Profit taking

A strategy commonly employed by amateur traders is the so called "profit-taking". The trader takes a position and after a certain percentage gain since the start of the trade is achieved the position is closed. And then what? Well, usually the trader will enter the position again, at a lower price - if he doesn't miss the chance..

I have investigated a simple profit taking strategy with the Dow Jones Industrial (January 1930 until October 2008). We hold the position until we achieve a profit of y% and close the position and buy again x days later. After that, the whole game starts again.

As the charts below show, the less we interfere the better the result.

The ceiling depects a buy-and-hold strategy (around 7% p.a. return) and lower mesh grid shows the annual returns generated with the respective profit-taking-strategy:

The profit-taking strategy is not working and leads in all cases to underperformance, assuming we use no leverage.

The strategy looks very appealing when presented with the right stock at the right time, but when we apply it mechanically for a long time, it does not appear to be working. On the contrary, it leads to a statistically significant underperformance to a simple buy-and hold when applying the Wilcoxon ranksum test.

The reason for the underperformance must lie in the fact that the Dow Jones generated a significant return over the last 78 years. When we interrupt those returns in an apparently random pattern, although it may not seem random to us but which in fact is random (such as when applying the profit taking strategy) it clearly leads to a significant underperformance.

As this mesh diagram below shows, the more we are long and the less we interfere, the better our return we achieve. The best return (as shown above) was achieved with a strategy where we were long most of the days.

We can conclude that even when we ignore transaction costs, the strategy does more harm than good.