Edward Renshaw
Professor of Economics
State University of New York at Albany
There is an old saying that prediction is very difficult, especially when it pertains to the future. There are times, however, when an investor in common stocks might be able to sleep more comfortably by keeping a wary eye on economic and financial indicators with a long history of being able to identify good years to have been in the market. In Table 4.1S we up-date a set of indicators that "would have worked well" from 1942-1989 and have (so far) continued to track the performance of the S&P composite stock price index rather well in the 1990s.
The nice thing about these indicators 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 and bond market indicators that seem to be of much forecasting value (Renshaw 1993).
The best years to have owned common stock since the crash of 1929 have been years containing a recessionary trough in economic activity as defined by the National Bureau of Economic Research. See Table 1.1S.
Economic recessions also have the distinction of being the only sure cure for inflation. Since the beginning of World War II there have been nine years when the Dec.-Dec. growth rate for the all item CPI was at least 1.4 percentage points less than the year-to-year increase in the CPI inflation rate. The following year financial returns for the S&P index have ranged from a low of only 5.4 percent in 1948 to a high of 31.0 percent in 1955 and have an average value of 20.4 percent. See Table 4.1S.
It should be noted, however, that some much needed stability has been imparted to the home ownership component of the CPI cost of shelter index as a result of a 1983 shift to a rental equivalent measure of home owner costs. The old procedure was to measure the price of new houses, mortgage interest rates and other cost elements that are borne by those relatively few families that are fortunate enough to be able to afford a new home during the month in question. This made the cost of home ownership very sensitive to changes in monetary policy. Since the change over to a rental equivalent value for the home ownership component we have not had any declines in the CPI inflation rate that were steep enough to trigger a good year signal for the S&P index.
The demise of CPI inflation signals may also be related to a more aggressive effort on the part of the Fed to prevent inflation from getting out of hand. Beryl Sprinkel, who was Chairman of the President's Council of Economic Advisers from 1985-89, was among the first economists to publicize a positive relationship between changes in the money supply and changes in stock prices. 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 (Table 11.1S). The financial returns in the following year have also been positive if the growth rate for real M1 was allowed to accelerate by four percentage points or more--as has sometimes been the case during severe recessions and periods of slow economic growth when the inflation rate was declining.
Evidence in support of a non random pattern to stock prices is easy to overlook because it usually involves first or second differences in a statistical indicator. The validity of this conclusion can be illustrated by examining large changes in the unemployment rate. There have been seven years since the end of World War II when the unemployment rate increased by 1.4 percentage points or more and each of these recessionary years was followed by a positive return for the S&P index. There have been nine years since 1940 when the unemployment rate declined by one percentage point or more and each of these rather vigorous years of recovery from a recession inspired enough investor confidence to be followed by a positive return for the S&P index.
The stock market is classified as being a leading economic indicator. In the post World War II period economic recessions have been of such short duration that by the time the Commerce Department's index of leading economic indicators declined by 2.2 percent or more, on a Dec.-Dec. basis, it was time to start looking for bargains in the stock market. Vigorous recoveries in the index of leading economic indicators amounting to six per or more have also inspired enough investment enthusiasm to push the S&P index higher in the following year (Table 7.5S).
While crashes in the midst of prosperity have tarnished the image of stock prices as a recessionary indicator, they will sometimes enable investors to acquire equities at bargain prices. Those years when the index of leading economic indicators increased and the average monthly value for the S&P index was down at least 1.6 percent from its preceding cyclical monthly high have (so far) always been followed by a positive financial return for the S&P index.
This degree of stock market "downness", when other indicators are signaling the continuation of a business expansion, deserves special mention since at the end of 1994 (when the S&P closed at 459.27) it was the only indicator in Table 4.1S that was signaling a good year to own common stock during 1995. There have been two other years when this type of divergence was a lone indicator pointing in the direction of a good year. Both of these years (1964 and 1987) were followed by double digit returns for the S&P index. They may help to explain why the S&P index, after closing out 1994 down 4.7 percent from its February 2 high of 482.00 quickly rebounded to a new historic high on Feb. 14, 1995.
