CECL – Managing increased sensitivity to macros assumptions across loss forecast models

CECL – Managing increased sensitivity to macros assumptions across loss forecast models

By Chris Varvares, Vice President and co-head of US Economics, Macroeconomic Advisers, IHS Markit

Interview ahead of CECL 2019 (Get 15% discount on the Congress using code: CECL99)

CECL 2019 is taking place in New York City on 27-28 March, 2019 – find out more here www.cefpro.com/cecl

Can you please tell the Risk Insights readers a little bit about yourself, your experiences and what your current professional focus is? What for you are the benefits of attending a conference like CECL?

I have roughly 35 years of experience observing, analyzing, modelling and forecasting the US economy. I am one of the co-founders of Macroeconomic Advisers, a firm acquired by IHS Markit in the fall of 2017.  My partner and I now head up the US macroeconomic practice for IHS Markit, including the regional team and the scenarios/stress-testing related services.  I served as president of the National Association for Business Economics In 2008-2009, a very interesting and challenging time for the profession. My current focus is on insuring that we are providing our clients with the very best of a range of forecasts, models and scenarios to permit them to more easily meet their business planning and risk management responsibilities.

A conference like this exposes attendees to a range of approaches and insights into how to meet this challenging new requirement while sharing experiences with other practitioners.

What, in your opinion are some of the key challenges with the increase in sensitivity to macro assumptions in forecast models?

As the outlook for the economy evolves through the business cycle, estimates of credit losses will necessarily change to reflect the new “expected” environment. CECL thus introduced a sensitivity of loss estimates to the forward forecasts. Using multiple scenarios, with appropriately changing weights, and using processes that look for material changes in assumptions can help reduce that sensitivity.

How can sensitivity to macro assumptions impact portfolios, and are there any portfolios more at risk than others?

Macro assumptions can change across many dimensions including interest rates, liquidity, wealth, employment and unemployment, spending by consumers and business, cashflows needed to support loan payments, asset prices and more.  For variable rate loans, large swings in rates can impact the ability to pay. Sharp declines in asset prices alters the “pay or delay” calculus on the part of borrowers. Unemployment for individuals or a sharp decline in cash flows for businesses can impair ability to pay and therefore increase defaults.  So yes, the nature of the changes in macro assumptions across scenarios can certainly impact portfolios differentially.

How can institutions recalibrate scenarios to track sensitivity?

Institutions with the resources to do so can and should review how various scenarios will impact their loss experience so that they have an understanding in advance of how a range of economic outcomes might impact the balance sheet.

What are some non-linearity considerations that need to be made when managing increased sensitivity?

Households and firms maintain financial cushions (when they can) to allow them to handle both periodic unexpected expenses (e.g. the washer needs to be replaced) and shortfalls in cashflow (e.g. our biggest customer just went with another supplier!). These can be liquid assets such as cash or easily sold securities, or the availability of credit. For small adverse surprises, the number of households or firms likely to miss a loan payment would be small because the financial cushions are adequate to fill the gap.  The larger or more prolonged the adverse event, the more likely the cushion will prove inadequate.  So, defaults rise with the size and duration of the adverse shock, and it turns out the pace of defaults rises more than proportionately with the size and duration of the shock. For example, a one percentage point rise in the unemployment rate may result in an x-point rise in the charge-off rate on the credit card portfolio, while a two percentage point rise in the unemployment rate may result in a 2.5x-point rise in the charge-off rate.  This makes it necessary to consider a range of adverse shocks of varying size to capture the inherent nonlinearities in the PD and LGD calculations.

How do you see the industry progressing over the next 12 months as we move towards implementation?

The early adoption of the standard has resulted in confusion about just what new processes will be deemed as meeting the standard. Thus, we are seeing many varied responses. Over time there may naturally emerge a consensus and lenders will tend to converge around a narrower range of processes, or FASB or other regulator will issue new guidance to clarify expectations to which all firms will need to adapt.

You may also be interested in…

Sign up for your free Risk Insights account: