Results of deposit pricing game theory exercise
In a recent article for BAI Banking Strategies, I broached a theory that deposit rates start rising prior to an anticipated increase in the Fed funds rate because of our tendency to try to outdo the competition.
To test the validity of this assumption, I asked readers to participate in an exercise in behavioral finance, requesting that respondents pick a whole number between one and 100. The winner would be the person whose number was closest to the average of all numbers submitted without going over the group’s average figure. Had this exercise been conducted with machines capable of only linear thinking, the average of all respondents would have been 50 because this is the normal average of the range between one and 100. However, since people are capable of non-linear thinking, they interject an element of anticipatory reaction to the process, thus producing different results.
Before I go into the actual results, I would like to explain the inverse relations between this game theory and rising rates. Since we are conditioned to think of price competition as lower prices for consumer goods, i.e. paying $18 for a book on Amazon compared to $25 elsewhere, our tendency is to think of price competition in terms of lower prices. Therefore, this game theory was designed to exhibit our natural tendency toward price competition, which is a mirror image of yield pricing. Thus, our tendency to lower prices for consumer goods is the same as our tendency to increase yields when competing for consumer deposits. And now, here are the findings from the over 60 responses I received:
The actual average of the group was 37.4, which means that the closest whole number, without going over, is 37, representing a 13% deviation from the normal average of 50. Two respondents picked 37 as their average for the group, which means that they correctly anticipated the reaction of the group as a whole to a rising rate scenario. The range of responses was across the entire spectrum, with one being the lowest and 100 the highest response submitted. Such large deviation from the normal average shows how challenging it is to anticipate competitive reaction to rising rates. In other words, when you don’t know how a competitor is going to price, you are highly likely to over- or under-react.
So, what does all this mean to bankers in a time when the anticipation for rising rates is high? The first implication is that when rates start rising, you should expect the average rates of your competitors to be slightly higher than the normal average rate of the group the week prior. The reason for that is that in order to avoid a price disadvantage, institutions attempt to outdo the competition by planning a greater increase than what they anticipate competitors will price. This dynamic will start slowly and gradually right before the Fed is “projected” to raise the funds rate, and will accelerate after the actual increase in rate.
The other implication, which is just as important, is the recognition that behavioral finance is a major factor in financial decisions. Regardless of the model you are using to forecast rate increases down the road, it is important that you incorporate financial behavior theory as a predictor in the projection. Just as this exercise demonstrated how behavioral finance shifted the group’s average from the normal of 50 to the actual at 37, your rate projections are going to be closer to reality when you add financial behavior as a predictor to your model.