In this case study, we highlight how MIAC Analytics™ was used to model current expected credit losses for acquired distressed assets subject to SOP 03-3 on behalf of a bank lender in 2014.

MIAC prices loans by modeling the timing and severity of credit losses in accordance with industry best practices, including the audit and reconciliation of vast time series data. Under the FASB’s forthcoming standard CECL. similarly rigorous methodologies will be required for accounting of all loans. We developed DR-Surveillance™, MIAC CORE™, and Vision™ to implement and manage these requirements on behalf of the world’s largest financial institutions.

In 2014, “Bank ABC” hired MIAC to coordinate their “Statement of Provision” protocol for SOP 03-3, to model their current expected credit losses for the Bank’s residential portfolio, CRE and C&I portfolio, unsecured consumer, credit card and auto portfolios.

The workflow at ABC parallels our ongoing CECL preparation for lenders:

CECL MODEL - Flow of Process

First, Bank ABC sent us their historical loan monthly data files across each of the assets that were purchased as distressed assets at a discounted price. The MIAC Borrower Analytics (“BA”) team collected and audited the data with MIAC’s specific loan type data models through the DR Surveillance platform. The BA team worked with Bank ABC loan servicing staff to help them rectify data issues that were determined to be either missing or problematic.

For those loan data fields that were not correctable by the bank’s staff, MIAC sent a small loan due diligence team to Bank ABC site. MIAC’s Due Diligence Group used VeriFi™ to manage the loan due diligence project. The improved data quality and enriched data were tightly integrated with MIAC’s DataRaptor™ and DR-Surveillance™ solutions. This enabled the historical performance metrics to be optimally comprehensive and reliable.

The relevant historical collateral performance measurements such as voluntary prepayment, foreclosure entry, REO entry, transition roll rates, exit from FCL cure rates, and realized loss severities where calculated and reported using DR-Surveillance™.

Bank ABC’s results were then measured against the collateral performance of a larger population of similar loans from MIAC’s historical databases – a very meaningful and useful procedure.

Importantly, the institution-specific historical performance metrics were utilized whenever the client dataset was sufficient to be statistically reliable. Where Bank ABC’s historical data was insufficient, MIAC leveraged its national datasets which contains historical loan performance data on approximately 85 million loans from 1999 through current.

That process yields a detailed set of comparative results such as these:

CECL MODEL

Bank ABC’s historical performance experience and geographical focus were used to make assumptions and adjustments to MIAC’s CORE collateral loss forecasting models.

The asset specific CORE behavioral models are embedded in MIAC’s Vision™ asset/liability model. The macro-economic factors stressed in our models are unemployment, GDP, HPI, and CPI and these will be regionalized where appropriate and defensible. MIAC provided detailed explanations of our models, methodologies and data feeds, and special custom statistical annexes for the Bank’s auditor and regulatory review.

The current portfolio, starting with its current state of delinquency/FCL, was then simulated across multiple economic macro-factor scenarios, similar to DFAST/CCAR analysis. The forecasted cash flows simulated all the delinquency states and their transitions and the completion of REO and loss recovery with the highest level of granularity to arrive at Bank ABC’s current expected credit losses.

Leveraging the benefits of MIAC CORE™ through our cash flow forecast engine, Vision, allows us to generate loan-level granular loss forecasts. Industry best practices for CECL will require firms to perform cash flow forecasting for collateral assets segmented into buckets of similar collateral attributes blended with lender realized loss experience and in consideration of macro-economic scenarios. Expected future credit losses are included within the cash flow forecasts, which facilitates detailed analysis of the amount of expected losses, as well as when they are expected to occur.

As institutions focus on implementing solutions for CECL over the coming years, they will look to employ industry best practices for measuring current expected credit losses. These are the analytical methods that MIAC has been practicing with our clients for nearly 30 years.

“Bank ABC” CECL Allowance
MIAC Analytics™ Case Study, August, 2016

Download CECL Case Study

Jeffrey Zuckerman, Vice President, Capital Markets Group
jeffrey.zuckerman@miacanalytics.com  (212) 233-1250 x278

More about MIAC CECL Capabilities

Read MIAC’s Perspective: CECL – Current Expected Credit Loss: A CORE Competency?
Download MIAC’s FREE CECL Process Guide

 

CECL MODEL