"October. This is one of the peculiarly dangerous months to speculate in stocks in. The others are July, January, September, April, November, May, March, June, December, August, and February." Mark Twain, "Pudd'nhead Wilson's Calendar", 1894
Which month is the most peculiarly dangerous for equity investment? Which month is the best?
To get some answers, I retrieved the closing weekly prices for the Toronto Stock Exchange 300 Index from the Canadian Asset Allocation Strategies database (WiseWords Publications, 1993). This database contains 30 fields of weekly data covering Canadian and other markets since January 1970. Weekly data, unfortunately, does not conform precisely to calendar months, so I chose my "months" as the 4-week periods ending on the first Friday of the following month. The results shown in Table1 and subsequent tables are therefore for holding periods of 28 days, not full months.
Table 1 shows the results covering a period of 25 years and 8 months, from January 1970 to August 1995, for the TSE300 Index. The Toronto Stock Exchange introduced the TSE300 in January 1977; for prior years, the database uses synthetic data prepared by the TSE. Over the 25+ years, the average 4-week return of the TSE300 has been 0.56%, with a standard deviation of 4.40%. Maximum return in a 4-week period was +16.8%, and minimum return was -22.5%. The returns shown do not include dividends.
Table 1 shows the average return for each 4-week "month", the sample standard deviation, the maximum and minimum returns achieved for the month, the number of years in which the month gave a positive return, and the percentage of years in which the return was positive.
The results indicate that December is by far the best month in which to be invested in the Toronto stock market. The average return for the month has been 2.40%. December has also given positive returns to investors 84% of the time, and the largest monthly loss has been only -1.22%. The other months that give positive average returns plus the probability of achieving positive returns more than half the time are February, May, July, August and November.
Average Returns for Each Month, 1970-1995
Toronto Stock Exchange 300 Index
|Month||Ave.||Std Dev.||Max||Min||# of
|# of +ve
|% of +ve
Conversely, September is the most dangerous month for Canadian stock market investment. Over the past 25 years, the average return has been a dismal -1.68%. Furthermore, September has given positive returns to investors less than 1 year out of every 3. The other months which have yielded negative average monthly returns and the probability of negative returns more than half the time are March, April and October.
Some Analysis of the Results
It is necessary to look more closely at these results, to see if they are meaningful, or merely random fluctuations. I used the Student's t-test to determine if the results shown in Table 1 represented statistically significant deviations from the TSE average monthly return of 0.56%.
The Student's t-test indicated that the positive returns achieved for February and December are statistically significant above the 90% probability level. The negative monthly returns for September and October are also significant above the 90% probability level. I also checked the results using the 4-week holding periods ending the last week of each month. Only February, September, October, and December had returns that differed significantly from the average monthly return of +0.56% in both sets of data.
Martin Zweig (Winning on Wall Street, Warner Books, 1986) has shown that changes in the U.S. prime rate can be used to forecast U.S. stock price movements. In my book Canadian Asset Allocation Strategies, I showed that movements in the Canadian prime rate are useful predictors of future Canadian stock price movements. How do rising or falling interest rates affect the monthly returns of the TSE300 Index?
Table 2 is subdivided into two parts, one for times when interest rates were falling, and one for times when rates were rising. The most noticeable thing about Table 2 is the strong partition of positive returns; 11 of 12 months produce positive average returns when interest rates are falling, while only 5 of the 12 months produce positive average returns when rates are rising. Over the past 25+ years, the average monthly return during times of falling interest rates has been +0.85%. The average monthly return during periods when rates are rising has been -0.05%.
TSE300 Average Returns for Each Month of
Rising or Falling Interest Rates, 1970-1995
|Falling Interest Rates||Rising Interest Rates|
|# of +ve
|% of +ve
|# of +ve
|% of +ve
Closer examination indicates that monthly influence remains strong in both sets of data. The period from November to February is very favourable for stock market investment whether rates are rising or falling. On average, the returns between March and August barely exceed the returns from Treasury Bills, even when the interest rate environment is favourable. September is a dangerous month; in fact, it appears to be worse when falling interest rates make an otherwise favourable investment climate! October is extremely unpredictable, and is probably a good month in which to avoid the stock market, especially if interest rates are rising.
Several conclusions flow from this study:
Equity investments, particularly RRSP-directed investments, should be made in November, shortly after Hallowe'en. Last-minute RRSP contributions should not be placed into Canadian equities unless interest rates are falling.
If your transaction costs are low enough, it may be worthwhile to routinely maximize equities in your portfolio in November, then increase your cash position in March. Further increase cash in September, then switch all the cash back to equities in November to begin the cycle once again.
This conservative strategy, aimed at capital preservation, will maximize chances for Hallowe'en treats, and minimize those Labour Day blues.