As Hurricane Sandy bears down on much of the eastern seaboard this week, there are important lessons for bank regulation.
Forecasters at the National Hurricane Center were quick to point out that just because Hurricane Sandy was rated only a Category 1 storm people should not underestimate its potentially damaging effects given its unusual combination of meteorological events. So too must organizations like the Office of the Comptroller of the Currency avoid the danger of ignoring potentially devastating risk events simply because the financial tsunami of 2008-2009 continues to weigh heavily on the minds of OCC management.
Imagine that in the aftermath of the Halloween Nor'easter of 1991 (the so-called Perfect Storm), the National Weather Service and decided to consolidate the National Hurricane Center with its National Storm Center and after its review of both organizations had decided to adopt the latter's hurricane forecasting system and abandon the former's models even though the NHC's hurricane monitoring capabilities were significantly more robust. Something similar has happened as a result of the consolidation of the Office of Thrift Supervision with the OCC as mandated by the Dodd-Frank Act.
The financial crisis exposed a significant number of missteps by OTS, to be sure, in reining in excessive risk-taking by some of the largest thrifts under its supervision. There is little doubt that restructuring the OTS was required.
However, in merging it in with the OCC, a little known but critically important activity being conducted by the OTS was abandoned, leaving the OCC virtually blind in assessing the impact of a major risk event for the industry.
In response to the thrift crisis of the 1980's, a comprehensive set of regulations in the form of Financial Institutions Reform, Recovery, and Enforcement Act also created a new agency to regulate the thrift industry, the OTS. It then embarked on a project to radically change the way it monitored thrifts' interest rate risk exposures. While a number of factors contributed to the thrift crisis, a severe mismatch in assets and liabilities gave rise to significant interest rate risk exposure during the period for much of the industry.
The OTS set out to construct its Net Portfolio Value model as part of Thrift Bulletin 13a, which required periodic reporting by thrift institutions of detailed aspects of their balance sheet under a variety of interest rate shock scenarios. The NPV model for its time was one of the more advanced regulatory efforts focused on assessing interest rate risk to come out of a U.S. bank regulator, and was unique among regulatory efforts of this kind due mostly to the fact that the thrift charter's focus on mortgage-related assets was prone to interest rate risk.
The NPV framework entailed modeling the cash flows of all assets and liabilities of the thrift balance sheet, including derivative instruments. The model was able to incorporate a wide variety of complex financial instruments and provide the OTS with detailed information on the sensitivity of changes in the economic value of a thrift institution to interest rate changes. Each thrift was required to submit information on each asset and liability on Schedule CMR, a companion to the Thrift Financial Report each thrift was required to submit quarterly. No other regulator adopted the OTS interest rate risk requirements, including the OCC, which was foretelling.
As of March 31 of this year, the OCC formally rescinded TB 13a and associated reporting requirements by thrifts and monitoring of interest rate risk exposure via the NPV framework. Currently, the OCC's official policy is to rely on a variety of handbooks and other bulletins to manage interest rate risk exposure of thrifts and banks.






















































As to the other comment that the OTS models were not as reliable as they might be, it sure beats not having anything to go by and if the OTS, now OCC made an investment in collecting better and more timely information, it would go a long way to providing OCC with insights into industry interest rate risk exposures that otherwise are not available at this time. Ignoring the problem simply because a model may not be as reliable as you would like is not a good reason to not focus on enhancing that capability.