I'm having an issue where, when comparing the exploration results of the highest ranked stocks in an exploration (using the exact same system code) on an/any specific date and then compare the stocks that were actually chosen by the system on the same exact specific dates, varies significantly.
The code is designed to rank according to two distinct rankings using a slightly modified Static Variable method outlined here in the last example by creating multiple static variables and combining them -- https://www.amibroker.com/guide/h_ranking.html
- The system rotates a list of 20 stocks monthly
- When I run the same code in exploration on any of the dates shown by the system, the purchase decisions for the list of stocks are different even though the buy criteria uses the same combined ranking i.e. PositionScore = combined ranking
In other words, if I run the system and look at any specific date and then run the exploration using exact same code, the 20 stocks that rank highest in the exploration do not match those chosen by positionscore. In fact, the stocks chosen from the PositionScore decision making process are way down the ranking list versus the top rankings shown in the exploration.
I hope it won't be difficult to picture this scenario without posting the entire code, otherwise perhaps I can post the code and take out some things. I usually don't mind sharing most code but I'm testing this system on information given from someone else and that person would prefer I do not make it public knowledge.
Thank you. Regards,
DaveDM