For 3Ys in a row, over 50% of Americans are exercising the recommended amount. Strangely, that is not making a difference on the obesity rate, which continues to rise. It reached an all-time high of over 31% last year according to the CDC. It seems the daily recommended amount of 150 minutes per week (21 minutes per day) of moderate exercise may not be enough after all.
As we continue to look for a magic formula to keep us healthy, bankers may be looking for a magic tool to help guide them in the countless decisions they must make around the financial model.
For regulators, a focus on modeling efforts at banks has ramped up sharply in recent quarters. Driving this regulatory shift is the simple fact that modeling at banks has mushroomed over the years into a myriad of tools to help manage and forecast a range of banking decisions. Bankers need to use models to improve and streamline decision making, but some of those models are far from perfect.
Indeed, the use of models now comes with the task of managing the risk of those models. Recently, regulators issued updated guidance for model usage to reflect the evolution taking place in their use by banks. Supervisory Guidance on Model Risk Management, offers new insights and guidance on how models can be used, applied, and managed.
While the FDIC said the new guidance applies to banks with >$1B in assets, it also pointed out that the advice pertains to any size bank whose model usage is "significant, complex or poses elevated risk to the institution." That is pretty broad indeed, so community banks too would do well to identify model usage at the bank and to certainly follow the guidance around critical ones.
According to one consultant, community banks have been slow to adopt robust model risk management procedures. This is perhaps because banks are reluctant to invest in expensive processes, when they do not perceive the risk as significant and it seems regulators are giving them a pass.
For community banks, one of the key takeaways is that your bank's model risk management process should reflect the complexity and extent of your bank's model usage.
The beauty of a model is that it can produce simplified results from complex data and make decisions easier. The danger is that those results can be wrong or skewed, leading to poor decisions. Even when a model's output is correct, it might be improperly applied. Resulting decisions can lead to losses or issues.
McKinsey & Co. research finds a simple model coding error at one financial institution led to decisions that resulted in several hundred millions in losses. That is a big bank, but the compounding effect and impact can reverberate at community banks just as much - albeit on a smaller scale.
One way to mitigate model risk is by having thorough documentation. A robust process can help control the development, implementation and use of models. This may not be easy, but it is very necessary given the complexity of the industry.
One area where banks often falter in their documentation is in explaining a model's limitations and weaknesses. Effective documentation needs to explain to users the model's background, its intended uses, and how it works. Documentation should be explicit in how a user might go wrong, including exactly what data and sources should be used, and how a user might make judgmental assumptions not supported by the model.
Model usage has been steadily rising in the industry, so community banks should be aware of the risk of using these models to inform decisions. In the end, model risk management can be as important as the model itself.