Who Bares the Risk for Bank Failures?
Fractional reserve banks are more similar to mutual funds then they are to safes. They exist to make investments using the money they get from investors and depositors. Assets purchased by the bank range in riskiness from reserves, which are the safest, to the very risky assets like sub-prime mortgages. On the liability side there are 3 major categories of investors:
1. Depositors, who get banking services and perhaps a small return.
2. Bondholders, who get a fixed return regardless of investment returns, unless the bank goes bankrupt.
3. Stockholders, who get whatever investment return is left over, as well as control over the company.
Because everyone but the shareholders have a fixed return, the shareholders have an incentive to maximize the riskiness of the investments the bank holds. Equity holders get all of the upside if the risky bets pay off, but the excess losses are imposed on someone else. Because bank failures have negative macroeconomic effects, the government has tried to regulate them to minimize them, often with unintended consequences. Most of the regulations focus on limiting the types and quantities of assets the banks can buy. I believe that such an approach is bound to fail, because shareholders still have an incentive to circumvent the regulations. The government could focus more on the liability side in a couple of ways.
Extra shareholder liability
Under the current institutional arrangement, shareholders have the most direct control over the company, but only stand to lose up to the value of the shares they hold. In the past, governments have required shareholders of banks to put up extra collateral, which is forfeit if the bank collapses. This collateral gives shareholders an incentive to encourage the bank to lend conservatively. Double liability is when the shareholders put up collateral equal to the value of the stock they hold and unlimited liability is when shareholders are liable for the entire losses the bank takes, if it fails.
Leverage refers to the ratio of stocks to bonds which make up a bank’s ownership structure. If a bank cannot pay the bondholders when the bond is due, it is forced into bankruptcy. So, the less equity (stock) relative to the amount of debt (bonds) a bank has, the more fragile the bank is. Most modern banks are leveraged 30:1 or so, meaning if they take a 3.5% loss on their investments, they are forced into bankruptcy. Needless to say, this is a rather risky institutional arrangement. If banks were required to have a 10%-20% equity cushion, they could absorb more losses in a crisis without requiring bailouts or bankruptcy.
Under the U.S.’s current de facto system, if banks take losses, bondholders are not liable for losses. Instead, if bondholders are at risk of taking a loss, the federal government simply gives the bank piles of money, imposing losses on taxpayers instead. According to the letter of the law, bondholders accept the risk that if the bank goes bankrupt, they will take some of the losses from the failed loans. In exchange, they get a higher return than the bank’s depositors. If bondholders suspect they may take losses if a bank acts irresponsibly, they will demand a higher return on their bonds and the bank will have an incentive to lower it’s riskiness.
Prior to the creation of the FDIC, if bank’s investments did extremely poorly, the depositors could take losses. Typically, because shareholders had double liability, and bondholders actually suffered losses, depositors only lost a small fraction of the money they put in the bank. But, in the worst bank collapses, depositors did take losses, and that threat of loss could trigger bank runs as people rushed to get their money out before bankruptcy. The FDIC ended bank runs and depositor losses, but at the cost of removing the incentive for banks to reduce the riskiness of their investments in order to attract deposits. As an aside, this is the reason why bank’s architecture was so impressive in the past. If the bank owned a lot of highly valuable real estate, that was a credible signal that they could pay their depositors back.
What is to be done?
Under the current system of regulation, stockholders are liable for the full value of their stock when the bank collapses, meaning they can lose everything. Neither depositors nor bondholders are liable for any of the losses, and are fully insured by the federal government. The asset side of banks is highly regulated in an effort to get banks to buy safer assets, but in my opinion, the SEC and Treasury have been captured by the banks and so any regulation, no matter how fierce it sounds, is likely to be ineffective. I think that the liability side of banks is perhaps the more effective method to controlling the level of risks banks take.
Mark A. Calabria wrote an article advocating a return to imposing losses on depositors above $40,000 per account. I think such steps would be reasonable in theory, but in practice I worry that they would not be time consistent for regulators. Meaning, in a crisis where bondholders were getting bailouts, will normal citizens be happy to sit by and watch their savings accounts disappear? Also, depositors are the least sophisticated of the investors in a bank and often have neither the time, the training, nor the inclination to analyze the riskiness of a bank’s portfolio. Perhaps an alternative is Steve Waldman’s idea to insure depositors rather than banks, but while such an idea would protect normal people, it would do nothing to reduce the macroeconomic impact of bank failures*.
I would prefer to use “speed bankruptcy” to deal with failed banks in a crisis. In speed bankruptcy, stockholders are wiped out and bondholders are handed a number of newly created shares proportional to the amount of bonds they held prior to the bank failure. Such a proceeding would avoid any changes to a bank’s assets, insure depositors effectively 100% without FDIC intervention, and avoid any debt deflation traps. Banks would simply open the next day under new management. The big “losers” under speed bankruptcy relative to our current system are bondholders, but they explicitly took the risk when they bought the bonds, so I don’t feel sorry for them. Once serious losses were imposed on the liability side of banks, banks would have a clear incentive to invest in safer assets and the government would not need to impose asset side regulation.
*Despite my Sumnerian outlook in general, I think that the central bank simply does not have the skill to maintain demand in the face of widespread bank failures. Maybe one day they will, but not today.