The Two Faces of Change: Current Expected Credit Losses

By Tim Lee, FI Consulting

FI-Blog-facesFI Consulting Manager, Joe Feldmann, was invited to speak at the 2014 Mortgage Bankers Association’s accounting conference on the recent Accounting Standards Update (ASU), Financial Instruments – Credit Losses, Subtopic 825-15.  Underneath the dryly named ASU is one of the most significant paradigm shifts to emerge in recent years from the accounting world (and, we would assert, the financial crisis): transition from the prevailing “incurred losses” framework for booking loan portfolios to a “current expected credit losses” (CECL) approach.

To provide a quick recap: Under current US GAAP, adjustments to the recorded value of loan portfolios held on book are made on an “incurred losses” basis. Loans must not be written down to reflect deteriorating fundamentals, such as increased credit risk, until after a qualifying event—such an actual or probable default—has occurred. The weaknesses of this setup became glaringly apparent during the 2007-2008 financial crisis, when lender after lender made repeated visits, sometimes within a quarter, to what commentators began to term “the confessional,” with write-downs in the billions of dollars. According to the St. Louis Fed, total reported annual profits at US commercial banks plummeted from $77.5 billion in 2006 to $1.0 billion in 20091, principally due to massive losses on loan portfolios.

The backward-looking quality of today’s loan portfolio valuation framework brings to mind the ancient Roman deity Janus, the god of transitions. Janus was always portrayed as having two faces: one looking forward, one looking back. Bound by the incurred losses framework, during the financial crisis corporate accountants were condemned to be a modern-day analogue to the backward-looking face of the old Roman god, looking with mounting consternation out the rear of their car (or, perhaps, chariot) at steadily higher piles of smoldering wreckage littering an increasingly pockmarked road. Much as they might have wanted, SFAS 5 bound them to record only what they saw out the rear window. This was true even while their screaming compatriots among bank management, the investor community and the regulators gripped the controls with white-knuckled dread, looking forward through the windshield at the subprime mortgage crisis truck bomb erratically hurtling over the double yellow line toward a spectacularly gruesome head-on collision with the hapless group.

The shift to a CECL paradigm permits the accounting community to join their investor and regulator counterparts in seeking a forward-looking view of credit losses in a portfolio. As proposed, FASB’s CECL guidance mandates the incorporation of both incurred losses and a “reasonable and supportable” forecast of future losses in the assets—neither worst-case nor best-case, but the likeliest middle ground consistent with current circumstances and relevant history. In a general sense, commentators so far appear to agree that the shift to CECL will add materially to loan loss reserves; estimates range from 10 to 50 percent above current levels. CECL is likely to bring the reporting of non-trading assets into closer alignment with the mark-to-market treatment of trading assets, which is expected to remain in place.

A crucial implication of the new CECL framework is the need for additional data and analysis to determine not just existing and imminent credit losses, but a robust forecast of expected future losses. For relatively homogeneous assets, such as mortgage pools, aggregated portfolio-level reviews may be fine. But for more idiosyncratic assets, like commercial or project finance loans, detailed loan-level forecasts and transparent analytics will become even more valuable to managers, investors and regulators operating under the CECL regime.

What does this mean for the future of loan portfolio accounting? There might be glimpses available from a surprising corner of the loan market: federal lending programs. Uncle Sam has participated in many lending markets since the Great Depression, including housing, small business, trade finance, and project finance. For over twenty years, each of these programs has been subject to a metric calculated by the White House’s Office of Management and Budget (“OMB”), known as the “credit subsidy cost.” The credit subsidy cost can be viewed as the expected value of default losses on a federally backed loan or portfolio, or synonymously a fair valuation of the government’s credit guarantee.

At a high level, our experience working with public sector clients has identified several additional types of information that have proven indispensable to developing a robust, auditable credit subsidy cost. These often include forecasts of macroeconomic and market-level data, such as house prices, interest rates, and unemployment and income levels. It may be appropriate to factor prospective regulatory changes, local conditions, or possible shocks into the analysis as well.

With the wreckage of the last financial crisis still looming large in the rear window, the shift to CECL represents an important step toward development of a 360-degree assessment of the landscape for holders of loan portfolios.  In future installments, we’ll turn to a more detailed discussion of how to not only take advantage of auditors’ newfound powers—like the Roman god, to look forward as well as backward—but to incorporate analytical methodologies and tools that, in effect, give night-vision goggles and heads-up displays to the modern-day corporate Januses surveying the financial world.

1Federal Reserve Bank of St. Louis, “Net Income for Commercial Banks in the United States.”