Keeping up with the speed of change within a bank often requires quick decisions on a variety of topics. Things can slow down when bankers need “perfect” or complete information about a decision. The reality is that perfect information is almost impossible to come by, and even if it was available, it ends up hurting the process anyway. There is also the aspect that asking for more information can act as a crutch to some bankers that either don't have the time to focus on the decision, or worse, don’t want to make a decision (consciously or unconsciously) as they don’t want to be held accountable. In this article, we highlight the background and one heuristic that will dramatically help your bank make better decisions.
The Fallacy of More Information
Theoretically, it often feels like the more information you have, the better decision you will make. In fact, several studies show the opposite.
The study gathered eight seasoned, professional horse racing handicappers to see how well they can predict the outcomes of horse races. The test was to predict 40 horse races in four consecutive rounds. During the first round, each handicapper would be given any five pieces of information of their choosing. For example, they often wanted information about the last race, jockey data, track conditions, and recent workouts. The handicapper would then make a decision about which horse would win and would register their confidence in their prediction.
On average, there were about ten horses in a race, so randomly, each handicapper should get one out of ten correct, just by chance.
In round one, the handicappers were correct 17% of the time. This was a consistent finding, validating the fact that a skillful handicapper with certain bits of information can do 70% better than chance. Interestingly, their confidence was 19%, which is more or less in-line with the accuracy of their predictions.
In Round 2, the participants were given ten pieces of information, and in Round 3, they were given 20. In the final round, the handicappers were given 40 metrics of information. The outcome?
Surprisingly, handicappers never averaged better than 17%. Also, enlightening was the fact that the more information you gave the handicapper, the more confident they become in their prediction. In Round 4, for example, handicappers averaged a 31% confidence rating. Thus, the additional information made them no more accurate, took more time to gather and analyze, and only served to make them overconfident.
Data-Driven Decisions –Creating Your Future Bank
The key is to be clear about what the decision means, what information is needed to make the decision and the availability/cost of that data to make the decision. One key takeaway is to give some thought as to data that is often required and then lay a plan to gather, clean and present that data on an ongoing basis in order to help inform better decisions. Profitability data is the perfect example of a key piece of data that many banks don’t gather now, which hampers them for effective decision making. Conversely, bankers are often waiting for information about peers or borrower information that has little bearing on the outcome.
Putting This Into Action
Information past a certain point tends to increase the banker’s confidence but not the accuracy of the decision. In other words, bankers often spend resources to have a false sense of security about their decisions while not impacting the outcome. This is a suboptimal position as it decreases your risk/reward equation.
In Part II we will provide bankers with a simple and proven heuristic that will help determine how to get the balance of information and accuracy correct that will hopefully lead to better, and faster, bank decisions.
Submitted by Chris Nichols on July 29, 2019