THE STOCK MARKET CRASH OF 1929

AND THE 1920s TECHNOLOGY BUBBLE

The usual and familiar accusations were made and have since been repeated. Speculators were 'evidently' responsible for driving stock prices away from their fundamental value, causing a bubble in the market. Speculators were aided by 'easy credit' from banks and stock-market brokers. Not just the stock market crash was blamed on speculators, but, even more, speculators and the stock market were blamed for the entire Great Depression.

Fundamentals

The fundamental value of stocks is the discounted present value of expected future dividends. Due to dividend smoothing and other considerations dividend payments usually lag earnings of firms. It is not surprising for example that during 1929 share prices of utility companies rose strongly, although most of these companies had not paid any dividends. Utility companies were seen to benefit from the expanding sale of electrical appliances and used their earnings to increase investment rather than distribute as dividends. RCA also did not pay dividends.

Even without complications from dividend-pay-out policies, the data show that from the early 1920s stock prices followed the increase in dividend payments. Estimates show that in 1929 stock prices implied expected dividend growth rates that were lower than the actual post-WW2 growth rates (Sirkin, 1975). Recent estimates of a fundamental value for the S&P500 index show that no evidence of a bubble in 1929 exists (Barsky and DeLong, 1990; Donaldson and Kamstra, 1996). At the time, professional economists provided no consensus view, but a number of them also argued that the stock market was in line with fundamentals (among them the famous Irving Fisher). After the October 1929 crash share prices recovered slightly. But from early 1930 to mid-1932 the stock market and its fundamental value followed the downward trend in economic activity and dividends, its level adjusted for an in increase in the risk premium.

The crash

The stock market peaked September 3, 1929. On September 5, 1929 Roger Babson addressed the National Business Conference and predicted that a sharp recession was in the offering. We now know that the index of industrial production actually peaked in June 1929. Restrictive monetary policy by the Federal Reserve was a major cause of the recession. Some of the tightening was attributed to anti-speculative reasons, but most of the Fed’s emphasis with respect to the stock market was on moral suasion and direct pressure on banks and brokers.

More important for the monetary policy tightening was the gold outflow, mainly from the United States to France. The Fed raised the discount rate during January-July 1928 from 3.5% to 5% and in August 1929 to 6%. Because prices were falling, real interest rates were much higher than nominal interest rates. Additional bad news was the failure of the business and financial empire of Clarency Hatry in Britain in September 1929. In October 1929 regulators denied the utility company Boston Edison a request for a stock split, fearing further price speculation on a price considered already higher than justified, and accusing the company of earning monopoly profits. On October 28, 1929 the Smooth-Hawley tariff was enacted. However, although surely bad news and although frequently mentioned as a major cause of the stock market crash and subsequent economic depression, the decision on the tariff came actually after the peak and the first panic in the stock market. US exports were also only 7% of GNP, and no evidence exists that import-export companies suffered more than other companies during the crash.

Many investors had bought stocks on margin, with borrowed money using stocks as collateral, and were liable for both interest rate costs, loan repayment and additional margin calls when prices decline. When stock prices started to drift downward during September/October 1929 the volume of trade increased. When prices declined and hope of a speedy recovery faded, the number of margin calls increased. Many were forced or attempted to sell their shares, and through sheer size of volume brokerage firms were swamped and prompt reporting of prices became impossible. Large scale or panic selling disrupted the market on a number of ‘black’ days in October 1929. Information on transaction prices was unavailable. Furthermore, speculators had a perverse incentive to stay away from the market in order to benefit from the distress sale and free fall in prices.

Conclusion

There is nothing particularly strange in the mechanics of the October 1929 stock market crash(es). Calculations with respect to a 'rational' fundamental value for the stock market in 1929 show that the market was not definitely overvalued at the time, given ex ante expectations of a hypothetical rational investor. In the following years the stock market simply followed the decline in economic activity and the general price level. Blame for the Great Depression cannot be found in the stock market, but rests with a failure of monetary policy.

Barsky & DeLong (1990,1993) simple bull-bear market model

Barsky and DeLong use a simple single-phase dividend growth model for share prices. In logarithms:

p_t = d_t - ln(r_t - g_t)

g_t = (1-q) Σ_i=0 qⁱ Dd_t-i

Rather than applying an unrealistic perfect-foresight assumption for future dividend growth, the expected dividend growth rate g is based on an adaptive model that extrapolates from historical dividend growth rates Dd. For many economists extrapolation is irrational, but they ignore the fundamental non-stationary nature of the real world that may make extrapolation unavoidable as a rational forecasting tool. Although the precise forecasting model can be improved somewhat (see for example Donaldson and Kamstra, 1996), the results do not seem to be very sensitive to the resulting small improvements. BDL assume the discount rate r to be either constant or on a (downward) trend. Again, the precise assumption does not affect the results very much in this long-run evaluation of bull and bear markets.

The main driving force in the model is the empirical fact that dividends follow an (approximate) random walk or unit root process and that forecasting future dividend values is almost impossible. (Which is contrary to the basic assumption made in most of the so-called bubble literature and studies focusing on the variance-bound test, using the unreal perfect-foresight-with-error model of rational expectations.)

Note: I would normally strongly recommend to always use a multiple-phase dividend growth model (see for example Panirgitzoglou and Scammel, 2002). Too many researchers using the single-phase model end up with having to fit the model with long-run growth values that are easily criticized as unrealistic. The more flexible multiple-phase growth model is generally robust to this long-run value problem. The reason is that the more variable near-term dividend growth rates have a much larger impact in the weighted-average present value calculation than the more distant future average growth rate.

Literature

• Barsky, R.B. and J.B. DeLong, "Bull and bear markets in the Twentieth Century," Journal of Economic History, vol.100 (2) June 1990.
• Donaldson, R.G. and M. Kamstra, "A new dividend forecasting procedure that rejects bubbles in asset prices: The case of 1929’s stock crash," Review of Financial Studies, vol.9 (2) Summer 1996.
• White, E.N., "When the Ticker ran late: The stock market boom and crash of 1929," in E.N. White (ed.) Crises and Panics: The Lessons of History. 1990.
• White, E.N., "The stock market boom and crash of 1929 revisited," Journal of Economic Perspectives, vol.4 (2) Spring 1990.
• Bierman, H., Jr., The Great Myths of 1929 and Lessons to be Learned. 1991.
• Bierman, H., Jr., The causes of the 1929 stock market crash: A speculative orgy or a new era? 1998.
• Cecchetti, S.G., "Understanding the Great Depression: Lessons for current policy," NBER working paper #6015 1997.

Internet

• * Webpages on the stock market crash link, link
• * Barsky and De Long (1990), Bull and bear markets in the twentieth century. Journal of Economic History. (PDF file) link
• * Cargill and Mayer (1998), The Great Depression and history textbooks. The History Teacher Volume 31 Number 4 August 1998 link
• * 1929 Stock Market Crash website link