why would a bank ever behave in the way SVB did?
What is missing from the coalition-of-depositors view is how banks actually operate. In reality, banks are not just coalitions of depositors, but rather are owned by shareholders and operated by managers. Thus, it is more accurate to say that shareholders and depositors pool their money together to finance the banks asset portfolio. It is the inclusion of shareholders in our analysis that makes things interesting.
Shareholders and depositors have very different interests. The coalition-of-depositors view of banks really helps illustrate the incentives of depositors. They want the bank to have well-diversified assets and sufficient reserve balances.
Shareholders might want something different. The great, Nobel Prize-winning economist Robert Merton can help us to understand why. In the 1970s, Merton pointed out that we could think of the equity value of a firm as being akin to a call option on the assets of a firm with a strike price equal to the value of the firm’s debt. A call option gives the holder the right, but not the obligation, to purchase an asset at a particular (strike) price at some date in the future. The intuition is as follows. If the firm was to shut down, it would sell off all of its assets. It would use the proceeds to pay off its debt. The money left over would go to its shareholders. As a result, one could think of shareholders having the right to buy the assets of the firm at the cost of the firm’s debt at some point in the future.
Why is this insight important?
The reason that Merton’s insight matters here is that options have convex payoffs (as the value of the firm’s assets rise, the value of the option increases at an increasing rate). This means that options are more valuable to the option holders when the underlying asset (in this case, the firm’s asset portfolio) is more volatile.
Applying this to banks, we can see that shareholders of the bank would prefer a more volatile portfolio of assets for the bank whereas depositors would prefer the bank would be a boring institution with well-diversified assets and ample reserves. While this is generally true of any firm, the thing that makes a bank unique is that its debt holders (the depositors) have the ability to redeem their claims on demand at any time. If a lot of depositors all show up at once to get their money, the bank might have to sell volatile assets to raise money. This risks insolvency.
This raises a natural question. If shareholders prefer more volatile assets and depositors prefer less volatile assets, how can the preferences of depositors and shareholders be aligned to avoid the insolvency risk just described?
Historically, shareholders of banks were subject to multiple liability. The most common version of this seems to have been double liability. The way that this worked is that a shareholder would buy X dollars worth of stock in the bank. If the bank become insolvent, not only did the shareholder lose his X dollars, but the shareholder was also responsible for compensating depositors using up to X additional dollars of the shareholder’s own personal wealth. This is referred to as double liability since a shareholder investing X dollars in the bank would have a maximum loss of 2X dollars.
It is easy to see how this sort of arrangement would help to navigate the conflicting visions of depositors and shareholders with regards to what banks should do. Multiple liability aligns the financial incentives of the shareholders with those of the depositors by making shareholders responsible for depositor losses.
In the current banking system, shareholders are subject to limited liability. This means that a shareholder who buys X dollars worth of stock cannot lose more than X dollars. Shareholders are not financially responsible for depositor losses. Perhaps that is some tacit recognition that this leads to misaligned incentives since the government also insures deposits in the United States up to $250,000.
A comparison of these two legal liability structures should demonstrate the differences in incentives. Under multiple liability, shareholders are financially responsible for any losses sustained by depositors. Under limited liability, shareholders have no responsibility for depositor losses. Furthermore, if the government is going to provide deposit insurance, shareholders do not have to worry about depositor losses much at all — at least not as it pertains to those with balances below the insurance threshold. Thus, given that shareholders already have a greater preference for risk-taking than depositors, we would expect that banks that operate within a multiple liability regime would engage in less risk-taking than in a limited liability regime with deposit insurance.
This is something to think carefully about as the U.S. government considers expanding the magnitude of deposit insurance in the midst of the recent turmoil in banking.
Nice work! Do you think multiple liability has a chance at being instituted?
Also, you wrote: "Shareholders are not financially responsible for shareholder losses." You meant "depositor losses," right?
"Depositors are therefore able to earn a rate of return on their savings while also having the ability to show up any time they want and get their money, if needed."
That's completely inadequate. The rate of return of sight deposits is zero and the ability to show up at any time is accorded by contract. Depositors are completely unaware of what is doing the bank with their money. And to take deposits (redeemable at sight notice) to invest in very long term assets (public debt and mortgages) should be a law violation and poses serious macroeconomic damage.