Wall Street financial models often resort to what is known as data mining. Information for various investing strategies of the past. They are collected and tested on a “what if’ basis.
This is also called back-testing when the strategies of the past are used to see what would happen, hypothetically, when projected into the future.
All this is based on many assumptions that the mathematical models are supposed to predict.
I have made a unique mini-career of looking at well over 1,500 investment strategies used by investors over the years. I have investigated advantages and disadvantages of each.
And I can tell you this: There are some worthwhile concepts as well as gibberish in all. But no panacea exists. I would say that most of the data mining is therefore useless, except for their marketing of investment management services.
No comments:
Post a Comment