What is a Robust Trading System?

The difficult part to begin to trade in any market is to define the strategy or trading system that we are going to apply to make money, as there are many trading systems but:

  • Which trading systems are really robust?
  • What are the main features of robust trading system?

We will start by solving the second question. We will be trading with robust trading system as long as we have verified that any fair value we assign to the variables that define our system, end up producing gains in the trading account.

Just the opposite to define a trading system that is not robust. This situation occurs when you have been able to verify that with certain values your system has worked incredibly well, however with other values for the variables, it has been a ruin for your account.

We can do a practical exercise to evaluate the robustness of two well-known trading strategies:

  • Trading system based on moving averages: We buy when the short moving average (MA1 of the chart) surpasses the long moving average (MA2). We close this operation and we open the opposite trade (sale) when the short moving average (faster) goes below the long moving average (slower).
  • Trading system based on maximum/minimum breakouts (Donchian channels): We buy when the price exceeds a certain maximum of “x” days and close the transaction when the minimum of “y” days is reached. For example, we can buy 1 lot when we reach the maximum of 80 days and close the transaction when we reach the minimum of the last 30 days.

In order to evaluate the robustness of both strategies we will be based on a study of 20 Forex pairs in the periods of 1993 to 2003 and 2003 to 2009. They are 20 years x 252 trading days x 21 currency pairs = 105,840 days of information. All of these multiplied by 1,600 tests of the possible values of the moving averages or maximum/minimum results in 169 million possible transactions.

It is a statistically significant sample. The results in generated profit values ($) are:

Moving averages system from 1993 to 2003:

Moving averages system from 2003 to 2009: 

The first conclusion we can draw from the above results is that some of the values of moving averages that worked very well during the years 1993 to 2003 were clearly losers from 2003 to 2009.

Now we analyze the results of the highs and lows breakout system from 1993 to 2003:

 And finally, the results of the highs and lows breakout system from 2003 to 2009:

In the results of the maximum and minimum breakout system, the benefits achieved are not as high as those reached with some values ​​in the moving average system during the 20 years of the study, but almost all combinations used in the system lead to long-term benefits.

In terms of profit factor, it is very rare that the maximum and minimum breakout system exceeds values ​​greater than 1.2. However there have been combinations of moving averages that have produced values ​​of profit factor of 2.5, but later those same values ​​of the moving averages produces only losses.

If you know anything about gambling, you probably know that the casino’s advantage over the roulette players is two zeroes: 20/18 = 1.1111; where the casino wins with the black and with the two boxes with zeros. These are profit factor values ​​very similar to those we achieve with the maximum and minimum breakout system, perhaps it is the constant (robust) advantage that you need to beat the market.


Leave a Comment