I read somewhere about a strategy in which a trader should search for stocks with a recent significant and constant increase in price, for instance 20% during the past two weeks. An example is shown in the figure below.
(CVC Limited in 2003, source: NinjaTrader)
In other words, we assume that the price has enough momentum to continue. It is suggested to buy such a stock if it trades e.g. 50 cents above the high of the previous day, i.e. the trend is confirmed.
I decided to backtest this strategy by creating a computer program that searches for such trends. For simplicity, I decided to buy the stock when the closing price is 50 cents above the high of the previous day (I enter on open of the next day). Then I set a stop-loss 2 USD below and the take-profit level 2 USD above.
For improving the strategy, we could try to trail the stop if it turns out that the price continues or/and change the stop-loss and take-profit levels. Nevertheless, this is something I do not want to do: I just want to make sure that the price has really a tendency of continuing its motion without any tuning or optimizing.
Unfortunately, backtesting of this strategy on S&P 500 stocks in years 2000-2017 gives negative results. My program detected 1405 set-ups but the percentage of the winning trades was about 48% (with the win/loss ratio of just 1.09). This shows that the performance is random.