Comment: Building Share Without Sacrificing Return

Banking's major retail problem can be summarized in the following statement of Jonathan Palmer, a former top executive at Barnett Banks. Says Mr. Palmer: "We found, 30 years ago, that the average number of products purchased by each customer was two ... today, the average number ... would probably range between two and three. So 30 years later, we haven't quite figured it out."

In other words, although banks serve nearly everyone, they continue to serve most in a rather superficial manner. The situation, however, is untenable. If the industry can't improve its share of customer wallet, it will not sustain today's lofty level of stock values.

Let us remember that stock prices depend not only on the return on equity (relative, that is, to the shareholder's required return), but also on the sustainable earnings growth rate. In turn, the earnings growth rate, which can be buttressed in the short run by expense cuts or expense restructuring via, say, outsourcing, depends in the long run on the growth in revenues.

Unfortunately, revenues in most retail businesses-lending and deposit- are projected to increase at rates much below the earnings growth needed to sustain current stock values. So unless banks redivide market share, taking a bigger slice of an inadequately growing pie, stock values will decline.

Yet the industry does not seem to be oriented toward improving share, since in many businesses it appears content to lose volume in order to preserve margins. This is strikingly apparent in the deposit business where rates paid have remained much below alternative prevailing market yields. In 1996, for example, the average bank rate on money market deposit accounts, one of the most profitable deposit types, came to 2.91%, while the money market fund rate averaged about 5%.

Banks have been somewhat insulated from the consequences of noncompetitive deposit pricing. In part this is because of the perceived value of deposit insurance. It is also because about half of a typical bank's most profitable customers (top 20%) are 55 years or older and thus are less prone than younger deposit cohorts to change their financial behavior. Nonetheless, even the most stalwart long-standing bank deposit customers have of late been exhibiting much more rate elasticity than was expected by defecting to the money funds.

Bank willingness to sacrifice share for margin has a certain narrow logic, if only in the short run. Consider what might have happened to the industry's 1996 pretax net (some $81 billion) and its return on equity (14.5%) if bank pricing had been more competitive.

For example, suppose that in 1996 the average bank MMDA rate equaled the money fund rate. (In the real world, of course, it would never rise to this level, nor should it.) Suppose further that the bank credit card rate was cut by some 200 basis points, and that the industry, eager to become more competitive in the small-business market, introduced enough small-business cash management accounts to halve all-in deposit margins. Add the assumption that bank loss provisions rose from their low actual levels (57 basis points) to their ten-year average levels (112 basis points).

What results? Industry earnings tumble from $81 billion to $45 billion. Correspondingly, ROE drops from 14.5% to 8.1%.

To be sure, the above numbers presuppose that "other things remain equal," which of course they never do. The dynamic change variable that can dispel bank gloom is the emerging capacity to utilize management-science tools in order to identify and target a higher percentage of better customers and serve many more of their needs at a lower cost than the competition.

There is abundant evidence that using refurbished data bases for superior customer targeting, while not the panacea that some maintain, can achieve outstanding results. Let us recall that institutions such as MBNA Corp. were consistently able to identify and attract customer segments that in most years had a lower propensity to default than the average credit card customer.

That ability both lowered loss provisions and cut operating expenses (for one thing, with a lower secular loss experience, MBNA needed many fewer collectors). Hence, it was possible to reduce prices in order to build volume while still maintaining and even adding to the bottom line.

We have helped card issuers to achieve results comparable to those of MBNA via the same sort of data base analysis. One institution, for example, used this approach to increase dramatically its creditworthy accounts. Another employed it to raise its yield on collections by up to 50%.

A large bank is beginning to use state-of-the-art marketing methods to single out and attract those mortgage applicants who are likely to create more value than the average mortgagor because these customers are less prone to prepay should rates decline. Once again, this capacity is leading to pricing actions that swell volume and facilitate cross-sell while raising profitability or, more precisely, expected value per customer.

Of course, the demands of good data-base marketing impose onerous skill requirements on institutions, requiring them to (1) marshal adequate data on customer behaviors, revenues, and costs from their many production systems and (2) massage and model these data, using decision-support tools like artificial intelligence, CHAID (Chi-squared automatic interaction detector), logistic regression, neural network, hazard modeling, and formal designed experiments such as conjoint analysis and fractional factorials.

For example, by analyzing a set of self-reported customer choices, conjoint techniques measure the relative importance customers ascribe to given features of a particular product. This knowledge enables a bank to reconfigure products to maximize their perceived value to discrete customer segments. And this can often be done at a reduced cost by eliminating expensive product features with low relative customer valuations or utilities.

The point, quite simply, is that while banks seem to confront a trade- off between market share and profitability, in reality that trade-off is a spurious one.

Banks can moderate prices to pick up share and raise cross-sell ratios without impairing returns or they can maintain and even increase prices and still expand share, provided they know enough about how the customer behaves and what he or she wants. And it goes without saying that this is feasible not just with traditional bank products but also with the newer ones-e.g., insurance, annuities, mutual funds, and, potentially, financial planning.

But knowledge alone is insufficient without the capacity for sustained implementation. In our view, the apparent failure of some-perhaps most- data-base marketing initiatives stems from an inability to implement the insights generated by the new marketing. And this inability is traceable to numerous skills deficits and to infrastructural, cultural, and organizational shortcomings.

In future columns, we will consider the steps required to embed a refurbished marketing function firmly into an organization that will work to nourish rather than repress it.

For reprint and licensing requests for this article, click here.
MORE FROM AMERICAN BANKER