Get Started
Blog Archives

Applying Few-Shot Learning Techniques to Mortgage Data

Recorded: September 22nd | 1:00 p.m. EDT

Few-shot and one-shot learning models continue to gain traction in a growing number of industries – particularly those in which large training and testing samples are hard to come by. But what about mortgages? Is there a place for few-shot learning where datasets are seemingly so robust and plentiful? 

 

A growing body of evidence suggests that the high dimensionality of mortgage data spaces may actually make them ideal candidates for few-shot learning. 

 

Suhrud Dagli and Jing Liu present the latest installment in RiskSpan’s Data & Machine Learning Workshop series. Suhrud and Jing share examples of how they are using few-shot learning techniques in prepayment modeling and in automating quality control checks on uploaded mortgage data. 

 


Featured Speakers

Suhrud Dagli

Co-Founder and CIO, RiskSpan

Jing Liu

Senior Analyst, RiskSpan


EDGE: QM vs Non-QM Prepayments

Prepayment speeds for qualified mortgages (QM loans) have anecdotally been faster than non-QM loans. For various reasons, the data necessary to analyze interest rate incentive response has not been readily available for these categories of mortgages.

In order to facilitate the generation of traditional refinancing curves (S-curves) over the last year, we have normalized data to improve the differentiation of QM versus non-QM loans within non-agency securities.

Additionally, we isolated the population to remove prepay impact from loan balance and seasoning.

The analysis below was performed on securitized loans with 9 to 36 months of seasoning and an original balance between 200k and 500k. S-curves were generated for observation periods from January 2016 through July 2021.

Results are shown in the table and chart below.

 


 

For this analysis, refinance incentive was calculated as the difference between mortgage note rate and the 6-week lagged Freddie Mac primary mortgage market survey (PMMS) rate. Non-QM borrowers would not be able to easily refi into a conventional mortgage. We further analyzed the data by examining prepayments speeds for QM and non-QM loans at different level of SATO. SATO, the spread at origination, is calculated as the difference between mortgage note rate and the prevailing PMMS rate at time of loan’s origination.

 

Using empirical data maintained by RiskSpan, it can be seen the refinance response for QM loans remains significantly faster than Non-QM loans.

Using Edge, RiskSpan’s data analytics platform, we can examine any loan characteristic and generate S-curves, aging curves, and time series. If you are interested in performing historical analysis on securitized loan data, please contact us for a free demonstration.


RiskSpan Named to Inaugural STORM50 Ranking by Chartis Research – Winner of “A.I. Innovation in Capital Markets”

Chartis Research has named RiskSpan to its Inaugural “STORM50” Ranking of leading risk and analytics providers. The STORM report “focuses on the computational infrastructure and algorithmic efficiency of the vast array of technology tools used across the financial services industry” and identifies industry-leading vendors that excel in the delivery of Statistical Techniques, Optimization frameworks, and Risk Models of all types. 

RiskSpan’s flagship Edge Platform was a natural fit for the designation because of its positioning squarely at the nexus of statistical behavioral modeling (specifically around mortgage credit and prepayment risk) and functionality enabling users to optimize trading and asset management strategies.  Being named the winner of the “A.I. Innovation in Capital Markets” solutions category reflects the work of RiskSpan’s vibrant innovation lab, which includes researching and developing machine learning solutions to structured finance challenges. These solutions include mining a growing trove of alternative/unstructured data sources, anomaly detection in loan-level and other datasets, and natural language processing for constructing deal cash flow models from legal documents.

Learn more about the Edge Platform or contact us to discuss ways we might help you modernize and improve your mortgage and structured finance data and analytics challenges. 


Automating Compliance Risk Analytics

 Recorded: August 4th | 1:00 p.m. EDT

Completing the risk sections of Form PF, AIFMD, Open Protocol and other regulatory filings requires submitters to first compute an extensive battery of risk analytics, often across a wide spectrum of trading strategies and instrument types. This “pre-work” is both painstaking and prone to human error. Automating these upstream analytics greatly simplifies life downstream for those tasked with completing these filings.

RiskSpan’s Marty Kindler walks through a process for streamlining delta equivalent exposure, 10 year bond equivalent exposure, DV01/CS01, option greeks, stress scenario impacts and VaR in support not only of downstream regulatory filings but of an enhanced, overall risk management regime.


Featured Speaker

Martin Kindler

Managing Director, RiskSpan


RiskSpan VQI: Agency Mortgage Risk Layers for Q2 2021

RiskSpan’s Vintage Quality Index computes and aggregates the percentage of Agency originations each month with one or more “risk factors” (low-FICO, high DTI, high LTV, cash-out refi, investment properties, etc.). Months with relatively few originations characterized by these risk factors are associated with lower VQI ratings. As the historical chart above shows, the index maxed out (i.e., had an unusually high number of loans with risk factors) leading up to the 2008 crisis.

RiskSpan uses the index principally to fine-tune its in-house credit and prepayment models by accounting for shifts in loan composition by monthly cohort.

