Flying Blind or Deposit Price Optimization

Arguably, at no time in history has the knowledge of depositor rate sensitivity been of more value than today given the importance financial institutions currently place on the deposit gathering function.  Deposit Price Optimization, the key to making informed and accurate deposit pricing decisions, relies on the expected future behaviors of depositors’ supply sensitivities (behavioral patterns) to interest rate changes and local market conditions, as well as the segmentation of each specific depositor by rate sensitivity.

Current mortgage market and liquidity turmoil, non-bank deposit competition, investment alternatives, and retirement planning needs, all make deposit funding and the price paid for deposits instrumental to reversing shrinking net interest margins.  In short, the need for financial institutions to possess the information that allows them to accurately predict the “optimum” price for deposits to reduce net interest margin pressures and reach deposit balance growth targets, while retaining and building customer loyalty, has never been greater.

When most financial institutions make deposit pricing decisions, they’re actually flying blind.  Working with limited information, the financial institution’s Deposit Pricing Committee can only “guesstimate” how deposits should be priced.  Financial institutions attempt to adequately price new and existing deposits by pricing deposits according to competitors’ posted rates.  What the industry refers to as “competitive pricing.”  Another strategy employed often is raising interest rates on deposits until a desired balance growth is achieved, called “walking rates.” As you can see with these strategies, most deposit pricing committees do not possess the most important data required to make optimal deposit rate-setting decisions, the knowledge of an institution’s own depositor rate sensitivity function, or in economic terms, depositor “price elasticity.”

Through years of experience with their depositor base, many institutions have a strong “implicit” sense of depositor price elasticity.  But it’s safe to say that no institution can know its depositor price elasticity with a high degree of precision within this framework.  The reason for this is not that the data doesn’t exist - it does.  The reason for this lack of knowledge is that there is no model yet available to process the volumes of information required to run and turn statistical analyses into understandable output.  In addition, there is no trained knowledge base that can interpret and implement the output provided by such a model (that is, turn the information into efficient deposit pricing decisions or deposit price optimization).  What is most needed then is a model designed to optimize deposit pricing decisions and make guesswork a thing of the past.    

Relying on Assumptions.  Our industry has developed strategies and recommendations to grow deposits, set deposit rates, and establish high-profile deposit promotions.  But all the strategies, suggestions, and formulas lack one fundamental ingredient - statistically based customer behavioral patterns.  The bottom line is that the approaches make assumptions about customer behavioral patterns that are not based on statistical or mathematical analyses.  The reason these approaches rely on assumptions is that no analytical or mathematical approach exists to take the assumption into a more quantified realm.

There are obvious reasons why institutions might not use deposit price optimization: the high cost of software development, statistical requirements for large amounts of data, and the lack of deposit pricing sophistication [an internal knowledge base] that the software requires.  The solution is an affordably priced software program that can process large volumes of data using highly rigorous and academically accepted mathematical algorithms to determine depositor price elasticity.

Just think what an institution could do with information derived from a sophisticated proprietary model designed to statistically analyze the factors that drive depositors’ deposit decisions and optimize deposit rates to increase profitability and net interest margins.  Armed with this knowledge, financial institutions would, for the first time, be able to project the effect of different interest rates on deposit balances. The model would assign an accurate economic value to deposits, provide the ability to test your own pricing scenarios to predict their effect on balances and profits, determine the best time to run deposit promotions, and assist in the planning of the need for “wholesale funds” in advance.

The reason a model such as this is needed in today’s marketplace is quite simple.  Current practice, whether it is the marginal cost of funds approach or the estimation of depositor retention rates, all rely on an institution knowing its depositor behavioral patterns, or depositor price elasticity.  Specifically, the marginal cost of funds approach is predicated on the need to minimize “cannibalization” and the need to raise additional deposits to fund growth objectives at the lowest all-in cost available in the market at that time.  This is not to say that the marginal cost of funds approach should be abandoned: but, the marginal cost of funds approach is based upon an assumption that the institution possesses a high degree of certainty of its own depositor price elasticity.  Therefore, without more precise knowledge of depositor price elasticity, the effectiveness of the marginal cost of funding approach is limited because it is inherently dependent on an imprecise assumption about depositor price elasticity.  The institution is flying blind because the assumption about depositor price elasticity is just that, an assumption.

Duration Another Key.  Another important calculation requiring an assumption of depositor price elasticity is the duration of deposits, a calculation highly dependent upon assumptions assessing depositor price elasticity.  Obviously, this dependency is difficult to quantify without a statistical analysis of depositor behavioral patterns. 

The regulatory insistence that duration estimates be backed by a reasonable set of hypotheses regarding bank pricing and depositor behavioral patterns increases the importance of using statistical analysis.  Without statistical analysis, it is nearly impossible to accurately determine depositor behavioral patterns.  In fact, most practitioners of the art of asset and liability management acknowledge the importance of, the difficulty in, and the lack of ability and knowledge base to calculate, the estimation of balance sensitivity parameters from bank data.  In other words, depositor price elasticity is one of the key ingredients in duration calculations, marginal pricing decisions, and deposit growth attainment.  Yet this one factor is assumed without an accurate measurement tool in the entire process.

Again, “flying blind.”

In summary, to calculate or even begin to understand price elasticity requires a model embedded with statistical algorithms and optimization routines to determine customer sensitivity to deposit rates and to determine which deposit rate is optimum to bottom line profitability.  In the future, deposit pricing committees must move away from “guesstimations.”  Deposit pricing committees should have at their routine disposal price elasticities for each depositor segment, thereby enhancing depositor product design and increasing the depositor value proposition. 

The overall goal of this article is to leave the reader with an appetite to hear more about deposit price optimization.  A discussion explaining in more detail how a deposit price optimization model works and how it integrates into the corporate decision making culture is now warranted.

Fly Blind or Deposit Price Optimization?

This article was first published in the December, 2007 issue of the Bank Asset/Liability Management newsletter. Subscription information for this newsletter is available at (800) 572-2797. This reprint is made with the permission of A.S. Pratt & Sons.