Strategy Test: Bollinger Band

This week's technical indicator test is for Bollinger Bands; a price overlay indicator. A basic long-side strategy - with and without stops - was tested where crosses of the lower BB band were bought (when the previous day's close was below the lower BB band) and crosses of the upper BB band were sold (where the previous day's close was above the BB band).


Test period: A complete bull-bear cycle defined by the S&P (March 20th 2000 to October 8th 2007).

Stocks: Active Trader list (AAPL BA C CAT CSCO DIS GM HPQ IBM INTC IP JPM KO MSFT SBUX T WMT)

Number of shares: 100

Commission: $9.95 (included in the loss calculation)

Trades: Round-trip only; partial trades were excluded.

The strategy was far more economical in the number of trade signals generated when compared to last week's on-balance-volume strategy test. The non use of stops generated 218 trades (compared to the frisky 1,300 of stochastics):

Total Profit: -$1,249
Winners: 139
Losers: 79
Win percentage: 64%
Profit Factor: 0.97

Adding a stop to the equation really messed up the returns, in part attributed to the jump in the number of executed trades: when a stop was used an average of 433 trades were executed, almost double the non-stop strategy. The following returns assumed no commission:

No stop: TP = +$3,089 on 66% winners
3% stop: TP = -$3,205 on 37% winners
4% stop: TP = -$5,983 on 36% winners
5% stop: TP = -$8,894 on 37% winners
6% stop: TP = -$10,216 on 37% winners
7% stop: TP = -$12,020 on 40% winners
8% stop: TP = -$10,909 on 41% winners

What of a three test period on no commission?

The three randomized dates were 2003/04 (sideways market), 2004/05 (bullish) and 2006/07 (bullish). There was a clear performance benefit from the non-stop strategy. There was no trend for returns based on a sliding stop scale, although a looser stop returned a higher win percentage - but not necessarily a higher profit:


Use of Bollinger bands has the potential to provide good trade signals when married with an indicator which performs well using stops. The high win percentage suggests Bollinger band buy signals occur close to tradable lows; the low profit return suggests it may not be the best exit tool and therefore may perform better with a trending indicator which catches more of the trend, such as the MACD.

This is something for a later Strategy lab.

Dr. Declan Fallon, Senior Market Technician, Zignals.com the free stock alerts, market alerts, and stock charts website