Rising Home Prices Contribute to More High-DTI Loans and Cash-out Refis

The Vintage Quality Index rose noticeably during the second quarter of 2021 — up to a value of 83.40, compared to 76.68 in the first quarter.

Unlike last quarter, when a precipitous drop in high-LTV loans effectively masked and counterbalanced more modest increases in the remaining risk metrics, this quarter’s sizeable VQI jump is attributable to a more across-the-board increase in risk layers.

A sharp rebound in the percentage of high-LTV loans, a metric that had been in steady decline since the middle of 2019, was accompanied by modest increases in borrowers with low credit scores (FICO below 660) and high debt-to-income ratios (greater than 45%).

The spike in home prices across the country that likely accounts for the rise in high-LTV mortgages also appears to be prompting an increasing number of borrowers to seek cash-out refinancings. More than 22 percent of originations had LTVs in excess of 80 percent at the end of Q2, compared to just 17 percent at the end of Q1. Similarly, nearly 25 percent of mortgages were cash-out refis in June, compared to 22 percent in March.

Modest declines were observed in the percentages of loans on investment and multi-unit properties. All other risk metrics were up for the quarter, as the plots below illustrate.

Population assumptions:

  • Monthly data for Fannie Mae and Freddie
  • Loans originated more than three months prior to issuance are excluded because the index is meant to reflect current market
  • Loans likely to have been originated through the HARP program, as identified by LTV, MI coverage percentage, and loan purpose, are also excluded. These loans do not represent credit availability in the market as they likely would not have been originated today but for the existence of

Data assumptions:

  • Freddie Mac data goes back to 12/2005. Fannie Mae only back to 12/2014.
  • Certain fields for Freddie Mac data were missing prior to 6/2008.

GSE historical loan performance data release in support of GSE Risk Transfer activities was used to help back-fill data where it was missing.

An outline of our approach to data imputation can be found in our VQI Blog Post from October 28, 2015.


Is Your Enterprise Risk Management Keeping Up with Recent Regulatory Changes?

Recorded: June 30th | 1:00 p.m. EDT

Nick Young, Head of RiskSpan’s Model Risk Management Practice, and his team of model validation analysts walk through the most important regulatory updates of the past 18 months from the Federal Reserve, OCC, and FDIC pertaining to enterprise risk management in general (and model risk management in particular).

Nick’s team present tips for ensuring that your policies and practices are keeping up with recent changes to AML and other regulatory requirements.


Featured Speakers

Nick Young

Head of Model Risk Management, RiskSpan


Data & Machine Learning Workshop Series

RiskSpan’s Edge Platform is supported by a dynamic team of professionals who live and breathe mortgage and structured finance data. They know firsthand the challenges this type of data presents and are always experimenting with new approaches for extracting maximum value from it.

Join us for a complimentary series of virtual workshops where RiskSpan professionals share what we’re learning about applying machine learning and other innovative techniques to data that asset managers, broker-dealers and mortgage bankers care about.

Catch up with these previously recorded workshops


Measuring and Visualizing Feature Impact & Machine Learning Model Materiality

RiskSpan CIO Suhrud Dagli demonstrates in greater detail how machine learning can be used in input data validations, to measure feature impact, and to visualize how multiple features interact with each other.

Structured Data Extraction from Images Using Google Document AI

RiskSpan Director Steven Sun shares a procedural approach to tackling the difficulties of efficiently extracting structured data from images, scanned documents, and handwritten documents using Google’s latest Document AI Solution.

Pattern Recognition in Time Series Data

Traders and investors rely on time series patterns generated by asset performance to inform and guide their trading and asset allocation decisions. Economists take advantage of analogous patterns in macroeconomic and market data to forecast recessions and other market events. But you need to be able to spot these patterns in order to use them.

Advanced Forecasting Using Hierarchical Models

Traditional statistical models apply a single set of coefficients by pooling a large dataset or for specific cohorts. Hierarchical models learn from feature behavior across dimensions or timeframes. This informative workshop applies hierarchical models to a variety of mortgage and structured finance use cases.

Quality Control with Anomaly Detection (Part I)

Outliers and anomalies refer to various types of occurrences in a time series. Spike of value, shift in level or volatility or a change in seasonal pattern are common examples.  RiskSpan Co-Founder & CIO Suhrud Dagli is joined by Martin Kindler, a market risk practitioner who has spent decades dealing with outliers.

Quality Control with Anomaly Detection (Part 2)

Suhrud Dagli presents Part 2 of this workshop, which dove into mortgage loan QC and introduce coding examples and approaches for avoiding false negatives using open-source Python algorithms in the Anomaly Detection Toolkit (ADTK).


In ESG Policy, ‘E’ Should Not Come at the Expense of ‘S’

ESG—it is the hottest topic in our space. No conference or webinar is complete without a panel touting the latest ESG bond or the latest advance in reporting and certification. What a lot of these pieces neglect to address is the complicated relationship between the “E” and the “S” of ESG. In particular, that climate-risk exposed properties are also often properties in underserved communities, providing much-needed affordable housing to the country.

