CECL is Here, But Where is EC?

Current Expected Credit Loss (CECL) is here but it appears that Economic Capital (EC) has been left behind.

CECL is the new mandate of the Financial Accounting Standards Board (FASB) that seeks to ensure the accounting for credit losses accurately reflects the inherent losses of credit-sensitive assets held by public companies. CECL replaces today’s “incurred loss standard”, and is required as of December 15, 2019 for any publicly-traded institutions issuing credit, including banks, savings institutions, credit unions and holding companies filing under GAAP accounting standards. CECL will bring major changes to these firms, not least in the way they estimate losses for their allowance for loan and lease losses (ALLL).

The essential analytical feature of CECL is that these debt investors must estimate the future expected loss on each credit investment due to interruptions of that instrument’s expected cash flows. As seen in the present value of a debt obligation, that obligation is worth its contractual or par value when there are no interruptions in the repayments by the borrower and future cash flows are discounted at the coupon on interest rate of the debt obligation. If those cash flows are interrupted, the present value falls below the par value. The difference between estimated value with interruption and par value then is the expected loss at the heart of CECL.

Those interruptions of cash flows come from default of the borrower and subsequent loss on the asset. Although the FASB does not mandate a specific methodology for estimating the CECL, the basic analytical challenge of this new standard is the estimation of each borrower’s probability of default (PD) and each asset’s loss given default (LGD). The mathematical product of PD times LGD is expected loss (EL), precisely the concept that is required for CECL.

CECL is new, but we’ve seen these concepts before. In Economic Capital.

The Idea of Economic Capital

Credit Economic Capital (EC) is based on the concept of unexpected loss (UL). UL is the variance of the statistical distribution of losses in a portfolio of loans and / or bonds. This distribution is skewed to the right with a long, “fat” tail because of the asymmetric nature of PD and its variance and of LGD and its variance (see the stylized figure of a credit loss distribution below). EC typically is chosen from a high percentile of cumulative losses corresponding to a low frequency-high impact event (e.g., 99.95th percentile corresponding to a loss event that occur five times in every 10,000 years).

EL is related to EC because EL, at the portfolio level, is the mean of this same distribution. Mathematically, portfolio EL also is the sum of the ELs of individual assets and a portion of the EC of the portfolio can be attributed to each asset in the portfolio.

Importantly, both EL and EC are derived from the PDs and LGDs of these assets.

Traditionally, EL and EC have been calculated over a one-year perspective since PD usually is estimated on a one-year basis. CECL requires EL to be calculated over the remaining life of each asset, meaning, in many cases, more than one year. While estimating remaining-life PD is not easy, the loss distribution of credit losses from which EL and EC would be measured is not fundamentally different from the figure above.

Banks and other financial institutions are scrambling to comply with CECL given its complexity and the short time left in which to comply. However, some of the easiest answers to estimating EL exist already their systems for economic capital, such as PD and LGD. Assuming that they still have such systems…

The Rise and Fall of Economic Capital

In the late 1980’s and 1990’s, Economic Capital was all the rage in bank risk management. Building off of Bankers Trust’s innovative work in risk-adjusted return on capital (RAROC), many large banks began to build comparable systems, typically on their commercial loan portfolios. I cut my teeth in risk management by building default prediction models for commercial loans at Citibank in the 1980s, as well as creating analytics for loss given default, both of which are inputs to the creation of the loss distribution from which EL and EC are drawn.

The fundamental purpose of EC was to enhance risk management at banks by redefining risk as loss volatility and to measure that volatility for all sources of potential losses. Over time, EC systems at banks became progressively sophisticated, including integrating all sources of potential loss into an institution-wide measure of loss volatility to correspond to the target debt rating of the bank. By attempting to capture all sources of risk in a single metric, EC was one of the building blocks of what is now known as Enterprise Risk Management.

For a brief period of time, EC had other applications. It was the denominator in RAROC as well as the basis for ranking of risk by customer, product type and business unit. At the end of 1999, I managed the first balance sheet collateralized loan obligation, issued by BNP Paribas, for which the goal was to transfer economic risk to the market through optimizing economic capital in the assets not securitized.

In the 2000’s, EC reached its pinnacle of importance, extending beyond banks to other institutions such as insurance companies. Many banks used it, and some still do, as the metric by which to allocate regulatory capital to business units and other organization entities and some use it as part of their risk appetite framework. Since then, the limitations of EC as a management tool became clear as its analytical complexity and lack of direct relevance to daily business activities made EC more of an academic curiosity. That lack of relevance was made obvious as EC failed at many banking institutions in the sub-prime mortgage crisis, precisely the kind of tail event for which EC was designed.

In a range a practices survey conducted on economic capital at banks and published in 2009, the Basel Committee on Banking Supervision anticipated one of the reasons why EC has failed to take hold in U.S. commercial banks – lack of commitment to EC by executive management[1]:

”The viability and usefulness of a bank’s economic capital processes depend critically on the existence of a credible commitment or ‘buy-in’ on the part of senior management to the process. In order for this to occur, it is necessary for senior management to recognise the importance of using economic capital measures in conducting the bank’s business…Moreover, senior management needs to take measures to help ensure the meaningfulness and integrity of economic capital measures. It should also seek to ensure that the measures comprehensively capture all risks and implicit and/or explicit management actions embedded in measurement processes are both realistic and actionable.”

I will write further on the reasons why EC has lost its relevance at many banking institutions in an upcoming blog but, for present purposes, we can say that EC has not lived up to its potential. It remains in place at many large banks but it no longer has the important role it once did. Arguably, it has been supplanted by “stress testing”, such as the Dodd-Frank Act Stress Tests (DFAST) and the Comprehensive Capital Analysis and Review (CCAR). Regulatory stress tests, widely understood to capture events in the tail of a firm’s loss distribution (see figure below), seem more intuitive that economic capital, in part because each stress test has its own macroeconomic scenario and thus are more “visible” to the participants in the process.

Can Economic Capital Make a Comeback?

CECL will be an enormous challenge for most institutions for numerous reasons, though banks that use PD and LGD in their credit risk management systems at least have the conceptual framework and systems to use these metrics in the quantification of the EL in CECL. Rather than starting from scratch, they can expand the use of the PD and LGD to CECL.

In such cases, EC could get a second chance, as well. Since CECL calls for EL to be ingrained at all public institutional investors in debt, there is a possibility that the concepts of PD and LGD will become second nature in many of those firms, including some that have had no prior experience with them. Since portfolio-level EC is a close cousin to portfolio-level EL, it is possible, perhaps likely, that CECL will make those institutions more open to proper quantification on EC and proper use of EC (more on these subjects later).

CECL may also be an impetus to use EL beyond loan loss provisions, especially in creating incentives for lenders to earn sufficient profits to “pay” for the increase in provisions that will naturally come with CECL. Those institutions also may be open—again—to using RAROC as a measure of profitability since EL is a part of the numerator of this metric, EC is the denominator, and CECL will be an important driver to the proper quantification of both.

EC and RAROC thus are two of the potential derivative benefits of using PD and LGD in compliance with CECL.

I welcome your thoughts and comments.

[1] Basel Committee on Banking Supervision. 2009. Range of practices and issues in economic capital frameworks. Bank for International Settlements.