Testimony before Subcommittee on Domestic and International Monetary Policy of the
House Committee on Banking and Financial Services
March 21, 2000.
My name is Robert J. Shiller and I am Professor of Economics at Yale University. My research, which concerns macroeconomics and financial economics, has also received support from the U.S. National Science Foundation and from the National Bureau of Economic Research, where I am Research Associate.
In his letter inviting me to testify, Congressman Spencer Bachus asked me to comment on the current monetary policy of the Federal Reserve as it relates to the equity markets, and, in particular, whether the Fed should address the wealth effect through general interest rate increases versus some alternative, such as raising margin lending requirements.
Federally mandated margin requirements were created by the Securities Act of 1934, with the objective of reducing speculative excesses. An important concern at that time was that speculative volatility could be exacerbated by a pyramiding-depyramiding process caused by margin loans. When margin borrowing is prominent, then if the stock market rises, speculators will be able to borrow against the increased value, thereby buying yet more stocks, creating a potential upward feedback. If the stock market falls, then speculators will be forced to sell some of their stocks to meet margin calls, thereby forcing the market down further. This downside feedback had been viewed as an important cause of the stock market crash of 1929. With margin credit approaching 30% of market capitalization in 1929, it was reasonable to think that such feedback would be prominent. This feedback process could be mitigated, it was thought, if margin requirements limit the amount of borrowing. Another important concern was to help protect the investor from taking on an overly risky position in the markets.
This pyramiding-depyramiding effect of margin credit must be much smaller today, when margin credit is only around 1.5% of market capitalization. The decline in margin borrowing since 1929 is not directly due to the margin requirements. The margin requirements are binding for only a small proportion of investors today. Instead, financial innovations of various sorts have reduced investors reliance on margin credit.
Federal margin requirements were created before a great many financial innovations, before the advent of financial futures and exchange-traded options, before the expansion of home equity lending, before credit cards. All of these allow retail investors to take effectively leveraged, more speculative positions, in the stock market without using margin credit. Generally, these alternative methods of leveraging positions do not have the same feedback effect that is represented in the pyramiding-depyramiding model, since there are no new opportunities to borrow when stock prices increase, and there are no margin calls when stock prices decrease.
Studies of the impact of changed margin requirements on stock market volatility have not conclusively found an effect of margin requirements on market volatility, see the 1997 paper by
Paul Kupiec which surveys the literature.. I find these studies not entirely convincing. Not knowing what brought on changes in margin requirements, we have no way of knowing whether the apparent ineffectiveness of margin requirements was that they were kept high during periods that would otherwise have been periods of exceptionally high volatility. The requirements would thus look ineffective even if they were in fact effective.
While the pyramiding-depyramiding motivation for margin requirements would suggest that margin requirements should be raised whenever such pyramiding seems somehow more likely to be a problem (as might be suggested by high volatility in the markets) and lowered when less of a problem, (as suggested perhaps by low volatility in the markets), in fact adjustments that were made to margin requirements by the Fed in the past appear often to have been justified in terms of a general policy of reducing margin requirements whenever the market is falling and raising them whenever it is rising.
Margin requirements for stocks were first imposed on October 15, 1934, when initial margin was set at 45%. Following that date, margin requirements for stock purchases were changed twenty-two times, until the last time the requirements were changed on January 3, 1974. Margin requirements have been as high as 100%, meaning that no margin credit was allowed at all, and as low as 40%. The history of the margin requirements can be seen in the accompanying figure.
When margin requirements were changed, the justification offered was often to restrain speculation, either upwards or downwards. Thus, the Fed apparently viewed margin requirements as a tool to stabilize the stock market. The Fed apparently also viewed margin requirements as in its arsenal of tools relevant to its broader mission of restraining inflation in the consumer price index and of stabilizing the economy.
