Profit mapping by John Ehlers
Optimization has been attacked by many technicians — and rightfully so — because peaking profit is virtually the same as curve fitting to back data. Used in this fashion, optimization can produce startling track records and still be useless for future trading. Market characteristics do change, however, and technical traders need a tool to help them adjust their preferred techniques to the changing market to improve profitability. Calculating the profit at any combination of parameters and making a three-dimensional map of the result is such a tool.
Common technical trading techniques use combinations of parameters to model market activity. The parameters are constructed so their variation has some meaning to the trader. An example of a trading technique is based on the crossing of two moving averages with different periods. The trading rule for
this technique might be "Buy when the shorter moving average crosses the longer moving average from bottom to top and sell when it crosses from top to bottom." In this case, the parameters describing the market are the periods of the two moving averages.
If the markets are changing, it makes sense that the parameters of the models also change. The dual moving average technique would produce higher profits if the two parameters were continuously adjusted. Conventional optimization would make an exhaustive search of all parameter combinations to find those that return maximum profits—exactly the curve fitting to avoid.