Linkedin    Twitter   Facebook

Get Started
Log In

Linkedin

Blog Archives
Home / Blog Archives

May 19 Workshop: Quality Control Using Anomaly Detection (Part 2)

Recorded: May 19 | 1:00 p.m. ET

Last month, RiskSpan’s Suhrud Dagli and Martin Kindler outlined the principles underlying anomaly detection and its QC applications related to market data and market risk. You can view a recording of that workshop here.

On Wednesday, May 19th, Suhrud presented 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).

RiskSpan presents various types of detectors, including extreme studentized deviate (ESD), level shift, local outliers, seasonal detectors, and volatility shift in the context of identifying spike anomalies and other inconsistencies in mortgage data. Specifically:

  • Coding examples for effective principal component analysis (PCA) loan data QC
  • Use cases around loan performance and entity correction, and
  • Novelty detection

Suhrud Dagli

Co-founder and CIO, RiskSpan

Martin Kindler

Managing Director, RiskSpan


April 28 Workshop: Anomaly Detection

Recorded: April 28 | 1:00 p.m. ET

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. Anomaly detection depends on specific context. 

In this month’s installment in our Data and Machine Learning Workshop Series, RiskSpan Co-Founder & CIO Suhrud Dagli is joined by Martin Kindler, a market risk practitioner who has spent decades dealing with outliers.

Suhrud and Martin explore unsupervised approaches for detecting anomalies.

Suhrud Dagli

Co-founder and CIO, RiskSpan

Martin Kindler

Managing Director, RiskSpan


April 21 Webinar: Automated Prepayment Model Calibration Using Machine Learning

Recorded: April 21 | 1:00 p.m. ET

Manually tuning MBS prepayment models is messy. In what amounts to an elaborate trial-and-error exercise, modelers must frequently resort to subjectively selecting sub-populations to calibrate, running back-testing to see where and how the model is off, and then tweaking knobs and re-running the back-test to see the impacts. Rinse and repeat.

RiskSpan’s Janet Jozwik and Steven Sun present an approach for running a set of back-tests on MBS pools that automatically solves for the right set of tuners to align model results to actuals. Learn how, by automatically covering every feasible combination of model knobs possible, you can visualize for every pool the impact each knob combination has on:

  • Modeled prepay vs. actuals
  • Model error
  • Refi incentive and other pool features

Janet Jozwik

Managing Director, RiskSpan

Steven Sun

Director, RiskSpan


A Mentor’s Advice: Work Hard on Things You Can Control; Learn to Live with Things You Cannot

March is Women’s History Month and RiskSpan is marking the occasion by sharing a short series of posts featuring advice from women leaders in our industry.

Laurie-Goodman

Today’s contributor is Dr. Laurie Goodman, vice president at the Urban Institute and codirector of Urban’s Housing Finance Policy Center. Laurie helped break barriers as one of the first women to work on Wall Street and built her own brand as a go-to researcher for the housing and mortgage industry.

Laurie serves on the board of directors of MFA Financial, Arch Capital Group, Home Point Capital and DBRS. In 2009, she was inducted into the Fixed Income Analysts Hall of Fame following a series of successful research leadership and portfolio management positions at several Wall Street firms.

Laurie offers this guidance to young women (though it is applicable to everyone):

#1 – Figure out the balance that works for you between your personal life and your work life, realizing that you can’t be all things to all people all the time. There are times when you will spend more time on your work life and times when you will spend more time on your home life and other non-work related activities. You can’t be a super-performer at both all the time. Don’t beat yourself up for that part of your life where you feel you are underperforming.

#2 – Develop a thick skin and don’t take things personally. This will make you a much better colleague. Many times, colleagues and others in your organization make comments that can be interpreted either as personal affronts or general statements on the project. Always look for the non-personal interpretation (even if you suspect it is personal). For example, “Gee, these results aren’t very useful” can be interpreted personally as “It’s your fault — if you had done it differently it would have been better” or non-personally, as in “The material just didn’t give us any new insights.” Assume it was meant non-personally.  

#3 – Develop confidence and advocate for yourself. Speak up in meetings, particularly if you have points to add, or can steer the conversation back on track. If you are not feeling confident, fake it until you realize that you have as much (or more) to contribute than anyone else. And use that confidence to advocate for yourself — your success is more important to you than it is to anyone else. Have the confidence to own your mistakes; we all make mistakes. if you own them, you will do everything you can to correct them.

