One area that I do this so-called "research" about is banking panics, specifically the Panic of 1907. For historical purposes in the U.S., it's a fascinating story. The Panic of 1907 changed the future of banking, specifically leading to the Federal Reserve.
For us in the post-2007 world, the Panic of 1907 might help shed light on the mechanisms that lead to a banking crisis. From my vantage point, 2007 mimicked 1907, but with flashier technology. Instead of runs on the trust companies, in 2007 shadow banks experienced a run. History rhymes.
The actual banking Panic (people screaming, running to get money) is worth reading about. It sounds like 2007, where a few players dictated what happened. It's a good tale. The United Copper Company tried to corner the copper market, like the Duke brothers in Trading Places did for oranges.
That failed. Panic ensued. Fun stuff.
If you want to learn about the personalities and the anecdotes surrounding the Panic, pick up Robert Bruner’s The Panic of 1907: Lessons Learned from the Market's Perfect Storm. It’s not Michael Lewis, but the stories are interesting.
So that’s the sexy stuff about the Panic of 1907. Yeah... That’s not what I study. Sorry.
I look at the build-up to the Panic. For reference, instead of looking at the Lehman and Bear Sterns of 1907, I look at the equivalent of the recent housing bubble. There are none of Michael Lewis's fascinating characters, but it’s the real driver.
So what happened before the Panic of 1907?
It was a classic boom/bust story. You've seen it once; you've seen it a thousand times.
The money supply climbed. Gold stock rose at 6.8% per year between June 1987 and June 1914. Gold production averaged about $110 million for the twenty years ending with 1890. The next twelve years saw an annual average production of $220 million (Conant and Nadler 1969, 703).
The stock of money, what we'd call M1, rose even faster at 7.5% per year during the same time frame. That might not seem like a much, but this period included six recessions and an absolute decline in money during 1907. This growth in the money supply exceeding gold came almost entirely from increases in national bank notes (Friedman and Schwartz 1963, 137). The national banks held gold and created even more paper notes. The word "leverage" might sound familiar.
The total volume of money rose from $1.5 billion in 1896 to $2.1 billion in 1900 to $2.7 billion in 1907 (Conant and Nadler 1969, 699). Up, up, up. It continued until the Panic. Starting in 1902 until 1906, the growth in the money supply surpassed 7% per year (Conant and Nadler 1969, 153). For comparison, Friedman and Schwartz calculated stable prices require 5% growth.
As with any macroeconomics story, pieces move together. These changes in the money supply matched changes in the import-export markets. Increases in money helped foreign trade. Exports of merchandise rose from $1.3 billion in 1900 to $1.74 billion in 1906 to $1.88 billion in 1907. At the same time, imports also rose from $1.2 billion in 1906 to $1.4 billion in 1907 (Conant and Nadler 1969, 698).
This is the standard, econ 101, free-market, specie-flow mechanism part of the story. That's not everything. 1907 has its own quirks.
Treasury Secretary Shaw used government deposits to induce banks to import gold starting in February 1906. The Treasury promised to increase governmental deposits at any bank that imported gold (Friedman and Schwartz, 155). As usual, the government tinkered. Maybe you could call it a “nudge.” That's the new, popular term.
On top of the help from Uncle Sam to make things fun, credit markets changed in a drastic fashion. Credit expanded at a greater rate than the growth of gold. Between 1890 and 1907 the United States increased bank credit $13 billion upon a gold increase of $3 billion. Most of this increase happened after 1897 (Johnson 1908, 455).
State banks were not required to keep reserves. In fact, banks could hold the reserves in banknotes. That doesn't seem like a reserve to silly old Brian, but what do I know?
Liabilities of State banking institutions expanded by over $5,000 million between 1900 and 1907 while reserves only grew by about $171 million. “Their deposit and credit operations were capable of expanding to enormous proportions without any definite relation to gold.” (Conant and Nadler 1969, 707). It’s a pyramid that lends itself to panics.
The increase in credit over the growth in the money supply was also affected by non-state banks. Commercial and trust bank reserves declined from 18% in 1897 to 10% in 1907 (Johnson 1908, 458). There was no minimum reserve requirement for trusts in New York City until 1906, when it became 15%. Even then, only needed to hold a third of this requirement in specie or legal tender. Meanwhile, national banks had 25% reserve requirements in specie or legal tender.
Due to this low requirement, trusts were able to issue credit to a greater extent than other banks could due to looser regulation. The capital-asset ratio was 4.8% for trusts in New York City compared to 5.8% at state banks and 7.5% at national banks. Part of this difference was due to the fact that trusts had much fewer restrictions on their assets. They could own stock and real estate directly.
It might not surprise you that almost the entire contraction in New York City happened at trust companies. (Moen and Tallman 1992, 616-618). This panic is sometimes called the Knickerbocker Trust Panic after the biggest bank failure at the time.
Maybe surprisingly to some, unlike the Great Depression, the stock of money was not greatly affected by the amount of bank failures. 0.0026% of banks suspended payments at some point compared to 0.034% in 1930 (Wicker 2000, 7). The (albeit wrong) story of bank failures told about the Great Depression don't apply to 1907.
In fact, by the time of the panic in October, there was a heavy influx of money and outflow of goods. As stated above, the net exports rose from $1.46 million in August 1907 to $114 million in December of the same year (Conant and Nadler 1969, 717). This corresponded to an increase of the money supply by $72 million of Government deposits, by $70 million from gold imports and $50 million in new banknotes (Sprague 1908, 367). Although there is not full agreement on this point, inelasticity of currency appears not to have been the cause of the problem since this influx happened during the heart of the panic, September through November.
So what is the causal mechanism? Is everything Treasury Secretary Shaw's fault? Probably not. Why did banks become so leveraged? I don't have a clue. No one else seems to know either. That's why it's called research.
To Be Continued...
This is just a teaser of all the fun stuff that happened during the Panic of 1907. Most of this has been known for 50 years, although economists might forget it. That’s a shame.
With all the focus on “big data,” it is sad to ignore all the data that happened before WWII. While the institutions were different, the same economic laws operated in 1907 as 2007.
(Correction: This post originally had the volume of money for the years 1896, 1900, and 1907 in millions, instead of the correct amount in billions.)
- Allen, W. H. ”What Caused the Panic.” The Sewanee Review 16.1 (1908):85-103
- Conant, Charles A., and Marcus Nadler. A History of Modern Banks of Issue. New York: A.M. Kelley, 1969.
- Friedman, Milton, and Anna (Jacobson) Schwartz. A Monetary History of the United States 1867-1960. Princeton: Princeton University Press, 1963.
- Johnson, Joseph F. ”The Crisis and Panic of 1907.” Political Science Quarterly 23.3 (1908): 454-67.
- Moen, Jon, and Ellis W. Tallman. ”The Bank Panic of 1907: The Role of Trust Companies.” The Journal of Economic History 52.3 (1992): 611-30.
- Sprague, O.M.W. ”The American Crisis of 1907.” The Economic Journal 18.71 (1908): 353-72.
- Wicker, Elmus. Banking Panics of the Gilded Age. Cambridge, UK: Cambridge UP, 2000.