Parabolic Whipsaw Oscillator (PWO)
As another thought to tie in with the last series of posts, a nice complement to the PTO and PSO would be a measure of the probability that a PSAR trade will be successful. While the PSAR looks great on a chart to the average trader, the system tester knows that the PSAR can be inconsistent due to frequent signal changes. It is logical that if you follow the PSAR with the prevailing trend, eventually you will have a successful trade. Unlike a roulette wheel, a sequence of failed trades to catch a trend are not truly independent, and will eventually lead to a profitable trade. This is due to the nature of trends and volatility: 1) long-term trends persist in the market and 2) volatility is mean-reverting. Thus it stands to reason that if we tracked the number of consecutive failed trades we may find that the odds of a winning trade are higher after a sequence of losses. The PWO is calculated as follows:
1) record the number of consecutive losing PSAR trades prior to a winning trade—note this is conceptually similar to recording the length of up and down runs.
2) take the percentile ranking of the current run of consecutive losing PSAR trades versus the last 100 losing PSAR run lengths (ie consecutive # of losing trades) this is the PWO.
When the PWO gets really high when going with the long-term trend, it is likely that a new trade will be successful. Conversely, when the PWO gets really low this implies an unusually high frequency of winning trades—this is a warning that future trades will be unsuccessful and possibly a unique indicator of an impending intermediate term market correction.