CECL Overview, Square’s Bid To Be A Bank, Equifax Data Breach

By September 10, 2017Blog

Headlines centered on Equifax’s data breach that compromised the personal data of 143 million U.S. consumers. The breach heightens focus on cybersecurity and secure methods for exchanging loan data while protecting personally identifiable information.

Square is following in SoFi’s footsteps by applying for an Industrial Bank Charter. Square’s application would allow it to own a bank without having to comply with Bank Holding Company Act requirements and face supervision from the Federal Reserve.

In securitization news LendingClub filed ABS-15G for its CLUB 2017-P1 deal and Kroll assigned final ratings to three classes of notes issued by SoFi in its $527 Mn ABS deal, SCLP 2017-5.

This past Wednesday, the FASB released an update to the current expected credit losses methodology (CECL) for estimating credit loss allowances. The new accounting standard applies to any bank, credit union, financial institution, and non-bank lender that reports in US GAAP and hold loans or other instruments held at amortized cost.

CECL implementation will require increased supervision and controls around credit loss forecasting. This week we explain the impact of CECL and explain how banks and issuers can comply with CECL using the PeerIQ analytics platform.

CECL – Why It Matters for Issuers and Banks

Leading up to the Global Financial Crisis, investors in forward-looking financial markets were estimating extreme credit losses for banks. Yet, the higher loss expectations were not reflected in bank reporting due the prevailing accounting treatment. The obvious disconnect between the company value and accounting value highlighted the shortfalls in the US GAAP approach.

Under the current US GAAP methodology that regulators seek to remedy, companies do not recognize credit losses until it is “probable” that a loss has been incurred. Under the new CECL methodology, losses are incurred when they are expected, based on “historical experience, current conditions, and reasonable and supportable forecasts”. This regulatory objective is to provide greater transparency regarding credit quality trends to financial statement users.

This past Wednesday, FASB released an update to the current expected credit losses methodology (CECL) for estimating credit loss allowances. This new accounting standard, which was initially published in June 2016 (in conjunction with regulators such as the FDIC, OCC, and NCUA), will apply to financial assets carried at amortized cost, including loans held for investment and held-to-maturity debt. Once in place, these assets must be held on the balance sheet net of an expected loss account. Changes are effective for fiscal years beginning after Dec 15, 2019, for all for-profit companies that file with the SEC.

A Double-Edged Sword

Once firms adopt CECL, management will have increased discretion around forecasts and ultimately net asset carrying value. This represents a dichotomy for investors. Assets should be carried at more accurate levels and better reflect the organization’s financial position. However, management estimates will significantly affect the balance sheet and income statement. For CFOs and management, higher levels of discretion come with increased accountability and investor and regulatory requests for greater transparency.

While a complex model is not required by the FASB, trade associations including the American Bankers Association believe that detailed analysis may be needed to address common questions from board members, investors, or auditors. For example, if loss provisions decrease while delinquencies increase, stakeholders might want a greater understanding as to why.

CECL Methodology vs Current US GAAP

The major change with the CECL methodology is that organizations are expected to include forward looking information when determining credit losses. Banks will need to calculate expected credit loss at the loan level for the entire life of the loan and then aggregate with similar instruments.

Since ECL is calculated for the life of the financial asset, rather than the annual rate, almost all held-to-maturity instruments that are not risk-free will have a credit loss allowance. These long-dated assets may appear more volatile than financial statement users are accustomed to because their impairment has large implications for the balance sheet and income statement. Under the new regulation it will be more important to have correct, auditable, and explainable expected credit losses.

Source: PeerIQ, FASB, FDIC

As you can see in the table above, accounting for available-for-sale securities (AFS) will change slightly as well. Total impairment for AFS will be capped at the amortized cost less fair value of the security. This is because an AFS could be sold at the fair value before maturity in the event the market is pricing an asset differently than management. Additionally, changes in ECL for AFS will come through on the income statement rather than other comprehensive income (OCI).

Due to the nature of the CECL changes, executives should use similar methods to monitor risk internally as they use to forecast cash flows for quarterly reporting.

