News & Commentary

10:14 AM
Chuck Nwokocha, Sageworks
Chuck Nwokocha, Sageworks
Commentary
50%
50%

How To Improve Credit Risk Modeling For Better Credit Quality

Banks have to evolve their credit risk modeling, and including cash flow in their credit risk analysis for commercial loans is one place to start.

As an institution’s portfolio grows, it can be easy to get away from the basics. Effective risk identification starts with the evaluation of individual credits through solid credit risk modeling.

This doesn’t necessarily require the licensing of oftentimes expensive, statistically-driven models and databases, which could be overkill or ill-suited for institutions that don’t have the volume of data to support a model.

Rather, credit risk modeling describes the analytical due diligence a bank performs to assess risk of borrowers. And since this risk is dynamic, the model must examine the ability of a potential borrower to repay the loan as well as non-financial considerations such as character, management ability, environmental conditions and factors.

Some financial institutions evaluate entire loan relationships, others prefer to rate each facility, and still others rate both relationships and facilities. Whichever approach the financial institution chooses, the following considerations are necessary for developing the credit risk model:

Credit Risk

Credit risk factors could vary slightly, institution to institution, depending on the portfolio size, concentration, borrower composition, loan types, location, etc. As such, each of the following should influence the credit risk model.

- Determining risk factors.

- Understanding credit quality (risk grading/risk rating).

- Likelihood that a business/borrower/relationship may default on its financial obligations.

- Accommodating different types of loans as well as industries (different industries require different capital structures).

Annual Reviews

Credit risk doesn’t end with the loan decision at origination; rather, risk profiles of borrowers can change throughout the year and the relationship. To identify risk within the portfolio, credit risk models should include an annual review of loans.

- Set a minimum review period that allows continual and regular monitoring and reassessing of risk.

- The reviews will lead to early identification of deteriorating conditions.

Trend Analysis

Companies rarely remain in a static condition. However, far too often, the decision to grant a loan relies on cash flow analysis only, which examines a company at a point or period in time. It does not include a trend analysis, showing whether the company’s performance is improving, stable or declining. Although extremely important, cash flow cannot be the only determinant in credit quality.

Even the definition of cash flow can be elusive and can carry many definitions. Cash flow is generally used to describe cash generated between two periods. Sometimes it refers to cash leftover after operations (i.e. profit). Credit analysis is much too complex to rely on just a single indicator, hence the suggested use of related factors. At a given point in time, the business may have positive cash flow. The grey dashed line represents a cutoff debt service coverage ratio (DSCR) of 1.25. A loan decision based only on the DSCR threshold of 1.25 would be unwise in the case of the deteriorating business. Instead, an analysis should use trend data to determine if the business is growing or deteriorating.

A more comprehensive risk analysis based on the 5 Cs of Credit would include this type of longitudinal analysis and capture risk from externalities. For example, it would have been unwise to make a loan to a real estate company at the height of the real estate crisis in 2008. A look at the Conditions would have indicated that the market environment and the state of the economy were in bad shape.

Another element to consider is industry comparisons. Performance ratios are more meaningful when viewed in context of the borrower’s industry.

Chuck Nwokocha is a senior risk management consultant at Sageworks, where he advises financial institutions on risk and portfolio management.

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Greg MacSweeney
50%
50%
Greg MacSweeney,
User Rank: Author
10/31/2013 | 1:51:55 PM
re: How To Improve Credit Risk Modeling For Better Credit Quality
There are many types of data that companies are using to determine credit worthiness. It's well documented that insurers can use credit reports to determine the likelihood that a driver will get into an accident (a person with bad credit is more likely to get into accidents, or violate traffic laws). I'm sure banks can use other types of data to determine if someone is a good risk.
Byurcan
50%
50%
Byurcan,
User Rank: Author
10/30/2013 | 11:54:28 AM
re: How To Improve Credit Risk Modeling For Better Credit Quality
Insurance companies can also use third-party data to root out fraud, correct?
ali15871
50%
50%
ali15871,
User Rank: Apprentice
10/30/2013 | 3:49:30 AM
re: How To Improve Credit Risk Modeling For Better Credit Quality
It also includes the what kind of information required for analysis and source of information to evaluate the credit worthiness of borrower.
Nathan Golia
50%
50%
Nathan Golia,
User Rank: Author
10/25/2013 | 7:57:24 PM
re: How To Improve Credit Risk Modeling For Better Credit Quality
Insurance companies are big on new third-party data sources because of their ability to make the transaction process quicker at the time of policy issuance Gă÷ things like prescription history, for example. I don't feel like banks have had the same issue with the user experience at the time of customer acquisition, but there are definitely options out there that could make it even sleeker and perhaps even be more predictive of the potential risk.
Greg MacSweeney
50%
50%
Greg MacSweeney,
User Rank: Author
10/23/2013 | 10:05:53 AM
re: How To Improve Credit Risk Modeling For Better Credit Quality
I think that banks are considering a number of non-traditional data sources to help determine credit, such as social data. However, i'm not sure how widespread it is and how much of that type of data has really made it into day-to-day analysis, or is still in the pilot phase.
Nathan Golia
50%
50%
Nathan Golia,
User Rank: Author
10/22/2013 | 8:24:50 PM
re: How To Improve Credit Risk Modeling For Better Credit Quality
What about the potential for different data sources than the traditional ones in determining creditworthiness? Certainly there must be some options.
Greg MacSweeney
50%
50%
Greg MacSweeney,
User Rank: Author
10/21/2013 | 10:31:06 AM
re: How To Improve Credit Risk Modeling For Better Credit Quality
The "annual reviews" and "trend analysis" seem to be the parts of a credit analysis process that are most lacking. There is a massive focus on credit worthiness at the beginning of the relationship, but I'm not sure how much ongoing analysis happens after a year or two. Granted, banks did a large sweeping review of existing mortgages 2-3 years after the credit crisis, but I doubt they have gone back again to do the same analysis. Am i wrong?
Byurcan
50%
50%
Byurcan,
User Rank: Author
10/18/2013 | 6:13:15 PM
re: How To Improve Credit Risk Modeling For Better Credit Quality
Good credit risk modeling is especially important in the current regulatory environment, and the importance of Know Your Customer initiatives.
Register for Bank Systems & Technology Newsletters
White Papers
Current Issue
Bank Systems & Technology Oct. 14, 2014
Bank Systems & Technology's new Must Reads is a compendium of our best recent coverage of customer analytics. Learn what big data means for banks, meet Wells Fargo CDO Charles Thomas, find out how to connect with your Gen Y customers, and more.
Slideshows
Video
Bank Systems & Technology Radio
Archived Audio Interviews
Join Bank Systems & Technology Associate Editor Bryan Yurcan, and guests Karen Massey and Jerry Silva from IDC Financial Insights, for a conversation about the firm's 11th annual FinTech rankings.