There is plenty of analysis in the financial blogosphere that purports the idea that August and September are unprofitable months and investors should sell at the end of July. This is the result of sloppy analysis and selection bias.
The chart below shows the equity curve of a strategy that buys at the close of July and exits at the close of September. The test period is from 01/1950 to 12/2015. The initial capital is $100K, full investment of equity is made and commission is $0.01/share.
The drop in equity is about 42% and the maximum drawdown is about 50%. Does this suggest that investors should stay out of the market in August and September?
NO, IT DOES NOT!
The strategy win rate is about 58%, given that 38 of the 66 trades since 1950 were profitable. Then, why is the strategy so unprofitable and suggests staying out from the market in August and September?
To make a long story short, although the win rate is higher than 50%, due to bear markets and corrections during summer periods, the average loss is much larger than the average win and that makes the strategy unprofitable.
However, as it is marked on the chart, the strategy remained profitable for extended periods of time, for example, from 1960 to 1973 with a gain of 17.3%, from 1980 to 1986 with a gain of 10% and from 1993 to 1996 with a gain of 15.3%
Therefore, the real question is not whether one should stay out of the market in August and September but whether a bear market will start this month.
If we are only interested in the next trade, i.e., what will happen to our portfolio this year, then the longer-term averages are useless unless they offer at least 95% confidence, which they do not. Looking at longer-term backtests to try to decide what to do now, shows lack of understanding of probability and statistics.
Backtests say something about the longer-term assuming market conditions remain the same. But do they? Or will they?
Below is a backtest of the same strategy from 01/1900 to 12/1932.
The strategy of buying at the end of July and selling at the end of September actually made 122% during those years. Why? Because, bear markets and corrections did not occur during the summer months.
Those who are motivated by wishful thinking and exercise selection bias will only show the first chart but not the second. But
It’s the market conditions, stupid!
Charting and backtesting program: Amibroker
Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service.