By David Pieper, Senior Technical Analyst, Tradesignal Ltd.
The Turn-of-the-Month Effect (TOM) is one of the most interesting and at the same time oldest price anomalies in the stock market. Norman G. Fosback introduced this calendar effect in his book “Stock Market Logic” in the seventies. The TOM effect is all about the simple idea of buying shares at month’s end and selling them again on the third day of the new month.
Is it possible to earn all of the equity return on just 4 days per month?
Several studies have proven the TOM effect to be true. In this article we concentrate on the work of Xu and McConnell*, which is based on long-term data from 1926 to 2005. Let us take a look on their findings, which confirm that a statistically significant TOM effect in equity returns appears to have persisted for over 100 years.
Figure 1 shows the mean daily returns for the US stock market. Please notice that -1 is the last trading day of the month, +1 is the first trading day of the month and so on. The chart proves impressively that the majority of US stock market returns between 1926 and 2005 were indeed achieved around the turn of a month:
- The value-weighted mean daily return for TOM days was 0.16%. On other days it was 0.01%.
- The equal-weighted mean daily return for TOM days was 0.23%. On other days it was 0.05%.
Figure 1: Average Daily Returns of the US Stock Market (1926-2005). The US stock market generates unusually high returns around the turn of the month relative to other days. -1 is the last trading day of the month, +1 is the first trading day of the month, +2 is the second trading day of the month and so on.Source: Xu/McConnell “Equity returns at the turn of the month”
- With value weighted returns, over the entire 80-year period of 1926-2005, market participants essentially earned no premium for bearing market risk except during the four-day turn-of-the-month periods**.
- Thus, on average, over the other 16 trading days of the month investors received no reward for bearing market risk.
** With equal weighted returns, investors did earn a small positive reward for bearing risk – due to the higher returns of small-cap stocks.
Maybe the TOM anomaly is due to higher risk or some other factors?
Xu and McConnel explored several potential explanations for the TOM effect. Here`s the summary of their results which all confirm the significance of the Turn of the Month anomaly:
- The TOM effect is not concentrated among small-cap or low-price stocks.
- The TOM effect is not confined to calendar year-ends or calendar quarter-ends.
- The TOM effect cannot be explained by higher risk at the turn-of-the-month: Standard deviation of return is not higher during the four turn-of-the-month days than over the other 16 trading days of the month.
The TOM effect also is not confined to the US stock market. In 30 out of 34 non-US countries, the TOM effect has occurred. Figure 2 shows that the TOM effect can be observed internationally.
Figure 2: The TOM effect occurs in several countries. The chart shows the mean daily returns around the TOM period compared to the performance of the other days.
Source: Xu/McConnell „Equity returns at the turn of the month“
Back testing made easy
Backtesting the TOM effect for a portfolio of stocks or equity indices including position sizing and simulation of transaction costs is made easy using the right software. With Tradesignal you can even optimize an existing strategy – with no need to program.
Figure 3: Tradesignal workspace for back test. With Tradesignal you can back test and evaluate the TOM effect for any market – without programming.
The following video shows you how easy it is to backtest TOM effect. A long entry is always made at the closing price on the last trading day of a month. An exit is made at the third day of the new month. This will automatically limit to twelve the number of trades per share. Of course you can test other holding periods and evaluate multiple risk-return metrics.
Author: David Pieper, Senior Technical Analyst, Tradesignal Ltd.
Tradesignal® and the Tradesignal logo are registered trademarks of Tradesignal GmbH.