Actually all traders have heard about Bollinger bands. They are depicted in Figure 1, where the curve in the middle is the 20-day SMA, while the upper and the lower curves are located at a distance equal to 2.0 times the standard deviation (note that other values can also be used). Various techniques have been created for trading the Bollinger bands, where perhaps the simplest strategy is buying a stock when the price touches the lower band. Here, a trader may look for a profit when the price touches the upper band again. Great, but is it really a profitable strategy?
Figure 1: Stanley Black & Decker Inc (ticker: SWK) at the end of 2014. Source: tradingview.com
I developed a computer program for testing this simple strategy on stocks from S&P 500 index in years 2000-2017. I considered only stocks priced at least 5 USD but less than 500 USD. The trading daily volume had to be greater than 100 000 shares. I set a stop-loss 50 cents below the lowest of the past two candles. Also, I considered only bull markets, modelled by assuming that the 100-day SMA of the S&P 500 index is above the 400-day SMA. Finally, I decided to enter a trade only if the win/loss ratio was reasonable, i.e. the distance between the upper band and the high of “yesterday” was greater than the distance between the high of yesterday and the stop-loss. In my testing, I bought a stock on open of the next day after the price “touched” the lower band. The output is shown in the table below, see the first column in the table. The results are actually positive, but I dislike the low percent of profitable trades and the huge maximum drawdown.
|Entry:||Bollinger bands||Bollinger bands + engulfing bullish pattern|
|Number of trades||32642||668|
|Profitable trades [%]||38.55%||52.54%|
|Average trade [%]||0.97%||1.17%|
|Number of days held||19||17|
Coming back to Figure 1: Let us now extend the strategy by entering a trade only if the engulfing bullish patter occurs at the lower band (denoted as A). In other words, we combine two different techniques and expect that this will increase the probability of success. In addition, I demand that on the next day the closing price is above the high of the pattern. In other words, we wait until the engulfing pattern is confirmed. The results are shown in the table above, the second column. Note that the maximum drawdown and the percent of profitable trades are much better than before – even though the win/loss ratio is lower (because we entered at a higher price than in the first case).
Summarizing: This trading strategy is easy to use and even appropriate for an automated trading. It combines two different approaches and this fact improves the performance.