There has been much discussion of late about the role of independent regulatory agencies set up by the Federal government. The legislature makes laws, but someone has to figure out how to implement broad laws in particular cases. When it comes to areas requiring technical expertise, can and should the legislature set up agencies outside of the control of the executive branch? 

The Federal Reserve, created in 1913, is one of the oldest of these independent regulatory agencies. Monetary policy is front-page news, but there are deeper issues lurking in the back pages of the financial press. In the last two years, for example, the Fed has spent more than it has received in revenue. Since the Fed can literally create money, there is no danger of bankruptcy, but ultimately, the public will end up paying for these losses directly or indirectly through some form of tax. Similarly, the size and composition of the Fed’s assets raise alarms. For a long time, the Fed’s assets were primarily US government debt, but post-2008, the Fed has acquired a significant amount of private debt. 

The temptation is always to declare that we are wrestling with novel problems, that this time is different. But, things rarely are different. Consider The Young Fed by Mark Carlson, an exhaustive study of the regulatory actions of the Federal Reserve before the onset of the Great Depression. It might as well be a primer on managing the regulatory problems facing the contemporary Fed. In fact, the regulatory problems the Fed faced in the 1920s can give great insight into the role of independent regulatory agencies today.

How to Save the Banking System 

The Federal Reserve System was born from the banking crisis of 1907. After the San Francisco earthquake, money flowed out of the East Coast banks to fund the recovery effort. Since the very business of banks is making loans with deposited funds, a large demand for withdrawals creates a liquidity crisis. A perfectly sound bank can reach the verge of failure simply because it is unable to convert its assets instantly into money. The collapse of the banking system was averted when J. P. Morgan stepped in to help.

The crafters of the Federal Reserve system sought a more stable solution to such problems. It seemed rash just to hope that a wealthy individual would always bail out the banking system. Fortunately, the template for the system already existed. In 1873, Walter Bagehot (pronounced badge-it) wrote the governing rule, subsequently dubbed Bagehot’s Dictum: in a time of crisis, the central bank, acting as a lender of last resort, should lend freely against good collateral to a sound bank experiencing liquidity problems.

The Fed’s structure, which looks oddly byzantine today, was designed to do exactly that. At twelve Federal Reserve Banks scattered around the country, officials thoroughly immersed in the banking conditions in their regions could determine which banks fit the criteria of being worthy of receiving funds from the lender of last resort. After all, not every bank in crisis is a fundamentally sound bank. Banks can easily find themselves in crisis if they make bad loans; letting such banks fail is sound policy.

Consider the situation in which this puts a Fed official trying to determine whether to help a bank in crisis. If a sound bank experiencing a liquidity crisis is not helped out, then this can easily create a contagion of bank failures. If you heard that a bank in your town had failed, you might worry that your bank would be affected and rush to your bank to withdraw your funds, causing your otherwise sound bank to fail (cf. It’s a Wonderful Life).

Is the answer to lend to every bank in trouble in order to prevent bank runs? That policy will instantly create a moral hazard problem. If I am the owner of a bank and I know that someone will provide funds if I run into financial difficulty, I have a high incentive to gamble. If my gamble pays off, I will reap high profits. If my gamble fails, someone else will pay my debtors.

It is easy to write a rule telling the people in charge of making decisions to only lend to sound banks facing liquidity crises. The hard part is determining whether a given bank meets that criterion. Why exactly is this bank in trouble? Are its loans fundamentally sound or risky gambles? Why are people pulling funds out of the bank? Do the bank customers know the bank is in trouble? Do you trust the managers of the bank to make wise decisions going forward, or would lending to the bank simply delay the inevitable collapse? Generic rules cannot answer questions like these. Someone has to make decisions in individual cases. Enter the regulators.

The 1920s

To see how little these basic questions have changed, you need look no further than the case studies in Carlson’s book. Far and away, the most marvelous example involves cattle. 

The Federal Reserve Bank of Dallas … ended up owning a substantial amount of cattle after defaults by both banks and ranchers. Managing the cattle involved considerable effort, expense, and staff. … The FRB Dallas’s insolvent bank department consisted of “4 office employees, 11 field representatives, and 13 livestock inspectors and caretakers.” The cattle holdings of FRB Dallas were apparently sufficiently large that efforts to sell them had a meaningful impact on the local market prices … [and] there were a number of complaints from the local cattlemen’s associations.

It is safe to assume that nobody involved in the creation of the Fed imagined it becoming a cattle rancher.

Acquiring problematic assets was commonplace in these years. In May 1921, arguably the most important bank in Salt Lake City, McCornick and Co., suddenly found itself in crisis after the death of its founder. At the time, the bank had 16.5 percent of all the loans, 21.4 percent of all the interbank deposits, and 14 percent of all the individual deposits in the entire city. But, when the Federal Reserve Bank of San Francisco looked into the books, there was reason to be concerned about over a quarter of their outstanding loans. 

