The nature and level of loss exposure is changing unfavorably in acquiring, in our view, though this increase is not yet apparent in the loss data. This rising exposure is due to a number of factors, ranging from the continued growth of e-commerce to the risk characteristics of gift cards to the contingent liability surrounding data security. These comparatively new risk issues are in addition to the traditionally volatile nature of acquiring risk.
It is our contention that there is a case in the U.S. for an increased focus on, and investment in, acquirer risk management, especially given the trends outlined above. Above all, merchant processors cannot afford to ignore risk-management capabilities and the effort necessary to achieve them.
First Annapolis recently completed research examining U.S. acquiringindustry writeoffs over the time period 1991 to 2004 and including 20% to 30% of U.S. acquirers, depending on the year. Losses in acquiring, of course, result from business failures in which the merchant is unable to meet its obligations (primarily consumer disputes or chargebacks) and from instances of merchant fraud. The weighted average loss level for acquirers over this time period was 2.1 basis points of sales volume, a level that compares favorably to earnings over the same time, which ranged between 8 and 12 basis points.
Our best estimate of 2004 nominal losses for US bank card acquirers across the industry is $350 million, give or take. These loss levels have not yet been impacted materially by gift-card-related risk or data-securityrelated contingent liability.
Episodic Losses
The simple average of the loss levels was 4.5 basis points, indicating that smaller acquirers experienced much higher losses than their larger counterparts. We believe there are four explanations for this difference. First, there are several small acquirers that, over the years, have specialized in lossintensive high-risk segments, and while large acquirers often also have such a sub-segment in their broader portfolios, the large acquirers’ loss levels are not dominated by this segment as a specialist’s would be.
Second, the large acquirers generate a greater proportion of volume from large merchants, which, over this time period, have generated lower levels of losses than regional and local merchants. (Risk management on such large merchants is also typically much more intensive than on small merchants.)
Third, small acquirers are, in general, less geographically diversified than large acquirers. Fourth, the large acquirers have developed technology advantages, in some cases, over many of the smaller players that, by virtue of their size, have difficulty making similar investments.
It is not entirely a given that small acquirers will generate higher losses than large ones, though, and an episode of business failure among large merchants would reverse the outcome, at least to a degree.
Losses in acquiring often have a binary quality to them in the sense that material individual losses are rare but can be severe when they do occur. So an acquirer can go many years with few (and perhaps no) material losses only to generate significant losses during an abnormal year, perhaps even from one or a small number of loss events. This episodic nature of losses has led some acquirers to go through cycles of investment and disinvestment in risk management that correspond to the recency of significant loss episodes.
In our analysis, after eliminating acquirers that specialized in high-risk segments and generated high loss levels in the ordinary course, we examined the maximum loss in each year. The maximum loss represented an acquirer that, in the ordinary course, generated losses consistent with the rest of the industry but in a particular year generated high losses that were an outlier relative to the rest of the industry. This maximum loss level was 7.4 times the weighted average loss level for the 14-year period.
These peak loss levels, in nine of the 14 years, caused the acquirer to generate negative earnings in that year. Two acquirers appear on the peak loss list twice, not in consecutive years in either case.
Complex future
In short, though average losses tend to be a mild financial issue for an acquirer, peak losses can be a material issue, and it is not possible to evaluate the riskmanagement outcomes of an acquirer over a short time.
Losses in acquiring do have some cyclicality. Average losses for the recession year 1991 were 14% above the 14-year average, and in 2001 were 48% above the 14-year average. By contrast, average loss levels from the expansionary period 1994 through 1999 were all below the 14-year average.
Likewise, the peak losses in 1991 and 2001 were 13.5 times and 11.5 times the average losses in 1991 and 2001, which is well above the 7.4 average multiple for the 14-year period. This cyclicality is primarily driven by business failure.
The future may hold a more complex risk environment for acquirers. Acquirers are generating a greater proportion of their volume from inherently higher-risk merchants now than ever before, to a large degree as the result of the shift in volume from physical to Internet retailing. cards are a form of so-called delayed delivery, which is when a payment occurs prior to the delivery of the goods or services. The exposure a merchant creates is highly related to its delayed-delivery characteristics, and as gift cards proliferate, they introduce delayed-delivery risk to segments that previously had little or no such risk (restaurants, for example). Gift cards also introduce a fraud risk that is especially difficult to monitor—specifically, the risk that merchants in financial distress will sell gift cards at a discount to face value to raise cash, thereby increasing exposure significantly in the run-up to a failure.
Also, data-security risks (though not traditionally an exposure risk units manage) create significant contingent liabilities for acquirers, both from potential card-association fines and from the potential for issuer damage claims that would follow a data compromise. It is difficult to size the datasecurity exposure, but it could exceed credit and fraud exposure.
None of these issues has made its full impact on industry loss levels yet because they are still evolving or emerging issues. In addition, although it is possible to overstate this point, much of the investment in risk management acquirers have made in the last decade has been related to productivity rather than to improvements in loss avoidance, per se.
Though it can be hard to observe because of the sporadic nature of losses and the absence of good industry data, we believe risk management capabilities already separate stronger- and weaker-performing acquirers. The rising complexity of acquiring risk cannot help but accentuate these differences, and therein lies the case for investment in acquiring risk management.
This NEWS ARTICLE
Digital Transactions
