Credit card issuers are laboring in an environment of rising delinquencies, possible interest rate increases, strong price competition, likely industry consolidation, and heightened regulatory demands.¹ These forces are generating strong pressure for better risk management, and the responses have been diverse, according to a survey of 13 major issuers, including 5 of the top 10 in the United States and 4 of the top 10 in Canada.² No single company excels across the board in risk management, we found. Four broad approaches that emerged in the survey suggest ways for issuers to use risk management to improve their profitability by as much as 5 to 15 percent.

1. Optimizing value, not minimizing risk. Many issuers still rely primarily on risk factors when they make underwriting decisions, assign credit lines and products, and manage accounts (Exhibit 1). These companies may leave money on the table by being either too stringent with profitable but slightly riskier customers or too cavalier with loans to the less risky. One alternative is to forecast the net present value of potential customers—an approach that helps issuers approve accounts for (and assign credit lines to) the people most likely to use them and to maintain a healthy balance. Another is options-based approaches, which suggest offering small initial credit lines to customers who seem risky but lucrative and then rapidly adjusting them.

2. Next-generation targeting. Intense competition for new accounts puts a premium on information about customers, but with more than half of the industry now using just one credit bureau, these traditional data sources don’t provide much of an edge. Some issuers are therefore tapping into risk-correlated data captured internally or by affiliates. Customers who exhibit prudence by purchasing fire extinguishers or maintaining strong auto insurance records, for example, are often good credit risks. Other issuers are bringing greater discipline to their modeling and customer segmentation efforts. Moreover, after a decade of increased automation, almost a third of the issuers we surveyed are adding an element of judgment to a select but substantial portion—20 percent or more—of the applications they review.

3. Customer-level management. The credit behavior of consumers is dynamic. Over a nine-month period, for instance, the FICO (Fair Isaac Corporation) scores of three of every seven consumers will change by more than 20 points, and the score of one of every five by more than 40.³ Yet only a third of the issuers we surveyed said they could react quickly to such changes by extracting customer-level information across accounts. The issuers that could were better able to manage their total exposure and losses (for example, by quickly extending or decreasing credit lines), to cross-sell products (by offering alternative ones to customers whose initial applications had been rejected), and to reduce costs (by consolidating collections calls). Issuers following such practices enjoyed not only higher revenues but also losses that were 15 percent lower than those of other issuers in the survey.

4. Smarter collections. How successfully issuers collect their debts varied widely. The roll rates⁴ of customers with bills 61 to 90 days past due, for example, are no less than 24 percentage points lower for top- than for bottom-quartile issuers (Exhibit 2). Traditional brute-force approaches (such as more frequent calls) for getting customers to pay up yield limited results. Leading issuers tailor a full range of collections strategies by segment, rigorously manage collections agencies, and employ innovative techniques such as mining text from collectors’ notes.

The range of performance we found across the industry suggests that different issuers rank the importance of deploying each of these approaches differently. But all issuers must undertake organizational change to make the most of their opportunities. Risk-management organizations are becoming increasingly independent, with centralized controls and resources. Few issuers, however, have aligned their risk-management incentives and their marketing and collections organizations, so these groups often don’t cooperate sufficiently.

Many issuers are expanding because scale benefits can be substantial: for example, issuers with larger applications-risk groups report lower average losses on new accounts. What’s more, an issuer can double its outstanding balances while increasing its risk-management staff by less than 40 percent. Meanwhile, talent is scarce and the impact of high turnover severe. Some issuers have responded by outsourcing selectively.

Given the size of the risk-management prize, tackling these problems is likely to prove worthwhile.