No Exceptions

A good way to assure long-term success is to follow your system without exceptions. If you want exceptions, write them down and test them thor­oughly instead of relying on your intuition alone. If you want to over­ride your model, it probably means you are not comfortable with it and should do more testing and refining.

Many traders will use rising volatility as a measure of impending market change. They will reduce their positions when volatility is "high." You should define this scenario precisely. For example, you could say that volatility is "high" if the 50-day standard deviation of closing prices is more than $1,500. Then, you could test historical price data to check what the markets have done when volatility has risen to this level.

You could choose to increase market exposure to 4 percent of ac­count equity if conditions are "right." Instead of relying on intuition, you could define a specific condition as "right," such as a 2 5-bar break­out accompanied by a 5-day RSI value greater than 70 or less than 30. You can then check historical data and develop some statistics on what to expect.

We saw in the tests with the 65sma-3cc model in chapter 4 that only 4 percent of the trade were the home runs that made all the differ­ence in portfolio performance. If you routinely deviate from your sys­tem, there is no assurance that you will not manage to miss all the im­portant trades in a given period.

The no-exceptions policy is necessary for another reason: traceabil-ity, which is discussed in the next section.