Last week, the White House issued an Executive Order of Climate-Related Financial Risk. The focus of the order was to direct government agencies toward both disclosure and mitigation of climate-related financial risk. The order reinforces the already relentless focus on ESG initiatives within our industry. The order specifically calls on the USDA, HUD, and the VA to ‘consider approaches to better integrate climate-related financial risk into underwriting standards, loan terms and conditions, and asset management and servicing procedures, as related to their Federal lending policies and programs.” Changes here will likely presage changes by the GSEs.

In mortgage finance, some of the key considerations related to disclosure and mitigation are as follows:

Disclosure of Climate-Related Financial Risk:

  • Homes exposed to increasing occurrence to natural hazards due to climate changes.
  • Homes exposed to the risk of decreasing home prices due to climate change, because of either increasing property insurance costs (or un-insurability) or localized transition risks of industry-exposed areas (e.g., Houston to the oil and gas industry).

Mitigation of Climate-Related Financial Risk:

  • Reducing the housing industry’s contribution to greenhouse gas emissions in alignment with the president’s goal of a net-zero emissions economy by 2050. For example, loan programs that support retrofitting existing housing stock to reduce energy consumption.
  • Considering a building location’s exposure to climate-related physical risk. Directing investment away for areas exposed to the increasing frequency and severity of natural disasters.

But products and programs that aim to support the goal of increased disclosure and mitigation of climate-related financial risk can create situations in which underserved communities disproportionately bear the costs of our nation’s pivot toward climate resiliency. The table below connects the FEMA’s National Risk Index data to HUD’s list of census tracts that qualify for low-income housing tax credits, which HUD defines as tracts that have ‘50 percent of households with incomes below 60 percent of the Area Median Gross Income (AMGI) or have a poverty rate of 25 percent or more.’ Census tracts with the highest risk of annual loss from natural disaster events are disproportionally made of HUD’s Qualified Tracts.

As an industry, it’s important to remember that actions taken to mitigate exposure to increasing climate-related events will always have a cost to someone. These costs could be in the form of increased insurance premiums, decreasing home prices, or even loss of affordable housing options altogether. All this is not to say that action should not be taken, only that balancing social ESG goals should also be considered when ambitious environmental ESG goals come at their expense.

The White House identified this issue right at the top of the order by indicating that any action on the order would need to account for ‘disparate impacts on disadvantaged communities and communities of color.’

“It is therefore the policy of my Administration to advance consistent, clear, intelligible, comparable, and accurate disclosure of climate-related financial risk (consistent with Executive Order 13707 of September 15, 2015 (Using Behavioral Science Insights to Better Serve the American People), including both physical and transition risks; act to mitigate that risk and its drivers, while accounting for and addressing disparate impacts on disadvantaged communities and communities of color (consistent with Executive Order 13985 of January 20, 2021 (Advancing Racial Equity and Support for Underserved Communities Through the Federal Government)) and spurring the creation of well-paying jobs; and achieve our target of a net-zero emissions economy by no later than 2050.”

The social impacts of any environmental initiative need to be considered. Steps should be taken to avoid having the cost of changes to underwriting processes and credit policies be disproportionately borne by underserved and vulnerable communities. To this end, a balanced ESG policy will ultimately require input from stakeholders across the mortgage industry.


Mortgage DQs by MSA: Non-Agency Performance Chart of the Month

This month we take a closer look at geographical differences in loan performance in the non-agency space. The chart below looks at the 60+ DPD Rate for the 5 Best and Worst performing MSAs (and the overall average). A couple of things to note:

  • The pandemic seems to have simply amplified performance differences that were already apparent pre-covid. The worst performing MSAs were showing mostly above-average delinquency rates before last year’s disruption.
  • Florida was especially hard-hit. Three of the five worst-performing MSAs are in Florida. Not surprisingly, these MSAs rely heavily on the tourism industry.
  • New York jumped from being about average to being one of the worst-performing MSAs in the wake of the pandemic. This is not surprising considering how seriously the city bore the pandemic’s brunt.
  • Tech hubs show strong performance. All our best performers are strong in the Tech industry—Austin’s the new Bay Area, right?

Contact Us


May 26 Webinar: Is Your Pricing Methodology Compliant With Rule 2a-5?

Recorded: May 26th | 1:00 p.m. ET

The SEC’s new Rule 2a-5 has important ramifications for anyone in the business of pricing hard-to-value instruments. It requires valuation practitioners to demonstrate good faith in implementing and following a defensible and transparent processes. But what does this mean as a practical matter?

On Wednesday, May 26th experts David Baum and Martin Dozier of Alston & Bird and Bill Moretti and Joe Sturtevant of RiskSpan explained and responded to your questions about:

  • What the new requirements are
  • Who is impacted and when
  • Implementation best practices
  • Potential issues with Rule 17a-7, and
  • Modeling considerations, including assumptions, back-testing, calibration, and data management.

Featured Speakers

William Moretti

Senior Managing Director, RiskSpan

David Baum

Partner, Investment Management, Trading and Markets Group, Alston & Bird LLP

Martin Dozier

Partner, Alston & Bird LLP

Joseph Sturtevant

Head of Valuation Services, RiskSpan


Get Started