There are many metrics one can use in a search for trading systems. However, just a handful of the metrics are useful and the rest are either derivatives of them or lead to excessive curve-fitting to specific series. Some new metrics that are proposed constantly even collapse to a very basic and known parameter, like the win rate or the profit factor for example, after all the algebra is done. Here is an example…

A customer using Price Action Lab contact me the other day and asked me if I could incorporate a “new” and “powerful” metric in the program search mode. The metric he described as:

return for every dollar invested = (win rate x avg winning trade) / (loss rate x avg losing trade)

Let us now find out, after reading my post “What Every Trader Should Know About the Win Rate, Profit Factor and Payoff Ratio”  how this “new” metric transforms:

Let: w = win rate, and (1-w) = loss rate. The right hand side of the “new metric” becomes:

NM = (w x avg winning trade) / ((1-w) x avg losing trade)          (1)

But avg winning trade/avg losing trade = r , the well known ratio, also know as “payoff ratio”. Thus, we now have for (1)

NM = w x r /(1-w)   (2)

But from my post we know that

pf = w x r /(1-w)   (3)

or that

pf/r = w/(1-w)   (4)

After combining (2) and (4) we get:

NM = pf

or that the “new” metric is just the profit factor, the good, old profir factor.

So much for this new metric.