Banking panics and financial crises
I have been doing research on banking for some time. Much of this research is on “runs on banking systems” which also are called “banking panics.” More recently I have started work on financial crises, partly because of the financial turmoil that started in the middle of 2007, which came to a head as a crisis in September and October 2008 and now is winding down (October 2009).
Runs on banks or similar institutions such as money market funds are a long-standing problem, going back at least to medieval times and probably back to ancient Greece and Rome. My research emphasizes banking panics and responses to those panics, with much of the research based on nineteenth-century U.S. experience.
I’ve given quite a few talks to non-economist groups about the current financial crisis. I also have made a few media appearances related to the crisis. The most recent media appearance was a taped interview at the University of Francisco Marroquin. The interviews are up on YouTube. The first interview on YouTube is about how it all began and why I think government-sponsored enterprises and a relatively small amount of securities tied to subprime mortgages are at the center of the problem.
In the second part of the interview on YouTube, I talk about financial innovation and why it is important for economic growth and the well being of people around the world.
I also have appeared on the network channels in Charleston SC and on TV3 and RTE in Ireland.
The talks that I given range from half an hour to an hour, with most about three-quarters of an hour long. Forty-five minutes is over three times longer than the combined YouTube interview from UFM, but I still gloss over a lot of important detail.
A talk a while ago now was on October 23 as part of the BB&T Speaker Series at the College of Charleston in Charleston, SC. “The Financial Turmoil in 2007 and 2008” is an Adobe Acrobat copy of the PowerPoint slides. The talk covers the general background of the turmoil and crisis and goes into a little detail about the run on money market funds. The run on money market funds was the initial event that led to the current financial crisis in October 2008.
In a technical paper “Why Do Bank Promise To Pay Par On Demand?” published in the Journal of Financial Stability, Margarita Samartín and I discuss the issues concerning money market funds in a section toward the end of the paper. You might find that section interesting even if you’re not a professional economist and the rest of the paper doesn’t interest you.
While at the College of Charleston, I also gave a talk titled “The Great Depression of 2008?.” While a substantial slowdown in the economy is all but certain at this point, something like the Great Depression of the 1930s is not particularly likely. It is possible to produce a severe depression: after all, a severe depression was produced in the 1930s. It would take seriously ill-informed government policies to produce one though.
Earlier talks on the financial turmoil included material of interest to particular groups, so I have left the slides from some of them here.
One of them was on October 7 at an American Bar Association event about commercial mortgage backed securities. The “The Financial Turmoil in 2007 and 2008” is an Adobe Acrobat copy of the PowerPoint slides. The slides summarize some basic aspects of the turmoil. It includes three slides that summarize developments in commercial mortgage backed securities.
In June 2008 I gave a speech to the national meeting of the Government Finance Officers Association. While also titled “The Financial Turmoil in 2007 and 2008” I discuss one particular development, the run on the Florida Government Investment Pool. This was similar to the runs on individual money market funds without the effects on financial markets.
The Florida Government Investment Pool operates much like a U.S. money market fund. This run was essentially the same as an old-fashioned bank run. These developments in Florida are of interest to government finance officers around the U.S., because many states have similar arrangements. In fact, the session was on the safety of money market funds generally, including government investment pools. The run was of interest for other reasons, including the similarity of the investment pool’s operations to a money market fund. Since the week of September 15, 2008, this has become of more interest because of the actual runs on money market funds. The run on money market funds was not entirely suprising. The reasons are explained in somewhat technical language in my paper “Why Do Banks Promise to Pay Par on Demand?”
I have a couple of papers on related topics written for non-economists.
“Wildcat Banking, Banking Panics and Free Banking in the United States” Federal Reserve Bank of Atlanta Economic Review 81 (December 1996), 1-20 is a relatively readable discussion of bank runs in the free banking period in the United States.
This paper also touches on the relationship between electronic money and free banking
Private remedies developed for bank runs. R. Alton Gilbert and I focus on private remedies for bank runs in the National Banking period in “Bank Runs and Private Remedies” Federal Reserve Bank of St. Louis Review 71 (May/June 1989), 43-61.