There is another type of general purpose stock market indicator that is sometimes useful in helping one to benefit from stock market crashes in the midst of prosperity. Since 1941 the S&P index has always registered a positive return after any year when its average dividend yield increased by 24 basis points or more. An increase in the dividend yield of this magnitude can be caused by an increase in dividends and/or a decline in stock prices.
Dividend growth rates are adversely affected by economic recessions. Since the beginning of World War II economic recessions have been of such short duration that investors could have always anticipated a positive return for the S&P index after any decline in the dividend growth rate for the S&P index amounting to 3.2 percentage points or more (Table 4.2S). The financial returns for the S&P index have also been positive after any year when its associated earnings declined by at least six percent.
In recent years analysts have begun to appreciate that the financial returns associated with the S&P index have either varied too much or too little on a year-to-year basis to be consistent with a random walk. The implication is that stock returns may be subject to the phenomenon of "mean reversion" or a tendency for stocks that have enjoyed high (low) returns to exhibit lower (higher) returns in the future.
One of the more interesting ways to try to take advantage of this phenomenon is to compute the annual high/low ratios for the S&P index and purchase stocks after there has been a downside reversal for this ratio ( Table 4.3S). Since 1941 the following year financial returns for the S&P index have (so far) always been positive after a downside reversal. Why this signal should work so well is speculative. The extensive overlap of high/low ratio reversal signals in Table 4.1S with other good year indicators, however, should help to bolster one's confidence in this signal.
For the 40 years when one or more of the indicators in Table 4.1S were pointing in the direction of a good year to own stock the average financial return has been equal to 19.3 percent. This can be compared to an average loss of five percent, for the 13 years identified by a hatch mark in the last column of Table 4.1S, when none of the indicators were signaling a good time to own stock. It should be noted, however, that the S&P index is more richly valued now in terms of dividends received, than was the case from 1942-83, and that good year returns of less than ten percent have been occurring more frequently than was formerly the case.
It should also be emphasized that systems for identifying the best times to be in the stock market often break down and fail to explain the future as well as the past. The stock market indicators in Table 4.1S were identified by the author in the 1980s and eventually published in The Practical Forecasters' Almanac. They are easy to update using information in the yellow pages or cyclical indicator section of the Survey of Current Business and Standard and Poor's Security Price Index Record.
The signals, which are identified in the footnotes to Table 4.1S, have provided a more reliable indication of what has happened to the market (so far) in the 1990s, I believe, than most of the tips from market-timing advisers--which have a "spotty record" for recent years (Damato 1994). If you are going to play the stock market game, it is nice to at least have a score card and make some effort to understand how this market has behaved in relation to economic and financial indicators that are often publicized by the news media upon release by government agencies.
Damato, Karen (1994). "Market-Timing Advisers Have Spotty Record for Recent Years," The Wall Street Journal, April 28, C1 and C7.
Renshaw, Edward (1992), Editor. The Practical Forecasters' Almanac(Irwin Professional Publishing).
-----, (1993). "Modeling the Stock Market for Forecasting Purposes," Journal of Portfolio Management, 20(Fall), 76-81.
