Fixed Targets and Stops – Effect on Win Rate

May 27, 2008: 6:41 PM CST

When testing a trading strategy or idea, it is helpful to know what the Win Rate (also known as the %Win) is, and what factors might affect this portion of your results.

But before we get too deep into the type of strategy you’re examining, it’s important to note the relationship between your profit targets and your stop-loss strategy to see what the effect will be on each of your results.  For this post, let’s specifically examine the Win Rate, or the Percentage of Winning Trades.

For the strategy, I used a momentum style strategy, where the indicator known as “momentum” had crossed above zero and rose for at least one day (bar).  The program was instructed to buy at market the next day and then play for a fixed target (optimized) with a fixed stop-loss (optimized), both in incremental dollar terms from $1 to $20 (a total of 400 tests were calculated on the DIA – Dow Jones ETF from 2004 to 2008).

Let’s view our % Win Rate to see what the relationship between our fixed stop-loss values and fixed profit targets becomes.

(Click for full-size image)

The Fixed Profit Target is scaled on the vertical (right) axis (in dollar terms) while the Fixed Stop-Loss parameter is scaled on the horizontal axis (also in dollar terms).

What if I told you I could help you achieve a 95% win rate?!  It’s absolutely possible, but not net profitable (as we’ll see in a future post).  According to the test, you can set your stop-loss at $20, $19, $18, or $17, or $16 with your profit target at $1 for a 95% or better win rate.  Unfortunately, this system generates fewer relative trades and the few trades that lose eliminate 20 profitable trades.

The opposite is true with a profit target of $15 to $20 with a stop-loss of $1.

While some trend following systems can capture that type of risk/reward relationship, it is better to lengthen your stop beyond $1 and have a more realistic risk/reward ratio.

Nevertheless, according to the test, the best win rate (% Win) comes with a profit target from $1 to $6 with a corresponding stop-loss value between $13 and $20 (which is the area I have boxed in on the chart).

When looking at the suitability of a system, the number or percentage of winning trades is only one factor, and perhaps may actually be a much smaller factor than the total dollars gained per win vs the total dollars lost when a trade is stopped out.

I will continue to run tests to see if this relationship holds across different time frames and markets but in general, the results lean towards the following:

The Larger the Stop, the Higher the Win Rate
The Smaller the Profit Target, the Higher the Win Rate

There is an optimal combination (per strategy tested) that leads to net profitability, but at the surface, win rate alone is not sufficient to guarantee profits, even with a very high % Win Rate.

Check back for more findings as I delve deeper into system testing and the relationship between stops and targets.

2 Comments

2 Responses to “Fixed Targets and Stops – Effect on Win Rate”

  1. Andre Reis Says:

    This is a fantastic analysis, very interesting. I’m no mathematician, but hopefully a nice next step would be to measure total profit instead of win rate, say, assuming 100 or 1000 random trades that hit either stop or target, according to the %win rate. That would give us a measure of what are truly the most profitable combination, because although the area you highlighted on the chart is interesting, few of us could stomach entering trades with 16:1 or 20:6 risk:reward ratio.

    A very interesting read for all swing traders out there.

  2. Corey Rosenbloom Says:

    Andre,

    Thank you for the comment. I’m trying to test close to random entries to see the relationship. Interestingly enough, with a time stop, I expect the system to be net profitable (have not tested yet – is coming). I’m not sure anyone could or even should play for such a small target with such a massively large relative loss, but we must be aware of the results nonetheless. It’s opening my eyes to new insights I haven’t heard of from anyone before so it’s interesting to see where this will take us.