The following papers are written primarily for professional economists.
Why do banks promise to pay the face value, or par value, of their liabilities on demand if it is all but certain that they will not honor this promise at some unknown future date? To my mind, this most basic question about banking is a simple question. Banks promise to give you a dollar whenever you want in exchange for your giving them a dollar. Banks keep fractional reserves, which means that they inevitably cannot honor this promise at some point. If enough people want currency at one time, the bank runs out of currency. Sooner or later, enough people come to the bank at one time that the bank cannot honor the promise.
If banks didn’t make this promise, bank runs (called banking panics in the literature) would be impossible. Yet, banks make this promise. Why?
In “Why Do Bank Promise To Pay Par On Demand?”, Margarita Samartín and I summarize the theories and provide some evidence from historical banking. The history is informative about the theories, suggesting that all the theories are informative about some of the history and no one theory is informative about all of banking history. None of the theories explains why money market funds operate to keep their net asset value at a dollar, which is the same thing as saying that money market funds operate in such a way that you can put in a dollar and get your dollar back later plus interest on demand.
The analysis shows that money market funds are subject to runs in the same way that banks are, as long as the funds guarantee to pay investors on demand no matter what happens to the value of the underlying assets.
Unfortunately, the paper and my mention of this possibility on this webpage were prescient about money market funds. The analysis in “Why Do Bank Promise To Pay Par On Demand?” answers why money market funds made themselves vulnerable to a run. Short answer: their customers prefer to have their funds available on demand without worrying about the current market value of assets. This is so even though it is not possible to do this always. It’s easy to state this preference, it is not so easy to make such a preference comprehensible in a coherent economic theory. No one has done it to date.
Each of the theories examined in the paper explains some historical banking arrangements, but none explains current U.S. money market funds. Curious, isn’t it? This paper is in press at the Journal of Financial Stability in 2008.
This analysis also indicates a key aspect of how electronic money is likely to work if it ever gets off the ground. Today’s money market funds in the U.S. provide an excellent example suggesting that electronic money is likely to be redeemable at par.
While it is easy after the fact to say that banks failed in runs because they were risky, it may not be so obvious before hand. Rik Hafer and I examine banking panics in the free banking period to see whether banks that appear risky before the panic are more likely to fail. We find that banks failing in a panic did appear riskier by general financial criteria before the panic. This is an occasion when the conventional wisdom at the time is correct. This paper, “Bank Failures in Banking Panics: Risky Banks or Road Kill?”, was published in Bank and Financial Market Efficiency: Global Perspectives, volume 5, Research in Banking and Finance, edited by Iftekhar Hasan and William C. Hunter, pp. 47-70. Amsterdam: Elsevier Ltd., 2004.
Runs on the state banking systems occurred in the free banking period. Iftekar Hasan and I document the existence and some implications of bank runs in “Bank Runs in the Free Banking Period.” Journal of Money, Credit, and Banking 26 (May 1994), 271-88. You may be able to access this online. You can get a copy at many libraries or from me.
A suspension of payments is one way to deal with a bank run: in short, just quit paying out. This is what the fund that “broke the buck” in October 2008 did, but other funds did not do so. Such a joint action is called a “suspension of payments”. “Suspensions of Payments, Bank Failures and the Nonbank Publics’ Losses” (with Iftekhar Hasan) provides evidence on the effects of suspensions of payments. Our evidence indicates that, if banks are allowed to just quit paying out, fewer banks fail in the end and the final losses to noteholders are less. This is a draft version of the paper published in the Journal of Monetary Economics in 2007. The data on bank failures, losses and related variables are available by sending an e-mail to me.
The primary assets held by these banks were state and federal bonds. Changes in the prices of these bonds were important for the banks. Rik Hafer, Warren Weber and I published a paper that collected these data together. “Weekly U.S. and State Bond Prices.” Historical Methods 32 (Winter 1999), 37-42. You can get a copy of this paper at many libraries or from me. All of the data we used in the paper are available in an Excel spreadsheet. If you’d like the additional available data on Illinois and New York, there is another Excel spreasheet which includes those data.