Year (1) (2) (3) (4) (5) (6) (7) Following Year
Return
1941 U D HLR 19.2
1942 CPI M U D E 25.7
1943 CPI U HLR E 19.3
1944----------------------------------------------------- 35.7#
1945----------------------------------------------------- - 7.8#
1946 U 5.5
1947 CPI M D HLR 5.4
1948 CPI L D 17.8
1949 M U D 30.5
1950 L HLR 23.4
1951 CPI U L D E 17.7
1952----------------------------------------------------- - 1.2#
1953 L 51.2
1954 CPI U L 31.0
1955 U HLR 6.4
1956----------------------------------------------------- -10.4#
1957 L D 42.4
1958 M U L D E 11.8
1959 U HLR .3
1960 L D 26.6
1961----------------------------------------------------- - 8.8#
1962 U L D 22.5
1963 HLR 16.3
1964 L 12.3
1965----------------------------------------------------- -10.0#
1966 L D 23.7
1967 M D HLR 10.8
1968----------------------------------------------------- - 8.3#
1969 L HLR 3.5
1970 U D E 14.1
1971 L HLR 18.7
1972----------------------------------------------------- -14.5#
1973----------------------------------------------------- -26.0#
1974 L D 36.9
1975 CPI M U L D HLR E 23.6
1976----------------------------------------------------- - 7.2#
1977 L D 6.4
1978 U L D 18.4
1979 L HLR 31.5
Year (1) (2) (3) (4) (5) (6) (7) Following Year
Return
1980----------------------------------------------------- - 4.8#
1981 CPI L HLR 20.4
1982 CPI M U D E 22.3
1983 L HLR 6.0
1984 U D 31.1
1985 M E 18.5
1986 M HLR 5.7
1987 L 16.3
1988 L D HLR 31.2
1989----------------------------------------------------- - 3.1#
1990 L D HLR E 30.0
1991 M D E 7.4
1992 M HLR 9.9
1993----------------------------------------------------- 1.3#
1994 L ?
CPI represents years when the Dec.-to-Dec. growth rate for the CPI was at least 1.4 percentage points less than the year-to-year growth rate for the CPI.
M represents years when the first differences in the Dec.-to-Dec. growth rates for the real money supply are equal to four percentage points.
U represents years when the unemployment rate either increased by 1.4 percentage points or declined by at least one percentage point.
L represents years when the revised index of leading economic indicators increased at least 6 percent, declined by at least 2.2 percent or was positive when the December down ratio for the S&P index was less than .984.
D represents years when the dividend growth rate for the S&P index declined at least 3.2 percentage points or when the average dividend yield increased by at least 24 basis points.
HLR represents years when there was a downside reversal for the yearly high- low ratio for the S&P index.
E represents years when the earnings for the S&P index declined by at least six percent.
