Friday, December 26, 2014

"What do I wish I knew?" and systematic modeling



You have to make an investment decision and the clock is ticking.  The decision has to be made now or at least you believe a decision is necessary. You may have a lot of information in front of you, but there will always be uncertainty concerning the result of any decision you may make. There is always uncertainty with the information available to make a decision. You just don't know whether you had all   or the correct set of information.  A key question to ask with any decision is, "What do I wish I knew?" and "If I know more information, would I better for it?" More is not necessarily better, but knowing what is needed is critical.

Investors want to have the best odds in their favor. You often wish you had more information, but you are not sure what. However, if you were given more information, you are not sure that it would be useful. The extra effort of finding any more information and delaying a decision also may not worth it. There costs with searching for more information.. Other are also playing the same game, so you also have to assess what information others have and whether you are at an information disadvantage. 

Too often we are dealing with problems of incomplete information.  There are time constraints with our decision-making. Delay is costly. There are information costs. New information may not be cheap. There is ambiguous information. Some information could be wrong, have some error, or subject to revision. The lack of information completeness could simply be that it is not being handy.

The key role of systematic trading is making all of the information you dream about handy and weighed in an appropriate manner. Systematic trading allows for preparation time for any decision.  The manager can gather all information he feels is necessary so there is not regret on what could be used for a decision. Its value of preparation is that decisions can be set-up for any set of scenarios. Any past information can also be tested to determine its impact on some make behavior.  The cost of information analysis and gathering can be minimized because it is reviewed and structured in a formal process.

Is the information reinforcing or unique to the problem at hand? Does it confirm or is highly correlated with information already available? These are all important questions on the whether new information is relevant, but it cannot be answer without some testing. A Rules-based approach allows for decisions that are based on past facts. Any information not used is based on preference, theory, and testing.

Of course, that does not mean that a systematic approach will have all of the answers. There can still be  surprise from a new event that has never been anticipated or does not have history.  Model risk and error still exist but it can be measured and quantified.

The choice is simple. Do not test and prepare decisions scenarios, and suffer information regret. Do your testing, form your rules, and play the odds of success without information regret.


No comments: