Benchmarking Techniques for Credit Portfolios

This article is the second in a series of two ‘Benchmarking’ articles. The first article discussed the process of benchmarking, and the importance of monitoring the key performance measures of your credit portfolios relative to the industry trends.

It is common practice for organizations to use averages and percentage measures for comparison, as these make it possible to compare portfolios of differing sizes against each other on a common basis. In the first article, we illustrated how the market positioning and the individual performance of a credit portfolio could be measured relative to the overall industry in a series of industry ranking graphs.

However, whilst averages and percentages allow benchmarking on selected measures, it would be ideal if absolute values, such as total balance, could also be used to benchmark performance and market positioning relative to the industry.

The issue in comparing absolute values is that each portfolio differs in terms of size. As an illustration, comparing the total balance of a portfolio with over 2 million accounts against one with only 100,000 accounts is clearly inappropriate. In order to enable this type of comparison, it is necessary to use a process termed ‘normalization’.

In this article, we will discuss how portfolios, irrespective of size, are ‘normalized’. Once this process has taken place, we show how a series of key performance measures can be benchmarked relative to the industry overall.

What is ‘Normalization’?
Normalization is a process used to quantify each portfolio’s performance in the industry, relative to its proportional size.

For example, if we have 5 portfolios of varying size:

• Portfolio 01 – 2,000,000 accounts
• Portfolio 02 – 900,000 accounts
• Portfolio 03 – 550,000 accounts
• Portfolio 04 – 300,000 accounts
• Portfolio 05 – 100,000 accounts

By summing the number of active accounts across all of the portfolios, we are able to calculate the proportion of each portfolio’s size relative to total number of accounts. In this example, the total number of accounts is 3,850,000 and Portfolio 02 makes up 23% of this total.

Using this as a base-line, we can now make the assumption that if Portfolio 02 has 23% of the total accounts, then it should have 23% of the total balances to be proportional. If Portfolio 02 has 30% of the total balances, then it has a higher proportion of balances that it should have comparative to its proportion of accounts.