Evaluating management: Bayesian reasoning and fallacy of obviousness

E

When I invest in companies, I don’t vouch for or give a character certificate to management. I look at the past and current behavior and then try to arrive at a judgement. In majority of the cases, past behavior is a good indicator, but we do get surprises from time to time.

If new developments make me change my view, I will not try to defend my past decision which was made on a different set of facts. The key is to rationality is to evaluate new facts appropriately and move on from there. As John Maynard Keynes said a long time ago – when facts change, I change my mind. What do you do sir?

Let’s move to the point of how to evaluate management quality in light of poor behavior? For starter, there is no formulae which will give the answer. The best analogy to judge management quality comes from the court system in passing verdict on defendants. A defendant is assumed innocent till proven guilty.

I personally try to look at management with a neutral view when I start analyzing a company. They are neither good nor bad. This is a very important point. I have seen majority of investors start with a presumption of a good or bad management and then collect evidence to prove it. It is very easy to make an assumption and gather enough evidence to prove your point.

The fallacy of obviousness

See this wonderful article which makes the same point. I would highly recommend reading this article. Some excerpts –

So, given the problem of too much evidence – again, think of all the things that are evident in the gorilla clip – humans try to hone in on what might be relevant for answering particular questions. We attend to what might be meaningful and useful

However, computers and algorithms – even the most sophisticated ones – cannot address the fallacy of obviousness. Put differently, they can never know what might be relevant. Some of the early proponents of AI recognised this limitation (for example, the computer scientists John McCarthy and Patrick Hayes in their 1969 paper, which discusses ‘representation’ and the frame problem).

In short, as Albert Einstein put it in 1926: ‘Whether you can observe a thing or not depends on the theory which you use. It is the theory which decides what can be observed.’ The same applies whether we are talking about chest-thumping gorillas or efforts to probe the very nature of reality

Equal priors

The key is to start without an assumption (50-50 probability for both scenarios or equal priors) and look at the meaningful (and not trivial) evidence to come to a conclusion. Once you have done that, your conclusion should not be set in stone, but treated as a hypothesis which can change based on new evidence.

If the management continues to behave well, your confidence is increased. If you start seeing negative behavior, your confidence goes down and at some point (which cannot be mathematically defined), you may lose faith in the management and exit the position.

The above approach is fancifully also called Bayesian reasoning.

One should think probabilistically when evaluating management and not consider these issues as black or white. That’s the essence of Bayesian reasoning.

The central point of this approach is to look at new evidence in light of your prior conclusion and change it in proportion to the evidence. In some case, the new episode may be a small one and will cause you to reduce your level of confidence a bit. In other cases, either the episode or series of episodes will be so awful, that you will be forced to change your mind completely.

—————-

Stocks discussed in this post are for educational purpose only and not recommendations to buy or sell. Please contact a certified investment adviser for your investment decisions. Please read disclaimer towards the end of blog.

 

Add comment

By Rohit Chauhan

Subscription

Enter your email address if you would like to be notified when a new post is posted:

I agree to be emailed to confirm my subscription to this list

Recent Posts

Select category to filter posts

Archives