By Edward Renshaw,
Professor of Economics,
State University of New York at Albany.
I use these working papers and reflections on the stock market as a supplement to courses in money and banking and financial economics and as a means of trying to do a better job of managing my own portfolio.
This introduction is composed of two parts:
Is it time to bail out of the stock market? This featured table, Table IN.1S, was put in place on June 27, 1996. It is summarized after the following highlights.
The best years to have owned common stock have been years containing a recessionary trough in economic activity (Table 1.1S). Since the stock market crash of 1929 the financial returns associated with the S&P composite stock price index have been over thirty percent in these years.
While it is easy to devise market timing switching models that would have generated impressive profits in the past, the out of sample performance of these models has been rather dismal (Essay 1S.)
It is easier to find statistical indicators that can identifying good years to own common stock than to discover reliable crash indicators (Essays 1S and 4S).
When low dividend yields are accompanied by tight money or rising interest rates, however, one has all of the ingredients that are necessary for the type of stock market correction which followed the Fed's decision on February 4, 1994 to raise short term interest rates to slow the growth of the economy to a more sustainable pace and prevent an acceleration of the inflation rate (Essay 1S).
In the post World War II period economic recessions have been of such short duration that by the time a majority of forecasters are convinced that the economy is in a recession, it is probably time to begin looking for bargains in the stock market. At this point in the business cycle you don't need leading indicators to identify good years to have invested in the stock market. Some of the coincident indicators that will soon be published by the Conference Board will also be helpful at verifying that there has been a downturn in business activity (Essay 1S).
Large spreads between the yields on low grade corporate bonds and one-month Treasury bills are of some value in predicting financial returns in both the stock and bond markets. They are not very reliable in warning investors about the possibility of a stock market correction in the midst of prosperity, however (Table 1.3S).
The safest way to profit from nonrandomness is to buy stock on margin after there has been a large decline in stock prices and numerous indicators of both a cyclical and value oriented nature are pointing in the direction of lower interest rates and a rapid recovery of stock prices (Essay 1S).
There is evidence to support the hypothesis that the stock market is more stable now than it used to be. The intriguing question is why? Among the suspects are a more stable economy, its changing structure, the invention of portfolio balance models and the globalization of the securities market (Essay 2S).
The financial returns for the 16 years with the smallest high/low trading ranges for the S&P index have all been positive (Table 2.1S).
On September 12, 1995 the S&P composite stock price index registered a new historic closing high of 576.51 and established another all time record of 1,156 trading days for a series of new highs without an intervening decline of nine percent or more. The previous record for a series of new highs without a decline of this magnitude is associated with the 409 day trading period from January 21, 1985 to September 4, 1986 (Table 2.2S). That stability record was temporarily marred by a cumulative decline of 9.4 percent over the next 17 trading days and was eventually followed by the biggest one day decline in the S&P index of 20.5 percent on October 19, 1987.
One reason for hoping that the stock market will continue to be more stable in the future than was the case on the average from 1930-79 is evidence in support of the hypothesis that business expansions are lasting longer than they used too (Essay 2S). The business expansion which began in March 1991 began to increase the average duration of post World War II expansions in June 1995.
Evidence in support of a more stable economy that might spill over and help to stabilize the stock market is most apparent when one examines cyclical fluctuations in nonagricultural payroll employment and its major components: employment in goods producing industries and service industries (Table 2.5S).
The globalization of the world's securities markets has dramatically lowered yearly spreads for Dec.-Dec. percentage changes in the stock market indexes for the more industrial countries of the world (Table 2.6S).
The unanswered question is whether inexperienced investors and money managers have taken control of the US stock market and are setting the stage for an old fashioned boom to be followed by a crash that will end the most prolonged new high bull market in the history of the S&P composite stock price index (Essay 2S).
What should young people invest in? The first priority should be a good education (Essay 3S).
It is not true in the short run and there is no reason why it should necessarily be the case that stocks will outperform bonds in the long run (Essay 5S). Every reputable study since the publication ofCommon Stocks as Long-Term Investment by Edgar L. Smith in 1925, however, has been forced to conclude that representative stock market averages have outperformed bonds and other types of financial assets over long periods of time (Table 3.3S).
In his book,Winning on Wall Street, Martin Zweig's main advice to investors is, "Don't Fight the Fed." That advice is even more applicable to investors in fixed income securities (Table 3.4S).
In thinking about what might happen to short term interest rates and stock prices in the years ahead one is well advised to remember that the Fed's interest rate reaction function has become more forward looking in recent years (Essay 3S).
