Sunday, January 10, 2010

Adaptive decision making - the battle of intuition versus rules

I have a bias toward systematic investing. I like the discipline with this approach. It provides structure that works well, yet there is a high degree of uncertainty. Rules provide guides for behavior. Rules control behavior and the potential for excess or mistakes.

Nevertheless, there still is a fair amount of intuition that is used with any model-based system. Data has to be gathered and processed. Some data is used other is discarded. A theoretical framework needs to be employed to determine what should be the appropriate output. Models that do not make sense have to be discarded.

You also have to know how to translate forecasts into decision rules which will size positions. Good forecasters do not always make good traders and good traders are not always good forecasters. You do not have to be right all of the time or even 50% of the time. You have to admit being wrong and scale trades according to the certainty at the moment.

Even if you have two modelers who have access to the same data, the models could be substantially different. This related to the expertise or knowledge of the developer. There is a lot of subjectivity with model building, yet there is little work discussing how intuition is used in decision making. There has to be adaptability with decision making. some would also say you need a naturalistic approach based on the environment and the level of experience faced.

One of the best researchers in the area of adaptive decision making and the the use of intuition is Gary Klein. He is not an academic but a consultant who has been on the front-lines of trying to make professionals better decision-makers. His work is powerful and important in any discussion of how real life decisions are and ought to be made. I was naturally interested when I saw that he had a new book published.

His new book, Streetlights and Shadows: Searching for the Keys to Adaptive Decision Making, may be the culmination of all of his thinking on the development of decision processes. It is not so much a theoretical piece than a response to all of the processes that have been tried to help make better decisions that may be wrong. He discusses ten claims on how to make better decision which he shows are either wrong or have some strong flaws. all of the claims make sense. All of the claims most would agree with, yet Klein suggests that many are wrong or need to be tempered.

I will go through the list of claims; however, you will have to read the book to find out the particular problems with them.

1. Teaching people procedures helps them perform tasks more skillfully. Everyone should agree with this one, yet only emphasis on process will not improve many decisions.

2. Decision biases distort our thinking. Of course, this is the foundation of behavioral finance, yet we need simple rules of thumb to help simplify our lives.

2a. Successful decision makers rely on logic and statistics instead of intuition. Logic helps but sometimes shortening the process also improves decisions.

3. To make decision, generate several options and compare them to pick the best one. Again, we have to develop our decision tree, yet good decision-makers often do not run through this process. They recognize a situation and move straight to an answer.

4. We can reduce uncertainty by gathering more information. No, more information may not always help us especially if we have to make a situation under a time constraint.

5. It's hard to jump to conclusions wait to see all of the evidence.

6. To get people to learn, given them feedback on the consequences of their action. No, we may need feedback on the consequence of our thought processes.

7. To make sense of situation we draw inferences from the data.

8. The starting point for any project is to get a clear description of the goal.

9. Our plans will succeed more often if we identify the biggest risks and then find ways to eliminate them.

10. Leaders can create common ground by assigning roles and setting ground rules in advance. Sometimes the mind meld of different ideas will be helpful.

Even for a quant. Or, more importantly for a quant, there is a need to think about the role of intuition in the decision process.

No comments: