Backtesting data: How I use it (ex: NXG, COCO, ACAD)

Some of you keep asking me whether I backtest my strategy. So, answering your question, yes, I do Backtesting, BUT I don't use it in my decision making. I mean, I do backtesting AFTER deciding to pick one symbol or another.

First, as we all know from funds disclaimers, past performance is not a predictor of future. In fact, it is very true, unless you have a good statistical sample. I backtest 2 years back and usually each symbol delivers 2-5 trades for that period which can not be considered statistically significant.

Second, any backtest is based on assumptions. For my strategy I backtest 5 min chart. I think that's the level of sensitivity I can (as a human) respond.

Finally, real trading is VERY different from backtesting.

You may ask then, why do I do Backtesting. Well, it's just another tool that helps me to understand what to expect IN GENERAL from a particular stock as sometimes patterns repeat at least in terms of magnitude.

For example, today, we traded NXG, COCO and ACAD

Here are past backtest results, and I knew them before entering trades:

Symbol: ACAD

Strategy: Pennystocks100 (5 min chart)

Period: Jan 01, 2009 - Dec 30, 2010

# Win: 1

# Loss: 1

% Winners: 50%

Profit-Loss%: 0.56%

Symbol: NXG

Strategy: Pennystocks100 (5 min chart)

Period: Jan 01, 2009 - Dec 30, 2010

# Win: 2

# Loss: 4

% Winners: 33.33%

Profit-Loss%: -3.36%

Symbol: COCO

Strategy: Pennystocks100 (5 min chart)

Period: Jan 01, 2009 - Dec 30, 2010

# Win: 1

# Loss: 0

% Winners: 100%

Profit-Loss%: 1.32%

Do you see any correlation with our results today? Not much, BUT today's results are within the reasonable range and I knew roughly what to expect. Psychologically, (by the way) it's a very important thing.

From the next year, I plan to publish Backtest data for every symbol, for every trade I am going to enter, in the same post with the Daily Alert after market close.