No other bank in the area had any interest in absorbing this large and potentially fatal problem. The negative consequences of a bank failure were potentially quite serious, given such a large bank was intertwined with the finances of many other banks in the area. After the Fed got involved, portions of the assets of McCornick and Co. were absorbed by assorted other banks, but the Fed itself acquired a large array of loans with dubious prospects of ever being repaid. (Here, too, the Fed ended up owning a ranch.) It took a decade to sell off the assets, and the Fed lost $200,000 (the equivalent of roughly $3.5 million today).

Whether the Fed should be engaging in actions that could potentially result in operating losses was a big item of discussion. As one Fed official noted in 1926, “In order to be of real value, there are times when a reserve bank has to risk a loss. The regional banks were not created as profit-making institutions.” But, when the Federal Reserve Bank of Dallas lost $500,000 on a single loan in 1923 (the equivalent of $9.1 million today), it is not clear how to evaluate whether the damage which would have been done to rural southeastern Oklahoma by not making the loan would exceed losses of that magnitude from making the loan. 

Is the answer that the Fed should more scrupulously follow Bagehot’s Dictum that loans should only be made against good collateral? While that advice seems perfectly sensible when you think about large banks with a highly diversified set of assets, what do you do about small banks on the periphery, which have few assets that seem perfectly safe? If the Fed demands the safest assets as collateral, the bank’s asset portfolio will become worse, increasing the chances that depositors and shareholders will be unwilling to provide funds to the bank.

Sometimes the Fed was placed in impossible situations not of its own making. In El Paso in 1924, the City National Bank was experiencing a bank run. The managers of the bank reached out to the community to raise funds, and to encourage people to help, they explained that the Federal Reserve Bank would be purchasing $500,000 of the bank’s bad assets. When he heard about this promise, the head of the Dallas Fed was completely surprised, resulting in an odd problem. The Fed could announce that it never made such an agreement, guaranteeing the bank would rapidly fail. Or the Fed could play along with the false announcement. The negotiations behind the scenes were, not surprisingly, quite convoluted. Part of the deal was the removal of the bank’s officers. The run was halted, but the Fed ended up losing $100,000 and there was prolonged litigation about the messy situation at the bank.

As the examples multiply, it rapidly becomes apparent that the Fed faced many problems of a sort that nobody could have anticipated. There would be no way to have written a rule in advance that would have helped manage the problems at the National Bank of Commerce in Frederick, Oklahoma. Over the early 1920s, the bank had acquired a noticeable, but not crippling, amount of bad assets. The normal banking practice in this case was to go to the shareholders and have them provide more funds to the bank in order to keep the bank, and not insignificantly, the assets of these shareholders, safe.

But, in this case, going to the shareholders presented a curious problem. One of the shareholders, Mr. Sims, was a former cashier at the bank. Carlson wryly notes, “Mr. Sims was a ‘legal resident’ of the state penitentiary as a result of his misappropriation of school district funds. Some of the other bank managers had testified against him at his trial. Mr. Sims’ wife controlled the shares of stock he owned, and she was ‘antagonistic.’” So, the bank managers realized that if they went to the shareholders and described the situation, Mr. and Mrs. Sims would “have created, maliciously, considerable comment which would have proven very disastrous.” The Kansas City Fed ended up consistently providing loans to the bank from 1921 to 1926.

Whither Independent Regulatory Agencies?

Losses, unusual assets, and questions that never seem to get resolved are not a new thing for the Federal Reserve. Exactly the same sorts of problems face other independent regulatory agencies. Was it wise for Congress to create these agencies to manage complicated problems?

Looking closely at the Fed’s actions in the 1920s brings the matter into sharp focus. On the whole, these regulators did a decent, but far from perfect, job. Would things have been better if the lender of last resort had been located in the Executive Branch? Moving forward, would the collapse of Lehman Brothers have been better managed if the Chair of the Federal Reserve had been a political appointee in the Treasury Department? The problems the 1920s regulators faced were orders of magnitude simpler than the problems faced by modern bank regulators. Bagehot’s Dictum, which seemed so sensible and succinct, was difficult to apply once the regulators were faced with problems of small regional banks or shadow banks. 

Who should make decisions in these cases? Looking closely at the sorts of decisions being made by independent regulatory agencies, it is hard to escape the conclusion that sometimes they make bad decisions. Regulatory agencies in the executive branch also sometimes make bad decisions. The Trump Administration’s executive order seeking to strip the Federal Reserve of its regulatory independence may have unintended consequences. Blunt executive orders are not a good way to improve decision-making in inherently complicated matters.