Imad Chahboun, Credit & Policy Analysis at Federal Reserve Bank of Boston discusses with us modeling techniques for CECL requirements to accommodate changes.
Disclaimer: The views expressed herein are those of Imad Chahboun and do not reflect those of the Federal Reserve Bank of Boston.
Imad, can you tell the Risk Insights’ readers about yourself and your professional experiences?
I have been part of financial risk quantification for more than 15 years where I had the opportunity to lead modelling activities for risk management and capital measurements. My credit risk experience covers retail portfolios such as cards and mortgages as well as wholesale and investment banks portfolios. I also spent a fair amount of time studying low (or no) default portfolios and assess associated risks. My main focus, in addition to aspects of quantitative modelling, has been on understanding the economics of each product and underlying risks. Recently, I have been supervising large and complex financial firms for wholesale and counterparty credit risk.
How can financial institutions best leverage modelling techniques towards CECL implementation?
Having modelled these products under Basel and stress testing frameworks, I think that firms will be able to build on a significant knowledge and understanding of these products credit risk economics. Although CECL is a new framework, the required modelling skills are not. For example, valuing asset based securities requires modelling underlying portfolio cash flows such as defaults, recoveries and prepayments, for the duration of these products’ lives. Modelling techniques such as vintage curves, roll rates, cash flow models/aggregations, etc. are already common practices relied on in the literature, by vendors. A key element is to stage, Internal and/or vendor data to produce default performance and recovery up to product’s maturity. Existing stress testing models allow forecasting PIT default risk for a foreseeable future which should be amended to meet CECL requirements such as product’s life forecast and new vs. existing products.
Can you outline some of the challenges and differences in reserving for long term vs short term instruments?
In my opinion, there are several challenges involved in modelling credit risk for a longer horizon. These include assumptions on the future state of the economy and the position in the credit cycle, the increased uncertainty of point estimates which will rely on subject matter assumptions and economic outlook. Data quality and relevance is another challenge where firms without sufficient internal data will have to justify the relevance of external data to own portfolios/products. The maturity of credit portfolio becomes relatively more important where, depending on the position in credit cycle, more mature portfolios will require less reserving and vice versa. Another challenge is model integrity testing where backtesting and benchmarking long term credit risk performance present additional challenge.
How do you see the role of the risk management professional within the CECL department changing over the next 6-12 months?
In my opinion and for near term, risk management professionals are expected to play significant role in assessing that systems, processes and modelling capabilities are being integrated and key millstones are well defined. They should also support GAP analysis regarding existing systems, models and production capabilities. In particular, assess how existing models and systems could be amended to satisfy CECL requirements.