Improve your Backtesting with these Simple Steps
Backtesting is a critical process a trader has to complete before he or she can put real money into a trade. Backtesting helps to gauge the effectiveness of the strategy by examining if the trade principles are repeatable more often, to produce returns.
But, traders often get shocked to see how a successfully backtested strategy often performs poorly during live trading, and unable to produce the expected results they had earlier predicted. It s a common theme that most backtested results doesn't correlate properly to live trades. This is the situation many traders are caught up with, and unable to comprehend about the effectiveness of backtesting.
However, there are few tricks up the sleeve, that can drastically improve the outcome.
In my two books, “Trade Smart with 10/20/30 Rule™” and “Trade Forex with Confidence”, I have introduced the 10/20/30 Rule™ that I have created, to ensure we get high correlation factor between backtested results and live trading performance. Correlation is absolutely necessary to see the strategy performs similarly when traded live and eliminate surprises. Therefore, the trader is now more confident in taking the trades as it comes.
The 10/20/30 Rule™ enforces strict rules in strategy building to ensure errors are not introduced to the trading system and also hidden flaws are uncovered and get fixed.
Here are some of the points you can do to improve your backtesting:
Too Few Data points Don’t take short cuts to backtest your chart patterns or indicators with just a few trade examples. You need a statistically good number of data points to make the assessment. If you are backtesting manually on a new strategy, then you would need between 30 – 50 data points.
Too many Data points Some traders do the gung-ho approach of running backtest for several years of data and make the strategy to work during bull market, bear market and consolidation period. To achieve this, they have to resort to adding too many rules or parameters for the strategy to work. Thus, it creates an over-fitted curve that can’t adapt to new market conditions and eventually fail.
Rules not clearly defined. You need to have a defined entry and exit rules and stay with it on your backtesting. Do not be make rough guesses and take the trades too early or too late. Hindsight errors can be introduced in backesting when traders record the entries and exits of their strategies at favorable levels and produce nice over-biased results.
WYSIWYG The caveat. Backtesting would only show whether the strategy is profitable or not profitable by running through the historical data. It does not indicate you on how to optimize the strategy further nor does it alert you about hidden flaws in the system.
Backtesting requires few little tricks for it to work reasonably well. A good backtested strategy is then primed up for next levels of testing like strategy optimization and in-sample/out-sample testing to ensure the strategy could produce the results you are after in live trading.
About the author
Ramesh is the author of 2 trading books “Trade Smart with 10/20/30 Rule™” and “Trade Forex with Confidence” which are available on Amazon. He created 10/20/30 Rule™ for traders to build and test their next money making trading strategy fast.