The Attractiveness of Risk-Based Strategies in Acquiring

In the issuing business, financial institutions have created tremendous value through highly nuanced risk-based pricing and solicitation strategies.

It is certainly also true that many acquirers in the United States and elsewhere have made very deliberate decisions regarding their appetite for, and approach to, risk. And, a segment of acquirers have actively pursued high-risk segments as their primary strategy.

However, based on recent First Annapolis research, it is not generally the case that acquirers have been successful in generating higher returns by taking more risk; rather, it appears that, on average, risk-centric strategies are unattractive.

Acquirers have been successful in most markets in pricing riskier merchants at higher levels than less risky merchants, and this has occurred both deliberately and de facto. Acquirers taking on higher risk accounts have faced less competition than for lower risk accounts, and have simply been able to capture higher discount rates. Furthermore, fees such as chargeback fees have become near universal in the United States, and more and more prevalent in major markets such as Canada and the United Kingdom.

Pricing Correlations

In First Annapolis’ 2006 pricing survey, 93 percent of U.S. acquirers use a chargeback fee (and 43 percent use a chargeback fee of $20 or higher). These fees act as de facto risk-based pricing. All of the company’s analysis of merchant- level data indicates high-risk merchants are as much as three times more price inelastic than average risk merchants. This means that as pricing rises on high-risk merchants, their attrition rates rise much more slowly than average-risk merchants.

Separately, First Annapolis’ analysis of 59 acquirers and 160 data points from recent years in the U.S. market indicates a positive correlation coefficient of .61 between an acquirer’s portfolio- level net spread and its loss levels. In short, as an acquirer’s risk level rises (as measured by its loss rates), its revenue improves. This result is, no doubt, intuitive to most acquirers. Higher Servicing Costs However, higher risk merchants are more expensive to service than other merchants because they require more credit and fraud monitoring, ultimately more collection, compliance, and work out activity, and, of course, because losses themselves impact profitability.

A high risk acquirer may have double or more the fraud investigation staff, for example, of a similarly sized average risk acquirer. Nevertheless, if acquirers were being rewarded for taking more risk, we would expect a positive correlation between loss levels and acquirers’ profit margins.

This means that as an acquirer took more risk, we would expect the acquirer’s profitability to improve and likewise, as it took less risk, we would expect its profitability to fall (over time and on average).

However, contrary to our prior expectations, First Annapolis’ analysis of the U.S. market measures a slight negative correlation between loss levels and profit margins.

Other Contributing Factors

It is possible that factors other than risk levels are influencing the results of this narrow analysis. For instance, high-risk acquirers are often smaller players operating at lower scale levels, and it may be that their scale levels are more of a driver of profitability than the risk-return tradeoff. However, among small acquirers, we would expect to see a relationship between risk and profitability, and we do not. It is also possible that acquirers taking large exposures but generating low loss levels would make loss rates a poor indicator of risk for this sort of analysis. In separate research, First Annapolis examined the level of credit risk exposure at 12 acquirers, and the rate at which this exposure generated losses. The company, in fact, concluded that the acquirers with the highest amount of exposure per dollar in volume tended to have the lowest level of losses per dollar of exposure.

Nevertheless, there was a positive correlation between exposure and losses for these acquirers. So, we concluded that the usefulness of losses as a proxy for risk may be weakened by the phenomenon of acquirers with large exposures and low loss levels, but we should still see positive correlations between losses and profitability. As we mention above, we see negative correlations.

First Annapolis also examined 10 individual U.S. acquirers’ time series results over a minimum of four years to see how differently individual acquirers may perform than average. The company also wanted to see if there were significant lag effects in profitability or loss levels that would influence our analysis.

Seven of these acquirers were top 10 acquirers, and three were top 20 acquirers. Two of the 10 were arguably pursuing higher risk merchants as a principle strategy. Of the 10, five had negative correlations between their loss levels and profit margins, two acquirers had essentially no correlation between losses and profits, and three had positive correlations. So, for the majority of these acquirers, over time, a risk-return tradeoff was not evident. However, three of the acquirers, by all appearances, deliberately managed their risk levels up or down over a period of years (one up and two down) and had a corresponding increase or decrease in returns. We believe this means that even if the acquiring industry on average is not benefiting systematically from riskreturn tradeoffs, individual management teams have gotten the algebra right.

Regional Differences

All of these results are market-centric. In other words, other geographic markets will be influenced by market conditions (principally pricing) that could result in different outcomes.

Admittedly, First Annapolis has less data outside the United States, but it examined 22 acquirers in Asia-Pacific, Western Europe, Canada, and Latin America. Averaging across all of these regions, losses were uncorrelated to profit margins. However, looking at individual regions, losses were significantly, negatively correlated to earnings in Asia and Latin America, and significantly, positively correlated to earnings in Canada. In Europe, losses were uncorrelated to earnings.

In Asia and Latin America, the results suggest that higher risk and elevated losses are not offset by higher revenue (and vice versa). This argues for risk reduction strategies in these regions. By contrast, in Canada, acquirers appear able to manage risk levels up and down and generate corresponding increases and decreases in earnings, unique in the world on the regional level. This suggests that a strategy of taking on more risk to generate greater returns may be an economically rational strategy choice in Canada.

Canada notwithstanding, First Annapolis believes these results present a picture that it is rare that acquirers are generating value through risk-based strategies, and that this observation holds up in most parts of the world.

However, the quantitative differences in correlation from region to region do emphasize that there are unique riskreturn characteristics in local markets, and if we had sufficient data to measure correlations in the component national markets of these regions (compare, say, correlations between Australia, Hong Kong, and India, for example), we would expect to see even greater variation. Some of these individual countries may have positive correlations between risk and return even if the region does not.

Though individual institutions may be able to manage the risk-return tradeoff in acquiring effectively, there is little evidence this is occurring systemically (excepting Canada), and little evidence that the higher revenues associated with riskier merchants are sufficient to offset losses and higher expense levels of risk-based strategies.

These observations do not argue against pursuing a risk-centric strategy, per se, but they do underscore the degree of difficulty in implementing such a strategy. The market in general does not fully price risk, so individual management teams implementing a risk-centric strategy must be smarter than the market. First Annapolis does not consider this a particularly happy margin of error, or to paraphrase its favorite Warren Buffet quote: “When a management team with a reputation for brilliance meets a business model with a reputation for poor fundamental economics, it’s usually the business model that keeps its reputation.”