We also asked Laurie what, if anything, she might have done differently. Her response:

Early in my career, when things went off track for any reason, I got very frustrated. I was unable or unwilling to distinguish between those aspects of my work that were under my control, and those aspects of my work environment that I could not control. As a result, in the early years of my career, I changed jobs frequently. As the years have gone on, I have learned to do the best work I can on issues that are under my control and accept and live with what is not. It has made my work life much more enjoyable and productive.

Our thanks to Laurie for her valuable perspective!

Keep an eye on https://riskspan.com/insights/ throughout March for insights from other women we admire in mortgage and structured finance.

RiskSpan is proud to sponsor POWER OF VOICE BENEFIT. Girls Leadership teaches girls to exercise the power of their voice. #powerofvoice2021

A Mentor’s Advice: Go Where Your Heart Leads and Learn to Say Yes (and No)

March is Women’s History Month and RiskSpan is marking the occasion by sharing a short series of posts featuring advice from women leaders in our industry. Amy Cutts

Today’s contributor is Dr. Amy Crews Cutts, President and Chief Economist of AC Cutts and Associates, an economics and strategy consulting firm based in Reston, VA. She started her professional career as academic and she used that experience to build her network, soon landing at Freddie Mac. There she honed her professional skills and reputation as an economist, writer, and speaker. Amy was engaged by Equifax in 2011 to create the office of the chief economist. She has been recruited to serve on corporate and nonprofit advisory boards and elected to serve on boards of directors of leading economics associations. Amy has become an internationally recognized expert on consumer credit and economic policy and is a sought-after speaker and advisor. She is a participant in the Wall Street Journal Survey of Leading Economics, and her work has been cited in federal regulations and in cases before the U.S. Supreme Court.

Amy offers this guidance to women embarking on their careers (though it is applicable to everyone): 

#1 – Look deep for your talents and passion. We tend to think of jobs as titles rather than the small things that make up the role. The best day of my career came when I embraced the joy in the small parts of the job, and from that I was able to move mountains within the company and create the role that suited me.  

#2 – Build your networks always. When invited to join others for lunch, go! There are always deadlines, but this is just as important. You never know when you will learn a critically valuable piece of information from a casual conversation. When given the opportunity to present (internally or externally), take it, and know that they respect you enough to have asked and they care about what you have to say. Even a small opportunity may be the start of something big, so jump in with both feet.   

#3 – Speak up and, when you have a strong opinion, add your voice to the discussion (but never, ever, make it personal). In a corporate setting, you get one chance to speak critically of a plan, a policy, or problem. After that you need to move on because you were heard, even if they chose to go against your advice. Good counsel is valuable in every organization. Be someone others want to get advice from. 

We also asked Amy what, if anything, she might have done differently. Her response: 

My biggest regrets have come from not being openminded enough. I didn’t know at the time that the days of secretaries were numbered, but I rejected the suggestion of taking a typing class in high school because I had bigger plans (who knew we would spend our days typing?). I never learned how to code well because, at the time I was in college, computer science courses were mainly for mainframe applications (unfortunately, I did take APL programming, an already dying language in the 1980s). I have rejected job opportunities because I did not fit 100 percent of the job description but would later see someone much less capable in that role, with the prestige or promotion I should have tried for.

Our thanks to Amy for her valuable perspective!

Keep an eye on https://riskspan.com/insights/ this month for insights from other women we admire in mortgage and structured finance.

RiskSpan is proud to sponsor POWER OF VOICE BENEFIT. Girls Leadership teaches girls to exercise the power of their voice. #powerofvoice2021

March 31 Workshop: Advanced Forecasting Using Hierarchical Models

Recorded: March 31 | 1:00 p.m. ET

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.

Suhrud Dagli and Jing Liu host an informative workshop applying hierarchical models to a variety of mortgage and structured finance use cases, including:

  • Changes in beta and covariance of portfolios across time
  • Loan performance across geographies and history – e.g., combining credit performance data from 2008 with unemployment-driven credit issues in 2020.
  • Issuer-level prepayment performance

Suhrud Dagli

Co-founder and Chief Innovation Officer, RiskSpan

Jing Liu

Model Developer, RiskSpan


A Mentor’s Advice: Raise Your Hand, Be Inquisitive, and Find Your Niche

March is Women’s History Month and RiskSpan is marking the occasion by sharing a short series of posts featuring advice from women leaders in our industry. Faith Schwartz

Today’s contributor is Faith Schwartz, President of Housing Finance Strategies, a strategic advisory services firm. She achieved a significant industry renown for having developed and led the HOPE NOW Alliance to unite the industry throughout the housing crisis. Faith was named among the “Most Powerful Women in Mortgage Banking” by National Mortgage Professional Magazine in 2018. She is also a HousingWire Vanguard Award winner.