How PeerIQ Helps

PeerIQ’s offering enables issuers and banks of all sizes to comply with CECL using our risk management tools – all of which are built with loan-level resolution, broad market data, and modern data science. PeerIQ today is helping financial institutions manage data relevant to implement CECL and can recommend various loan-level estimation methods.

Additionally, PeerIQ enables financial institutions, investors, and regulators to speak a “common language” by providing a fully transparent forward-looking cashflow and pricing model.

For instance, PeerIQ’s analytical tools will enable risk managers to perform scenario analysis, and output a price yield table with CPR and CDRs to facilitate communication with the financial statement users.

How the PeerIQ Credit Model Works

The PeerIQ Credit Model uses a transition-rate model to estimate future cashflows. We start with the observation that at a given point in time, a loan has one of several states (current, delinquent, fully prepaid, or charged-off). Then we estimate the probability of a loan moving between the various states in a network. For example, a loan might move from “current” to “30-day delinquent”, or from “30-day delinquent” to “60-day delinquent”. Alternatively, a loan might “cure” and move from delinquent to current status.

 

Source: PeerIQ
Notes: C=Current, P=Prepay, D3=30day delinquent, D6=60day delinquent, D6+=90day and more delinquent, D=default.
Arrows indicate directionality of possible transitions.

We use these variables in order to estimate the probability of a loan transitioning between states, for each period, at a loan level in order to forecast future cash flows.

We estimate the probability of each transition using set of candidate variables for selection such as age, credit score, vintage, income, employment, term, grade, channel, and dozens of other variables. We also incorporate broader market data such as TransUnion dataset, ABS performance, and macro data such as unemployment rates to link issuer-specific dynamics to broader economic trends.

Conclusions & Takeaways on CECL

Overall, we applaud the coming changes to US GAAP and expect investors to respond favorably. Here are a few takeaways to keep in mind:

  • Organizations should be able to use similar methodologies for valuing securities at the financial reporting level as they are using for internal risk and portfolio management.
    • Subjective forecasting inputs can have a substantial influence on the income statement and balance sheet.
    • Changes in ECL could affect profitability, bond covenants, or even dividend payments.
  • Investors can trace changes in underwriting standards because assets are required to be separated by vintage.
    • ABS investors will see how public issues are valuing assets on their books with regards to expected losses.
  • Organizations are expected have models that forecast future interest rates, prepayment curves, default curves.
    • PeerIQ’s loan-level credit-model enables banks to comply with managing their consumer credit risk.
    • PeerIQ’s credit model provides issuers, community banks and large banks with an institutional grade approach to forecast losses.
  • Longer assets have higher burdens associated for holding them on the balance sheet.
    • The probability of a credit loss typically increases with the term of the loan.
    • Companies that intend to keep loans on their balance sheet may be more inclined to offer shorter term loans in order to reduce ECL allowances providing an opportunity to non-bank lenders who move loans off their balance sheet via securitization, whole loan sales, or marketplace lending (which would be accounted for as loans Held for Sale).

Feel free to reach out to PeerIQ to learn more about CECL implementation and PeerIQ’s forward looking credit models.

Conferences:

  • Ram will speak on the “Trends in Online Consumer Lending: Less Tech, More Fin?” panel on Monday, September 18 at ABS East in Miami, FL.
  • Ram will speak at the 1010Data Discovery Summit on October 5 in New York, discussing PeerIQ’s data partnership with 1010Data at 11:15-11:55AM.
PeerIQ in the News:
Industry Update:
  • Square Makes Its Big Move on Banking (PMYNTS, 9/7/2017) Square applied for an industrial loan company charter, leading to debate over whether fintech companies should be allowed to use this structure over traditional bank charters.
  • Cat-Bond Investors Sizing Up Irma (AB Alert, 9/8/17) Holders of catastrophe bonds already are handicapping which issues would face the biggest losses in the event Hurricane Irma takes a heavy toll on Florida.
Conferences:
  • 2017 US Open Stats & Fun Facts (WalletHub, 8/28/17) The US Open generates over $290 MM+ in annual revenue and is the most watched sporting event (over all matches) on the planet.