Modeling Uncertain Variables

The Use of Modeling Inherently Uncertain Variables

I agree with David Babson’s idea of the unpredictable of investing in that some variables are much more predictable than others. The focus of investing should be bottoms up at a company specific level that should especially take into account predictable trends. However, I disagree that the so called unpredictable variables are thus of little importance and are not worth focusing on. What makes a market is that some parts of the valuation equation are either unknown or difficult to predict. It does not make them either useless or unimportant, especially since these unknowns are the unexpected variables that will move the stocks. Therefore, it is possible the unpredictable variables become the most important factors because they are unknown.

The main reason unpredictable variables are important is it is always crucial to know what expectations are regardless of whether you can handicap outcomes. It may be difficult to know what variables such as GDP or interest rates will be, yet to know what the market is pricing in can often give an indication of whether there is more upside or downside already priced in. If these variables are impossible to predict it would be advantageous to take a contrarian approach to what the market is expecting if the expectation is outside of the variable mean. Thus, it is important to accept that some variables are unpredictable but at the same time useful to try to predict whether what people expect is too extreme given that no one really knows.

Further, by knowing what potential outcomes are possible from unpredictable variables, investors can be surer of what will happen when different events either happen or become expected. Regardless of how predictable any variables are, there is still a lot of uncertainty in investing. This is why investing becomes in essence, a numbers game of risk and return. Thus, by developing different scenarios and the outcomes of them, one can attempt to create a forecast for whether the return outweighs risk and in such a way whether something is or remains a good investment.

It doesn’t really matter whether or not effectively we can solve for a certain variable to value a stock. Even if we cannot this missing link in the valuation is important. Those who wish to beat the market do not want people to be able to solve for it because then the market would be efficient and the need for asset managers or ability to be an active investor would be gone. David Babson is arguing for stocks over the long run, and in such a multi-decade outlook, things like GDP or politics become effectively just noise in the scope of very long term company specific trends. Yet, such investing is not as normal today with people demanding short term returns, and does not work as well in a more efficient market than existed when people like Babson and Buffet were highly successful. In today’s market, there is no such thing as too much information, especially the information that is contested and uncertain.