# identifies financial returns following years with no good year signals.
Source: The Practical Forecasters' Almanac(Irwin Professional Publishing, 1992), Table 2.91. Some of the signals for the Commerce Department's index of leading economic indicators have been adjusted to conform with changes in the index discussed in the October 1993 issue of the Survey of Current Business (SCB). The data needed to update this Table can be obtained from the cyclical indicator section of the SCB and S&P's Security Price Index Record.
Year Earnings Dividends Payout % Change Difference Financial
(dollars) Ratio Dividends % Points Return
(1) (2) (3) (4) (5) (6)
1942 1.03 .59 57.3 -16.9@ -22.9 19.2
1943 .94 .61 64.9# 3.4 20.3 25.7*
1944 .93 .64 68.8# 4.9 1.5 19.3
1945 .96 .66 68.8 3.1 - 1.8 35.7
1946 1.06 .71 67.0 7.6 4.5 - 7.8
1947 1.61 .84 52.2 18.3 10.7 5.5
1948 2.29 .93 40.6 10.7 - 7.6 5.4
1949 2.32 1.14 49.1# 22.6 11.9 17.8*
1950 2.84 1.47 51.8# 28.9 6.3 30.5
1951 2.44 1.41 57.8# - 4.1@ -33.0 23.4
1952 2.40 1.41 58.8 .0 4.1 17.7*
1953 2.51 1.45 57.8 2.8 2.8 - 1.2
1954 2.77 1.54 55.6 6.2 3.4 51.2
1955 3.62 1.64 45.3 6.5 .3 31.0
1956 3.41 1.74 51.0# 6.1 - .4 6.4
1957 3.37 1.79 53.1 2.9 - 3.2 -10.4
1958 2.89 1.75 60.6# - 2.2@ - 5.1 42.4*
1959 3.39 1.83 54.0 4.6 6.8 11.8*
1960 3.27 1.95 59.6# 6.6 2.0 .3
1961 3.19 2.02 63.3# 3.6 - 3.0 26.6
1962 3.67 2.13 58.0 5.4 1.8 - 8.8
1963 4.02 2.28 56.7 7.0 1.6 22.5
1964 4.55 2.50 54.9 9.6 2.6 16.3
1965 5.19 2.72 52.4 8.8 - .8 12.3
1966 5.55 2.87 51.7 5.5 - 3.3 -10.0
1967 5.33 2.92 54.8# 1.7 - 3.8 23.7*
1968 5.76 3.07 53.3 5.1 3.4 10.8*
1969 5.78 3.16 54.7 2.9 - 2.2 - 8.3
1970 5.13 3.14 61.2# - .6@ - 3.5 3.5
1971 5.70 3.07 53.9 - 2.2@ - 1.6 14.1*
1972 6.42 3.15 49.1 2.6 4.8 18.7
1973 8.16 3.38 41.4 7.3 4.7 -14.5
1974 8.89 3.60 40.5 6.5 - .8 -26.0
1975 7.96 3.68 46.2# 2.2 - 4.3 36.9
1976 9.91 4.05 40.9 10.1 7.9 23.6*
1977 10.89 4.67 42.9 15.3 5.2 - 7.2
1978 12.33 5.07 41.1 8.6 - 6.7 6.4
1979 14.86 5.65 38.0 11.4 2.8 18.4*
Year Earnings Dividends Payout % Change Difference Financial
(dollars) Ratio Dividends % Points Return
(1) (2) (3) (4) (5)n (6)
1980 14.82 6.16 41.6# 9.0 - 2.4 31.5
1981 15.36 6.63 43.2 7.6 - 1.4 - 4.8
1982 12.64 6.87 54.4# 3.6 - 4.0 20.4
1983 14.03 7.10 50.6 3.3 - .3 22.3*
1984 16.64 7.53 45.3 6.1 2.8 6.0
1985 14.61 7.90 47.6# 4.9 - 1.2 31.1
1986 14.48 8.28 57.2# 4.8 - .1 18.5
1987 17.50 8.81 50.3 6.4 1.6 5.7
1988 23.76 9.73 41.0 10.4 4.0 16.3
1989 22.87 11.05 48.3# 13.6 3.2 31.2
1990 21.34 12.10 56.7# 9.5 - 4.1 - 3.1
1991 15.97 12.20 76.4# .8 - 8.7 30.0*
1992 19.09 12.38 64.9 1.5 .7 7.4*
1993 21.88 12.58 57.5 1.6 .1 9.9
1994 30.63 13.18 43.0 4.8 3.2 1.3
(5)n. The first differences in the growth rates for dividends in column (4).
#Years when the payout ratio increased by 2.3 percentage points or more. The financial returns in these years have usually been positive.
@Years when the dividend growth rate was negative. The financial returns in these years and the following year have been positive.
*Financial returns following negative differences in column (5) of 3.2 percentage points or more.
Source of basic data: Standard and Poor's Security Price Index Record.
