Magazine article The Journal of Lending & Credit Risk Management

How to Identify and Evaluate Industry Risk in a Loan Portfolio: A Five-Step Approach

Magazine article The Journal of Lending & Credit Risk Management

How to Identify and Evaluate Industry Risk in a Loan Portfolio: A Five-Step Approach

Article excerpt

Industry analysis has become an important component of portfolio management practices at many financial institutions. The key to analyzing industry risk at the portfolio level is having a clear and comprehensive approach that articulates industry risk in a way that all areas of the bank affected by industry analysis can understand and use. In this article, the author has outlined a five-step approach to help financial professionals identify economic industry risk in a loan portfolio and to prioritize the industries that should be examined more closely.

As the economy moves into its seventh full year since the last recession, many financial institutions are attempting to pinpoint the areas of their loan portfolio most at risk when the next economic downturn occurs. The economy has a direct impact on a commercial loan portfolio through its industry composition. For example many of the large loan losses experienced by financial institutions during the 1990-1991 recession can be traced back to their concentration in real estate and construction industries. Isolating and managing industry risk in a commercial loan portfolio may be one key in weathering the next downturn. Economic industry risk in a loan portfolio can be identified in five steps. This exercise allows the lender to prioritize the industries that a financial institution should analyze more closely.

Step One: Identify Industry Composition and Concentrations

The first step in managing industry risk is to identify the current industry mix of the loan portfolio and determine if these industries are related. One of the true challenges of industry analysis is deciding how to define an industry as industries can be classified in many ways and at many levels. For example, Standard Industrial Classification (SIC) codes can be used to specify an industry either at the two-digit (Paper), three-digit (paperboard containers and boxes), or more detailed, four-digit (corrugated boxes) level. However, a more meaningful way to analyze industries may be to group them into industry sectors.

Grouping industries into sectors facilitates concentration analysis. A sector should be comprised of correlated industries that move together over the business cycle. Ignoring the correlation among industries can lead to a serious underestimation of credit risk.(1) For example, if 25% of exposure in a loan portfolio is in the carpet and rug industry, 25% in lumber, 25% in real estate, and 25% in construction materials, an argument could be made that, using these industry definitions, the portfolio is evenly distributed across industries and no concentrations exist. However, all of these industries will be affected by a decline in housing starts, which may lead to increasing losses in all areas of the portfolio. Defining these four industries as one sector shows the portfolio is concentrated in one industry that will be effected by similar economic events.


Creating industry sector definitions begins with identifying and obtaining a single time series for all industries that dates back through at least one business cycle. Some examples are real revenues, sales, operating margins, cash flows, or business failure rates by industry. After obtaining the data, correlation analysis can be performed to measure the degree of association between industries. For example, if correlation analysis for growth in cash flows yields the results in Figure 1, paper, printing and publishing, and wholesale paper products can be put in one industry sector while food, restaurants, and grocery stores may be placed in another industry sector.

A financial institution may choose to use more complicated statistical procedures like factor analysis or cluster analysis to group industries into sectors. Factor analysis groups data into categories according to their underlying similarities. Cluster analysis will create clusters of industries such that industries in a given cluster are similar to each other and not similar to the industries in other clusters. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.