The nice thing about good year indicators for the stock market is that you don't need a computer or stock market guru to interpret their predictions. At the one year horizon, it is only the more extreme values of stock market indicators that seem to be of much forecasting value (Essay 4S and Table 7.5S).
Since the beginning of World War II the financial returns for the S&P stock price index have always been positive during any year when M1 expressed in constant dollars was allowed to increase by 1.6 percent or more (Essay 4S).
Since 1928 there has (so far) never been a dividend yield for the S&P stock price index under 3.4 percent at the end of any year containing a recessionary peak in business activity (Table 5.2S).
History suggests that investors should be cautious whenever the monthly dividend yield for the S&P stock price index has either established a new record low or tied the preceding low (Table 5.2S). The stock market correction of 1994 turned out to be rather mild and never did push the monthly yield for the S&P index above three percent. The unanswered question is whether investors will be content with a yield of less than three percent when it becomes clear that the US economy has slipped into another recession.
In 1979 Nobel Laureate Franco Modigliani and Richard Cohn criticized financial analysts for using a nominal rate of interest to capitalize corporate earnings and suggested that investors had been systematically undervaluing the stock market by 50 Percent. "In the presence of inflation," they argued, "one properly compares the cash return on stocks, not with the nominal return on bonds, but with the real return on bonds." Since this article was published most of the superiority of stocks over bonds can probably be attributed to a dramatic decline in real risk premiums (Table 5.3S).
Lower dividend yields and smaller risk premiums should help to narrow the longer run return differential for stocks and bonds. If the stock market is perceived as not being more risky than the bond market and continues to provide many investors with tax deferment advantages, it wouldn't be very surprising if the prices paid for equities eventually get bid up to the point of providing investors with a smaller before tax return, on the average, than long term bonds (Essay 5S).
The tendency for major bull markets to "fade away" instead of ending on a big bubble, or a spectacular series of new historic highs makes it very difficult to identify bull market peaks in close proximity to their occurrence (Table 6.1S).
Since the mild recession of 1960-61 the S&P composite stock price index has consistently lost more of its value after a recessionary peak in business activity as defined by the National Bureau of Economic Research than before the peak (Essay 6S).
Lots of momentum in the second quarter of a calendar year containing a new high bull market has so far always spilled over into the third and fourth quarters (Table 7.2S).
Hope seems to spring eternal before a presidential election (Table 7.4S).
Stocks and long term bonds have performed better, on the average, during years with low or declining inflation than during years of accelerating inflation (Tables 10.1S and 10.2S).
The stock market is one of those games where records are made to be broken (Essay 9S). Can there be a bad October without a disappointing September? See Table 8.1S.
The answer this question is: probably not.
In recent years financial analysts have begun to appreciate that fluctuations in representative stock market indexes are not exactly random. Since the beginning of World War II all of the 12 calendar year declines in the S&P composite stock price index that were not offset with dividends received have occurred either in a post presidential election year or in the following year.
Hope springs eternal, it would seem, in the two years preceding a presidential election. After experiencing a net decline of 1.5 percent in response to the Fed's preemptive strike against the possibility of accelerating inflation in 1994 the S&P composite stock price index was up an impressive 47.7 percent as of May 24, 1996.
Calendar year "anomalies" are also associated with the occurrence of new historic highs for this index. There have been 13 occasions since 1927 when the S&P index recorded one or more new historic highs during May and then experienced a cumulative decline amounting to two percent or more before recovering to another new high. For the 12 resolved cases an investor wouldn't have had to wait more than 41 trading days before this type of correction was over or more than 94 trading days before a badly timed purchase of a portfolio resembling the S&P index could have been liquidated at a profit. (See Table IN.1S).
A similar pattern of short lived corrections after new historic highs can be observed for the months of March, April and June. The logic behind these short lived corrections is speculative. Some analysts have suggested that there is nothing like the prospect of a good old summer time for maintaining a May time spirit of "new high" optimism with regard to the stock market.
When changes in stock prices are examined from a monthly perspective, however, it becomes clear that the termination of summer vacations and the start of a new school year has often been a poor time to own equities. Since the S&P index was extended back to 1928 it has declined 39 times and increased only 29 times during September.
After the prompt recoveries from the May time corrections of 1929, 1959, 1961, 1972 and 1986 investors in a portfolio similar to the S&P index would have experienced September losses of 4.9. 4.6, 2.0, .5 and 8.5 percent, respectively. These declines would have more than wiped out September gains of 2.4, 2.9, 3.2, 3.3, 3.9 and 1.0 percent in 1928, 1964, 1965, 1967, 1968 and 1983.