Values for the S&P Index
Year ------------------------ High-Low Financial
High Low Close Ratio Return
(1) (2) (3) (4) (5)
1928 24.35 16.95 24.35 1.437 41.9
1929 31.92 17.66 21.45 1.807 - 7.9
1930 25.92 14.44 15.34 1.795 -23.9
1931 18.17 7.72 8.12 2.354 -41.7*
1932 9.31 4.40 6.89 2.116 - 9.0
1933 12.20 5.53 10.10 2.206 53.0*
1934 11.82 8.36 9.50 1.414 - 1.5
1935 13.46 8.06 13.43 1.670 46.3*
1936 17.69 13.40 17.18 1.320 33.3
1937 18.68 10.17 10.55 1.837 -33.9*
1938 13.79 8.50 13.21 1.622 30.0
1939 13.23 10.18 12.49 1.300 - .8*
1940 12.77 8.99 10.58 1.420 - 9.9
1941 10.86 8.37 8.69 1.297 -11.2
1942 9.77 7.47 9.77 1.308 19.2*
1943 12.64 9.84 11.67 1.285 25.7
1944 13.29 11.56 13.28 1.150 19.3*
1945 17.68 13.21 17.36 1.338 35.7
1946 19.25 14.12 15.30 1.363 - 7.8
1947 16.20 13.71 15.30 1.182 5.5
1948 17.06 13.84 15.20 1.233 5.4*
1949 16.79 13.55 16.76 1.239 17.8
1950 20.43 16.65 20.41 1.227 30.5
1951 23.85 20.69 23.77 1.153 23.4*
1952 26.59 23.09 26.57 1.152 17.7
1953 26.66 22.71 24.81 1.174 - 1.2
1954 35.98 24.80 35.98 1.451 51.2
1955 46.41 34.58 45.48 1.342 31.0
1956 49.74 43.11 46.67 1.154 6.4*
1957 49.13 38.98 39.99 1.260 -10.4
1958 55.21 40.33 55.21 1.369 42.4
1959 60.71 53.58 59.89 1.133 11.8
1960 60.39 52.30 58.11 1.155 .3*
1961 72.64 57.57 71.55 1.262 26.6
1962 71.13 52.32 63.10 1.360 - 8.8
1963 75.02 62.69 75.02 1.197 22.5
1964 86.28 75.43 84.75 1.144 16.3*
1965 92.63 81.60 92.43 1.135 12.3
1966 94.06 73.20 80.33 1.285 -10.0
1967 97.59 80.38 96.47 1.214 23.7
1968 108.37 87.72 103.86 1.235 10.8*
1969 106.16 89.20 92.06 1.190 - 8.3
1970 93.46 69.29 92.15 1.349 3.5*
Values for the S&P Index
Year ------------------------ High-Low Financial
High Low Close Ratio Return
(1) (2) (3) (4) (5)
1971 104.77 90.16 102.09 1.162 14.1
1972 119.12 101.67 118.05 1.172 18.7*
1973 120.24 92.16 92.55 1.305 -14.5
1974 99.80 62.28 68.56 1.602 -26.0
1975 95.61 70.04 90.19 1.365 36.9
1976 107.83 90.90 107.46 1.186 23.6*
1977 107.00 90.71 95.10 1.180 - 7.2
1978 106.99 86.90 96.11 1.231 6.4
1979 111.27 96.13 107.94 1.157 18.4
1980 140.52 98.22 135.76 1.431 31.5*
1981 138.12 112.77 122.55 1.225 - 4.8
1982 143.02 102.42 140.64 1.396 20.4*
1983 172.65 138.34 164.93 1.248 22.3
1984 170.41 147.82 167.24 1.153 6.0*
1985 212.02 163.68 211.28 1.295 31.1
1986 254.00 203.49 242.17 1.248 18.5
1987 336.77 223.92 247.08 1.504 5.7*
1988 283.66 242.63 277.72 1.169 16.3
1989 359.80 275.31 353.40 1.307 31.2*
1990 368.95 295.46 330.22 1.249 - 3.1
1991 417.09 311.49 417.09 1.339 30.0*
1992 441.28 394.50 435.71 1.119 7.4
1993 470.54 429.05 466.45 1.097 9.9*
1994 482.00 438.92 459.27 1.098 1.3
*Financial return following a downside reversal for the yearly high-low ratio in column (4).
Source of basic data: Standard and Poor's Securities Price Index Record.
Go on to Essay 5S:
Return to the Introduction