To All:
Months ago I came to the conclusion that, in order to get consistently good nets, the practice of using stocks as relateds was not the best. I came to this conclusion based on what I saw happen to nets that were subjected to the inconsistent price movements of the related stocks because of news that had nothing to do with target stock. An example of this would be that the stock price of company XYZ, that had always traded in tandem with the target stock, was suddenly demolished because of news that their CEO resigned, news that had absolutely nothing to do with the target stock.
To get away from this type of scenario, I started to exclusively use indices for relateds, which included those for industry group and general market. It occurred to me last night that industry groups fall in and out of favor from time to time, and the stocks in that group no longer trade with the trend of general market. It seems to me if your target stock's industry group was in favor last year and now is currently being ignored, that this type data in a net that spans two or three years would produce inconsistencies within the net similar to those mentioned above. Would it not be better to compare the price movement of the target stock exclusively to its industry group and related industry indices, and exclude the general market indices? Wouldn't this give the net a better foundation to make accurate predictions? After all, if your target stock's industry group was trending with the general market six or eight months ago and now is trending against it, and you are using general market indices as relateds, wouldn't the net see that as contradictory data?
Comments please.
Jay |
