A comment I made has resulted in quite a few queries. The statement was: “I seek to exit when the market has two consecutive days where the true range is greater than mean +3 x the standard deviation of the average true range”.

Most of the posts could not understand why I chose to find a spot to exit. After all, isn’t volatility supposed to be the life blood of traders? And, why didn’t I just reduce my position size?

Reduction of position size would be automatic - that would be taken care of by the position sizing algorithm.

So position sizing is not why I look to exit. The reasons lie in the tools I use to identify high probability trades. I use stats to identify:

  • when it is a low risk buy/sell because the market is in a zone where the correction is likely to end.
  • when it is high risk buy/sell because the higher timeframe is likely to correct. The higher timeframe line turn will cause my timeframe to at least attempt to change its trend.
  • If I am entering intra-day as a responsive trader, the prices at which the high/low of the day may be said to be in place. To do this I need a reliable guide for the range of the day.
  • If I am entering as a breakout trader, what volume size tells me it is too low for the breakout to be valid.

Etc etc.

All of the above rely on stats providing a picture of what is ‘normal’. But what happens when the abnormal is the norm? In that situation, all my prior data (are next to useless - we are not comparing apples with apples). On the other hand, the ‘new’ data do not form a large enough sample size to say that this is the new reality - so I abstain.

This approach works for me. It may not be to your liking or suit your personality. But I would rather be out than in a trade and say: ‘I am more uncertain than normal because I have no yardstick by which to measure the price action’.

I am attaching one chart to show you what I mean. The blue data in Figure 1 show the ES’s stats prior to the current increase in volatility; the red shows the current figures. Notice that currently the ES mean is greater than mean + 3 standard deviations of the old stats. In this environment my preference is to stay out.

10-09-2008-atr.jpg

FIGURE 1 S&P Stats

Related Posts