Valid forecasting of deposit accounts

Deposit pricing is challenging because it involves making assumptions about tomorrow. If your assumptions about the future behavior of your competitors and consumers are false, you will be making a very costly decision today.

To avoid this, you need to ensure that your deposit forecasting model is based on scientific and empirical principles that ensure reliable results.  A reliable forecast must contain two critical research elements to ensure the validity of the model and the sample size: inclusion of consumer financial behavior as a factor in the forecasting model and expanding the sample size to increase the reliability of the observations.

The most common method of forecasting future behavior of deposits is by measuring the relations between rates (independent variable) and balances (dependent variable) and using the coefficient as the basis for projection. Such a simplified model is not valid because it ignores the impact of behavioral finance (intervening variable) on the relations between rates and balances.

Behavioral finance encompasses the way consumers make financial decisions based on economic conditions. Scientific research in behavioral finances shows that consumers change their financial decisions on deposits in a predictable manner based on the state of the economy. In other words, changes in economic cycles will cause consumers to shift funds from term to liquid accounts, and vice versa, and exhibit variable sensitivity to rates.

The only way to achieve reliable results in deposit forecasting is by incorporating behavioral finance as a factor in the projection model of rates and balances. Omission of behavioral finance from the forecasting model will produce false results because rates and balances do not operate in a vacuum – they involve decisions by consumers.

An observation made from a small sample is much more likely to be misleading because what you see is not all there is. It’s a golden rule in scientific research that large samples are more stable; therefore, forecasting trends in deposit rates and balances must be done with the largest available sample.

Even though deposit pricing is specific to a market area, it is critical to observe trends in deposits on a broader scale, preferably nationally, because trends are more likely to be evident in a larger sample than in a smaller sample. Moreover, with the increasing presence of direct Internet banks, every financial institution should constantly monitor trends in deposit rates and balances on the national level because that is the only way to monitor direct banks.

Using a valid and relatable forecasting model and trend analysis of deposit rates and balances is becoming a necessity because what you don’t see can cost you. Consider forecasting and trend analysis as an early warning system of what’s ahead for deposits even if you currently price only locally. Pricing trends are like ocean waves – they start from a faraway point, but eventually will reach the shore.

Finally, keep in mind that there is no such thing as absolute certainty in the forecasting of deposits or anything else for that matter. There are only degrees of probability. So, the question you need to ask before you price for tomorrow is: how wrong do you want to be?

Mr. Geller is an expert in behavioral finance and the author of the book Money Anxiety. He can be reached at [email protected]