Systemic and Operational Analytic Models for Superior Insight
Digital Risk introduces Veritas™, the industry’s only multi-dimensional mortgage analytic platform that models systemic and operational risk for insight superior to single-scoring methodology.
Veritas is built on a vast repository of specific borrower, property and local real estate market data which provides wide-spectrum insight into the many facets of risk. The result is the most reliable method for understanding risk at the root cause whether systemic or operational.
The time has come for a platform that analyzes many more impactful variables that can predict future risk with much more accuracy and provide the reason why that risk exists. Veritas is that platform, bringing data and analytic confidence to the entire mortgage lifecycle, including origination, default and servicing.
The Many Dimensions of Borrowers. One Size Fits None
The mortgage meltdown turned traditional thinking on its head with regard to what constitutes risk. However, most industry players continue to rely on outdated methods to determine risk. If history taught us anything it’s that a high FICO and low CLTV ratios don’t guarantee low risk as demonstrated by the large swath of high FICO borrowers that defaulted in record numbers over the last four years.
Even today the industry still depends heavily on that one static score despite its inability to take into account the ever-changing life factors that differentiate borrowers, influence operational risk and are essential to predicting a borrower’s ability to repay their loan in the future.
The traditional credit score is relatively static with respect to time and other factors, giving the impression of “one borrower, one score” regardless of the many factors that discriminate amongst borrowers and in turn, influence risk. Veritas takes into account the ever-changing factors of the borrower’s circumstance, that previously have not been measured, and incorporates them into a more accurate, meaningful and dynamic view of the borrower.
32 Unique & Discreet Borrower Dimensions
One of the most distinctive capabilities of Veritas is that it can meaningfully discriminate amongst the many different kinds of mortgage borrowers across a host of 32 distinct dimensions, or clusters, enabling the analyst, the underwriter, the servicer or the investor to make decisions previously thought impossible.
For example, Veritas can identify high CLTV borrowers who are likely to service their mortgage, versus those who are likely to default. In the same paradigm, it also determines how CLTV will impact a borrower’s likelihood of redefaulting upon a loan modification and which high CLTV borrowers will be responsive to principal write down. For example, it can discriminate between those who will keep their loan current after principal write down and those who will not, allowing their loan to redefault even after principal write down.
Veritas can identify borrowers most likely to default or redefault, and provides comprehensive analysis as to “why” such an event would occur.
The Industry Data Repository
The Veritas repository contains data from more than 5 million loans originated from 2006 – 2011. This database represents information from industry standard loan performance repositories, Digital Risk’s own proprietary database and a public record data source that aggregates public data.
All information is fully anonymized. Specific borrowers and properties can’t be identified, nor can individual non-public, personal information, such as credit histories, be examined.
- Borrower database include: the number of relationships, types of credit relationships, how encumbered the borrower is by all forms of credit, how he/she services the credit relationships, the true monthly debt servicing obligations and how the owner likely responds in distress.
- Property database includes many parameters property type, age, structure, equity (related to CLTV), value across multi-year period, and encumbrance level.
Real Estate Market Data Include:
- How housing prices have changed across 6,000 zip codes
- How those price changes have evolved over time
- What are the characteristics and conditions of the market
- What types of properties are selling
- What is the average time on market
Absolute Reliability. Regression Testing Proves the Hypothesis
Digital Risk, along with two banks as development partners, has performed a number of analytic stress tests over the past year to ensure accuracy among the models. To test the answers Veritas provided against true outcomes, a sample of 100,000 modified loans were analyzed to determine likelihood to default. The Veritas results were in the 95th percentile successfully identifying those loans that had, in fact, redefaulted or were highly likely to redefault in the near future. This allowed our partners to take preventive measures to keep the borrower in their homes.
How is Veritas Different?
Analysis of the data demonstrates that current one-dimensional credit score is not as meaningful without the discreet, multi-dimensional borrower models and methodology built into the proprietary platform. While a single static score can display known borrower behavior, it does not take into account the ever-changing life factors that discriminate among borrowers, influence operational risk and are required to understand how those factors may affect the future.
Veritas undoes many of the foundational beliefs about the loan lifecycle, from origination to servicing and even pricing the loan for sale into the secondary market. Veritas is the first multi-dimensional engine of its kind to blend an assessment of systemic risk with an index of operational risk, achieving an ideal blend of people, processes and technology to create a more complete picture of loan risk.
Veritas allows the mortgage industry to calculate future risk using a multi-dimensional model to instill data confidence. Analytic and risk modeling services are provided as part of Digital Risk’s mortgage risk, compliance and transaction management solutions.