y Charles Marc Abbey First Annapolis For years, if not decades, acquirers have generated higher pricing for riskier merchants; and over the last six or eight years, acquirers have also used pricing mechanisms such as chargeback fees and the gross and gross-gross discount calculation methodologies to create de facto risk based pricing.

However, few acquirers have developed anywhere near the sophistication found, say, in the issuing business where issuers have made careers of measuring and exploiting behavioral differences in granular risk segments. Research recently completed by First Annapolis quantifies merchant behavior supportive of risk based pricing approaches.

Our hypothesis is that for risk-based pricing to be actionable beyond the de facto pricing mechanisms already in place at acquirers, more risky merchants have to be measurably less price sensitive than non-risky merchants.

To quantify this behavior, we developed a measure of price elasticity. Specifically, we measured the relationship between pricing and attrition within certain risk categories, using chargeback rates as a proxy for risk.

We examined merchant level pricing, chargeback and attrition data for hundreds of thousands of merchants. Essentially, for each chargeback category (see Figure 1 for a simplified example) we created samples-each of which included an average pricing level and an average attrition rate.

We then created a simple regression analysis relating attrition to pricing level, and the slope of this regression line represents how much attrition increases with a given increase in price.

This analysis demonstrated some significant variation in price sensitivity from segment to segment.

This type of analysis can include numerous other segmentation criteria: industry, merchant size, geography, etc., and indeed many of these criteria are highly relevant to elasticity.

There are many limitations to this type of analysis:

• The data include a finite range of prices, and price changes outside of this relevant range may have different attrition characteristics.

• Chargebacks are, of course, an imperfect proxy for risk as the financial condition of the merchant, return characteristics, and level of delayed delivery are key risk metrics that are not yet incorporated into the analysis.
• Individual acquirer's portfolios exhibit different levels of sensitivity to pricing than the industry in general.

• It is our speculation that these merchant behaviors are created primarily by the competitive environment (e.g. risky merchants are less attrition-prone precisely because they find it harder to find willing acquirers); therefore, these measures of elasticity will change over time as acquirer strategy and risk tolerance change.

Nevertheless, there are clearly segments with behavior sufficiently distinct to drive fine pricing strategies. As the industry matures and the organic growth of the market slows, there will be a handful of organizational capabilities that separate average acquirers from superior ones.

Portfolio management and optimization will be one such capability, and risk-based pricing has the potential be a key portfolio management strategy.

Charles Marc Abbey is a Partner at First Annapolis, a Baltimore-based consulting and merger and acquisition advisory company.