How long to Hold on to a Trade?

Traders often face the problem of exiting a trade too early or too late. Holding time is something that can be analyzed more to improve your trading results. Every underlying has its own statistical properties.

Standard deviation

The table shows the standard deviation for close-to-close returns over the last 3 years.

What does it have to do with holding time? Well, if you held a trade for one day the standard deviation tells you how far away from your entry price it will likely go in 95% of observations.

This number shows you what one can expect from an open position in a certain time interval. If your goal is like me to profit from intraday movements, I will exit within 24 hours. If you’re trading the FDAX with a standard deviation of about 1% and an average High-to-Low-Range of 1.2%. To increase the likelihood of reaching your targets you have to be able to profit from 0.6 or lesser percentage moves because it is very unlikely that you will catch the bottom or high of any given day.

Risk management has to be built around what is achievable. A system with a high winning percentage can take 0.05% goals for example and a system with less winning percentage 0.6% as an average goal in this example. The reason is that you don’t want a system with 5% winning rate or 0.1 risk reward.


The higher the kurtosis the more often the outliers occur. What this means is that underlyings with a high kurtosis can be expected to have bigger trending moves outside of the normal boundary and are therefore better suited for trendfollowing.

Trading inside boundaries

With a good trading plan your exits should already clear. There is a certain “Reason2Sell”. Be it moving average trailing stops, market profile levels or a fixed target.

If you’re using the standard deviations to be in line with reality and what you can expect it’s not possible to exit a trade too early or too late over a statistically significant amount of trades.

What you can improve on is choosing the right stops aligned with the expectancy of the underlying. If you are a trendfollower and trade different underlyings you could have a look on the statistical properties first and then decide on what is likely to be reached, take kurtosis into account and formulate a plan around it to find good risk/reward that have a high chance of working out.

Choose a higher kurtosis to get bigger outlier opportunities.

The wrong underlying

If you’re having trouble with holding trades too long or too short, it could very well be that you’re stuck in the wrong underlying. Let’s say your problem is to exit too early and you’re trading Gold. It might be that the move that you want to have, say 1%, just doesn’t happen in the time you’re sitting on the screen. So, you would need a more volatile underlying like Sugar, Oil or Stocks.

Turning a losing system to a winning one

For example, looking at the last 100 trades traded inside a system, the following statistics are captured:

  • Average trade holding time: 5 days
  • Underlying traded: ES (SP500)
  • Average win: 1%
  • Average loss: 0.5%
  • Winning percentage: 30%

What you need now is the standard deviation, kurtosis and average high2low for the weekly time frame. Just go to, search for S&P 500 Futures and click on historical data.

There you go. For the example you will get about 2% weekly standard deviation and 2.65% average high-to-low.

Now, what you can see is that this system exits too early, because it doesn’t even capture the standard deviation.

With a kurtosis of 1.6 outliers are very rare. The plan should change now. Instead of trying to capture 1% moves, you will try 2-3% moves instead to align more with reality and turn this system into profitability by just changing the exit strategy towards holding the trades longer.

ES, Kurtosis, Standard Deviation

Leave a Reply

Your email address will not be published. Required fields are marked *

© Niclas Hummel Trading 2020 | Terms and Conditions | Privacy Policy