How To Evaluate A Mechanical System

By ProfessionalTradingSystems.com

Every trader has his own set of parameters with which he evaluates whether a particular system is good for real trading or not. Most people use different coefficients and metrics and in this article I`d like to show my own criteria.

Mainly I want to know 4 basic things about every algorithmic trading system:

• How much can you gain per year – to obtain this measure you just need to calculate the simple  average of the performances of all years included in the backtest. So, if you have 3 years of backtestested period with 30%, 40% and 50%, then obviously the Average Annual Return is 40%. Knowing the the potential without knowing the risk is very dangerous, that`s why we need the second measure.
• How big is the downside – Actually this is the size of the Maximum Drawdown and is measured from the peak-to-trough. Here we understand what risk is involved in trading the system, assuming of course that MaxDD could be bigger or smaller in the future. Here are are talking about the MaxDD for the entire period of testing not drawdown for any particular year.

• The above mentioned measures are united in one single coefficient –  Return-to-Risk-Ratio. When you divide Average Annual Return to MaxDD you can see the quality of any mechanical system much better. Now you know how much you can gain adjusted to the risk you have to take trading the system.
• How much the strategy depends on brokers quality – When a system makes 1-2 pips per trade on average it is very dependent on broker`s quality, its spreads and trades`filling. Slippage of 0.5 pips per trade is normal for some strategies and you can wipe out half of your gain when you subtract slippage form your backtestetd results. Usually this kind if strategies are scalpers.  The more you gain in pips on average, the better. Our Portfolio for example has 7-8 pips per trade on average, so if you subcontract a very big slippage of 1 pip it will still remain a very profitable set of systems.

As every part of my trading I try to stick to very simple rules and measures. I am focusing only on the most important things which I`d like to be answered and then I find a simple way to answer to my questions via math ratios.

So with our example I want to know much can I gain per year, how much is the risk and how much I depend on choosing a broker.

Remember that when you see an equity curve you almost instantaneously know whether a system is good and bad for real trading. Nice growing line is invaluable. The measures are just tools to qualify and compare different systems with more ease.

Article by Professional Trading Systems –  Forex Mechanical and Automated Systems