How are entry and exit dates chosen for spread trades?
Several criteria go into choosing which of several potential entry and/or exit dates to publish. For this BP/EU spread, there may be multiple combinations of entry and exit dates that are perfectly valid. In other words, there may be several strategies with entry dates such as those to which you refer. When we choose which one to publish, we might normally give the most weight to win/loss percentage, then to total average profit, then to average-profit/day, etc. This particular one had the best overall numbers.
For example, the extreme low visible in the seasonal pattern would seem to be the best entry. But sometimes a low such as that has generated the most dymamic moves and the best average profit --- but a win/loss percentage of only 11 out of 12 instead of 12 out of 12. So, it becomes a judgement call --- which is in the best interests of the customer? That is also why we provide the seasonal pattern, etc., so that each trader can make up his own mind. Real-time traders are not tied to dates we publish; those dates are optimized. I tend to think of them as residing at the peak of a distribution bell curve: The date published may not have been the best date in any of those 12 years; but it was the one date over all others that generated the optimum results if used in every year.
Real-time traders can use these strategies to anticipate the reccurence of these price movements. It is, however, up to them also to decide if and how much they consider current news and fundamentals --- based on their own trading style and expertise, their own perspective of the market, the ebb and flow of news, their own resources, etc. We do not presume to know any more than anybody else about what the market will do tomorrow let alone a week or two or three from now. We provide the research; traders must decide for themselves if and how to take advantage of that knowledge.