If the Fed was using margin requirements to attempt to stabilize markets, then one might think that the Fed would loosen margin requirements when stock prices were low (relative to some measure such as earnings) and raise margin requirements when stock prices were high. However, this is not mostly what the Fed has done. The figure shows also the price earnings ratio, to compare with the margin requirement on the same figure. The price earnings ratio, from my new book Irrational Exuberance, is the Standard and Poor Composite Index, corrected for inflation by dividing by the consumer price index, divided by a ten-year moving average of lagged Standard and Poor earnings, also corrected for inflation by dividing by the consumer price index. Such a price earnings ratio is more revealing than conventional price earnings ratios over this long time period. As one can see, there is only a very weak relation between margin requirements and the price earnings ratio. For example, when the margin requirement was at its highest level, in 1946. the price-earnings ratio was only about average. The margin requirement jumps up and down massively with much greater frequency than does the price-earnings ratio.
A closer determinant of the margin requirement is the recent past price change in the stock market, as can be seen by closely examining the figure. In each of the twelve times margin requirements were raised, stock prices had risen in the previous six months (using monthly
Standard and Poor Composite data, between the month seven months before the margin increase and one month before the margin increase.) In nine of the ten times margin requirements were lowered, stock prices had declined in the previous six months. Thus, the Fed used margin requirements to lean against recent price changes, not to lean against historically high or historically low markets. This explains why the margin requirement series shown has so much more of a high frequency character, changing rapidly up and down more strongly than does the price earnings ratio.
The margin requirements that we have today were last adjusted by the Federal Reserve in 1974. It would seem highly unlikely that, if these margin requirements were set at the "optimal" level in 1974, they are still at the right level in 2000. If the Fed does not believe in margin requirements, then it might seem that the initial requirements should be set at zero, not left at 50%. However, maintaining margin requirements does serve as a symbol of concern about the dangers of investing on margin to individual investors, and eliminating them might be interpreted as a signal of lack of concern about dangerous speculation.
Despite this ambiguity as to the "optimal" margin requirement, I think that there may be some small advantage to raising margin requirements today as a symbolic gesture, if the Fed were inclined to do so as a way of dealing with an overheated market. Judging from Alan Greenspans testimony, he is not about to advise such a thing since he will take no position on whether the stock market is overpriced. He has said, in his recent Humphrey-Hawkins testimony, that his opinion about the level of the market is no better than anyone elses.
Greenspan often appears to believe that the market may be overpriced, but he has always stopped short of saying that it is. His famous "irrational exuberance" speech of December 5, 1996, he in fact merely posed the question whether markets might be overpriced. In his recent testimony, he has instead maintained that his concern with the stock market is only whether it is feeding a wealth effect that might ultimately contribute towards an inflationary environment.
In my forthcoming book Irrational Exuberance, I argue that the market currently appears to be overpriced, and is quite likely to disappoint in coming years.. Moreover, I argue that there is a role for leaders such as the chairman of the Federal Reserve to express their opinions about market overpricing or underpricing. Those leaders who have opinions about such overpricing should not be shy about expressing them, since it is partly through such expression that individuals gain a sense of how to price assets. The Federal Reserve in our society represents a force of moral authority and careful research unaligned with commercial interests.
In summary, I feel that it may be useful if the Fed were today to raise margin requirements as a symbolic gesture to emphasize that the market may be overheated. While the Fed should probably not change margin requirements as often as it used to do before 1974, on the rare occasions, such as today, when the stock market is very distorted by speculation, some action to change the margin requirements might have a slightly beneficial effect on the markets.
Kupiec, Paul, "Margin Requirements, Volatility, and Market Integrity: What Have We Learned Since the Crash?" Federal Reserve Board, Finance & Economics Discussion Series 1997-22, http:www.federalreserve.gov/pubs/feds/ 1997/199722/1 99722pap.pdf.
Shiller, Robert, Irrational Exuberance, Princeton University Press, 2000 (available for order now on online services, in bookstores by early April.)