By the end of these correction years, however, there were only three years (1929, 1983 and 1986) when the S&P index didn't close above its peak value for May. The post election year of 1929 ended what is often regarded as the greatest bull market in recorded history.
After its second peak at a value 166.21 on May 25, 1983 the S&P index: bounced back to a new high of 172.65 on October 10; then drifted down to a year end value of 164.93; snapped back to 169.28 on January 9, 1984; retreated to a low of 147.82 on July 24, 1984; and then soared upward to a new all time peak value of 336,66 on August 25, 1987 that was only temporarily marred by a one month plunge of 8.5 percent during September 1986.
The data in column (2) of Table IN.1S, however, are not inconsistent with the hypothesis that the U.S. stock market may be more stable now than was formerly the case. From 1928-67 there were four new high corrections beginning in May in the minus 6.5 to 10.3 percent range and only three smaller corrections in the minus 2.2 to 4.4 percent range. The largest correction since 1967 has (so far) only been equal to 4.4 percent.
The more worrisome aspect to the new high corrections which have begun in May is that they have often been the prelude to more serious crashes. For the 12 resolved cases one could have always sold an index fund tracking this index after it had recovered to a new high from a May time correction and then repurchased it at a discount of at least nine percent at some point in the future.
The unanswered question is whether the stock market will continue to follow this pattern. While it is easy to develop systems for moving in and out of the stock market which would have been superior to a policy of buying and holding a portfolio similar to the S&P index, the out of sample performance of market timing strategies has not been very encouraging. The changing character of the U.S. economy and its financial markets makes it very difficult to answer the bail out question.
The worst follow-on crashes in Table IN.1S are associated with the great depression of the 1930s and the prolonged recession of 1973-75. The good news with regard to economic recessions is that they are occurring less frequently and are not having as devastating an impact on the stock market as these two recessions.
Investors, moreover, have begun to appreciate that the S&P index is classified as being a leading economic indicator and that years containing a recessionary trough in business activity have been the best years to have owned common stock, on the average, in the post World War II period.
Stock market crashes in the midst of prosperity can also provide some wonderful buying opportunities. During the last 20 years an investor would have been better off, on the average, to have purchased an index fund tracking the S&P index after a one-day drop in the index amounting to one percent or more than to have followed a dollar cost averaging strategy of buying stock at the end of each trading day or a trend chasing strategy of only buying stock after a one day gain for this index amounting to one percent or more.
The fundamentals may also be better. Since the short lived recession of 1990-91 the earnings associated with the S&P index have more than doubled. Many corporations are not only flush with cash but are more inclined to repurchase their own shares, rather than pay dividends that are subject to an additional tax. This in conjunction with a maturing baby boom that is more aware of the fact that stocks have outperformed bonds over long periods of time may have reduced the probability that the stock market correction which began in May of this year will soon be followed by a major crash.
Table 4.2S. Earnings, Dividends and Financial Returns for the S&P Index. Table 4.3S. Annual High-Low Ratios for the S&P Stock Price Index.
Number of Trading
Days from Peak to S&P Value End of
New High S&P Peak Cumulative ----------------- ----------------
Peak Date Value Decline (%) Trough New High Sept. Year
(1) (2) (3) (4) (5) (6)
5/14/28 20.44 -10.3 28 59 21.37 24.35
5/04/29 26.48 - 8.6 19 40 30.16 21.45*
5/29/59 58.68 - 4.0 7 23 56.88* 59.89
5/17/61 67.39 - 4.4 41 54 66.73* 71.55
5/12/64 81.16 - 3.1 18 31 84.18 84.75
5/13/65 90.27 - 9.6 31 94 89.96* 92.43
5/18/67 94.58 - 6.5 19 58 96.71 96.47
5/08/68 98.91 - 2.5 8 17 102.67 103.86
5/26/72 110.66 - 4.4 37 50 110.55 118.05
5/06/83 166.10 - 2.5 9 13 166.07* 164.93*
5/25/83 166.21 - 2.9 9 14 166.07* 164.93*
5/29/86 247.98 - 3.4 8 19 231.32* 242.17*
5/24/96 678.51 - 2.4e 17e ? ? ?
e represents an estimate as of June 26, 1996 which might be exceeded.
*identifies cases where the end of period value for the S&P index was less than at its May peak in column (1).
Go on to Essay 1S:
Go to Renshaw Home Page