Retail Bankers Must Review Their ABCs

The following dozen observations are designed to provide a cursory overview of retail banking profitability, together with some suggestions on how to improve it.

(1) On balance, retail branch networks are adequately profitable. That is, taking into account all revenues (net interest income plus fees) less direct, indirect, and overhead costs, risk charges (based on average losses on loan products), and liquidity charges (debits for liquidity providers), the after-tax return on the equity imputed to thesse networks exceeds the bank hurdle rate.

(2) Nevertheless, without their trove of small-business checking accounts, a goodly number of these networks would not show an adequate rate of return. In other words, the purely retail component of retail branch networks sometimes detracts from rather than augments shareholder value.

(3) That's because most branched-based loan products - personal, auto, second deed of trust, student, etc. - either lose money or provide inadequate returns when fully costed, largely because of the small size of the loan balances. Additionally, some deposit products are also losers - e.g., short-term CDS and IRAS - for much the same reason. (At some institutions, under-one-year CDs break even only when the account balance reaches $25,000.)

(4) Most retail banking total relationships are unprofitable. At one typical institutions, Oliver, Wyman found that 29% of the customers provided 700% of the profit, while the other 71% of course subtracted 600% from this return.

(5) Account profitability drives relationship profitability. In only rare instances do banks that lose money on one product make money on the total customer relationship. That's in large part because most retail customers use only one product.

The fulcrum product is the checking account. In cases of multiprodcut usage, when the bank loses money on the checking account, it tends to lose even more money on the total customer relationship. By the same token, a profitable checking account tends to reflect an even more profitable overall relationship.

(6) Most regular checking accounts lose money. Although, on average, banks tend to make money on regular checking, returns are once again unbelievably skewed. Only 20% to 25% of regular checking accounts typically make money; the rest lose. That's because, in most competitive urbvan areas, an account balance of between $1,750 and $2,500 is required for breakeven performance, and balances of such size are not readily forthcoming.

(7) The above resulta are based on full costing for products and customers. However, managements often argue that since many indirect and overhead expenses will not disappear if selected products and customers are eliminated (a questionable hypothesis), it makes more sense to compute profitability based on only the marginal, or variable, costs of service. Yet many products and customers that lose money on a full-cost basis also lose money when only marginally costed. For example, in an analysis of returns at a representative institition, it was found that 65% of personal installment loans, 82% of auto loans, 60% of savings accounts, and 66% of CD accounts showed red ink even when only the variable costs of service were taken into account.

(8) Since many products lose money on a marginal as well as on a full-cost basis and since most customers who are unprofitable on a single-product basis are also unprofitable on total-relationship basis, banks have little to lose by repricing products to at least cover their marginal outlays.

If the customer leaves because of the repricing, the bank benefits because its cost savings exceed the forgone revenues. If the repriced customer stays, so much the better. Oliver, Wyman calculated that one client could add earnings equal to nearly a third of its total pretax retail profits by repricing just those accounts that lost money on a marginal basis. (9) In point of fact, however, most repriced custmers will not leave. Customer sensitivity to higher fees and prices is vastly overrated. In a survey of why customers close their checking accounts conducted some years back, higher fees were cited as the reason in less than 7% of the cases.

(10) Once having raised prices and fees, banks must collect these higher charges. To do this requires careful monitoring of fee waivers. Many banks exempt whole classes of customers from fees. But what hurts are not these policy waivers, but rather idiosyncratic fee forgiveness - waivers, generally by platform personnel, of charges in response to pressure from customers, both unvalued as well as valued ones.

Often these idiosyncratic fee waivers will lose a bank between 30% and 50% of posted charges. Control over this practice pre-supposes systems modification to identify the persons responsible for given waiver decisions.

(11) In the short run, repricing products to at least recoup marginal xosts is a permissible strategy. In the long run, however, products and customers must yield profits sufficient to cover full costs and produce a competitive shareholder return.

And while much can be done to trim delivery costs - by eliminating unprofitable branches, by upgrading teller and platform productivity through the use of more part-timers, by improving scale in support operations, and by paring unneeded staff functions and redundant managerial layers - there are also many opportunities to operate creatively on the revenue component of the profit equation through changes in the customer mix.

(12) Roughly speaking, customers can be divided into three categories: "A" customers are adequately profitable; "B" customers are breakeven types; and "C" customers lose the bank money.

In a typical institition, 25% to 30% of the customers are A-rated; 30% to 40% are B-rated and the remainder are Cs. A shift in the customer mix from 30-30-40 to 40-30-30- that is, a ten-percentage-point increase in the number of As - can often raise overall retail profits by as much as 60%.

Attracting more A-rated customers depends in part on reorienting the job of bank marketing. Many marketing departments should be worrying less about their efficiency - the cost of attracting the incremental customer - and more about their effectiveness - success in attracting the right kind of customer.

The standard for self-evaluation should be: "Is my department bringing into the bank a mix of new customers that is richer in As and leaner in Bs and Cs than was the case in the past?

Most large banks now have enough data (and smaller banks can readily acquire these data) to estimate the likely pattern of product usage and balance levels of various groups of potential customers. Based on this analysis, marketing can target potential customers according to the expected value, or expected profitability, of the relationship.

This value is arrived at by projecting cash flows (interest revenue and fees on deposit balances and loans less acquisition costs and applicable interest and noninterest costs of service) and discounting the net flows to their present worth, using a discount factor derived from the bank's hurdle rate.

If the net present value of the flows exceeds the amount of capital that must be allocated to backstop the customer relationship, that relationship should be pursued. If the net present value falls short of the needed capital, the relationship should be avoided, if at all possible. Although this type of analysis is increasingly used to gauge profitability in credit cards, it is not widely employed for the totality of retail banking relationships. It should be.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.