Yesterday we discussed bank models in general and things bankers should consider when running models internally. Today, we turn our attention to the little dot that has appeared on our risk management radar screen identified as "loan portfolio credit risk modeling" or "stress testing."
When using a credit stress model, it is imperative banks are capable at evaluating the validity of results generated. At a minimum, results must be translated into actionable items, to pay for the modeling process through better loan pricing, enhanced performance or reduced risks.
There are many data points to consider when credit stressing a loan portfolio to manage its risk, but one of the most basic is by doing a cost-benefit analysis. One way bankers can stress test and analyze a portfolio of loans is to purchase software and install it. This approach generally costs somewhere around $50K to $150K, plus an annual maintenance fee of 20% and the cost of about one-half of an FTE to run the model. At the low-end, that amounts to a total cost of about $110k or so. Another way community bankers can get this accomplished is to outsource the processing and analysis to us at a fraction of the cost. Significantly lower cost, robust analysis and a proven methodology are all primary reasons why hundreds of banks have already engaged us to complete this project.
Another issue we often see with banks that utilize in-house credit risk software relates to the complexity of the loans themselves. No system will ever replace a good lender and stand alone software without expertise to help bridge the gaps can easily turn into a glorified Excel spreadsheet that "sorts" instead of "analyzes" and "stresses."
It may not be intuitive, but loans originated in a $100mm institution are very similar to those originated in a bank that is $50B in size. The primary difference between the two is the size of the loan. Given that, the complexity of modeling should be roughly comparable for both institutions.
Another problem banks find when running in-house credit analysis software is getting help. Most software is provided by technology companies who are not schooled in credit analysis. Asking questions about specific credit concepts are often beyond the capability of most software help desks.
Another overlooked aspect of running in-house credit analysis software relates to the hardware it is installed upon. Applications that are sophisticated enough to properly analyze the credit of hundreds and hundreds of loans require dedicated servers and large relational databases. The sheer number of calculations and relationships involved is mind boggling.
Staffing is another big problem bankers run into when using in-house software to do loan portfolio risk analysis. To be effective, these systems must be managed by seasoned credit professionals skilled in loan analysis, credit management and modeling/simulation systems. Some banks will try to get by with a loan processing clerk or try to dump it onto the CFO as "part of ALCO." Credit analytics is completely different and requires a unique skill set. As a result, banks keeping this process in-house should be prepared to hire a competent analyst with credit modeling/simulation experience at an annual cost of about $80K per year. Since credit simulation is still reasonably new to the marketplace, the talent pool is also small and shallow.
Finally, training is a critical component of any in-house supported software solution, but particularly so given the complexities of credit analysis. This process is at the very core of "Safety and Soundness" and regulators have made it clear they are not messing around. Banks must have adequately trained and seasoned staff (and back up staff) available to support any in-house solution or risk regulatory consequences. In short, inadequate training can result in a failure to comply with policies, regulatory requirements, producing erroneous information, improper credit management decisioning and a myriad of other problems and issues.
When it comes to detecting loan portfolio credit risk, trying to do it all in-house is a complexity most community bankers can do without. Give us a call or email and we will train you to operate the radar screen and show you how easy it is to get started down the path of outsourcing with a solution that really works.