Faith offers three core pieces of advice to young women (though it is applicable to everyone):

#1 – The business world is complex and there is much to learn.  Raise your hand often, be inquisitive, and always seek to understand the economics of every last detail of the business.

#2 Whatever your line of work, develop your unique expertise, develop a way to measure and report it, and share it often with those who are less knowledgeable. You will become a “go-to” resource for your colleagues.

#3 Establish yourself as an inclusive leader. So many people forget it is your peers, staff and leaders who make up the ecosystem of your company. Listen to many ideas and then come to your own conclusion. This will help you downstream as you lead new and exciting initiatives.

We also asked Faith what, if anything, she might have done differently. Her response:

My lessons learned are many: Be a better listener; think about how to most effectively communicate internally and externally – and to know the  difference. Understand the full corporate culture where you work to best adapt your own style and stay effective. Over my career, I have learned how to adapt, how to evolve in a matrixed leadership role and how to continue to be an influencer, regardless of title. To this day, I much prefer the title of “senior advisor” in most any role I play.

Our thanks to Faith for her valuable perspective!

Keep an eye on https://riskspan.com/insights/ this month for insights from other women we admire in mortgage and structured finance.

RiskSpan is proud to sponsor POWER OF VOICE BENEFIT. Girls Leadership teaches girls to exercise the power of their voice. #powerofvoice2021

Edge Enhancements: Spotlight AGENCY EDGE

RMTA2021 Winner2021 is off to a great start, but the Edge Team is not resting on its laurels.

On the heels of a year that saw more than a 30 percent increase in Edge subscribers, including a doubling of Agency Module users, we continue to add more of the Ginnie and GSE data you need.

Edge’s enhanced datasets make customizing S-curves even easier.

For example:

Loans with a principal deferral pay more slowly than loans without them when faced with the same refinancing incentive.

But how much more slowly?

Edge lets you quantify the difference, so you can adjust your models accordingly.

7 of the 10 largest U.S. broker/dealers use Edge to analyze Agency prepays.
Find out why.

AICPA

What Will Non-QM Underwriting Look Like After the Pandemic?

On January 21st, RiskSpan senior managing director Bill Moretti moderated a panel at IMN’s Non-QM Virtual Forum.

The discussion, entitled “Underwriting Credit Standards, Assessing Ability to Pay & Evaluating Default Risk: Are You Protecting Yourself Against a Second Wave or Going All Out?” featured a cross-section of industry participants. These included a rating agency (DBRS Morningstar), a wholesale originator (Oaktree Funding Corp.), an aggregator/securitizer (Annaly Capital Management) and two technology companies (RiskSpan and LoanScorecard), each of whom obviously approaches the underlying question from a slightly different perspective.    

Underwriting standards, of course, are best explored in the context of evaluating credit risk. Unprecedented market disruption in response to Covid clearly laid bare the fact that many (if not most) mortgage borrowers are not in a financial position to endure a significant curtailment to income for any sustained period. The discussion focused on what this means for new-origination underwriting standards after Covid.   

The panel tackled a number of questions, including: 

  • How is the market assessing risks in new loans given the current market conditions?
  • How are income declines considered? 
  • What role does forbearance play? 
  • Should underwriters be looking harder at borrower net worth or taking other assets into account? 
  • Should PPP payments be counted as borrower income? 
  • Should DSCR calculations be revisited or modified for investment-property borrowers? 
  • Is the market taking alternative and other non-traditional data into account when assessing credit risk? 

A review of non-QM performance in 2020 relative to that of prime and GSE (CRT) loans revealed some interesting insights: 

  • Non-QM loans reached a significantly higher 30-day DQ peak (13% in April and May) compared to 4% and 3% for prime and CRT loans, respectively. An analogous gap was observed prior to Covid, however. 
  • 60+ day DQ (including BK/FC/REO) showed something similar with non-QM loans peaking at 14% in June/July compared with 4% and 5% for prime and CRT loans, respectively. These rates improved by the end of the year but remain elevated relative to pre-Covid levels. The panelists attribute this to forbearance activity.  
  • CDRs began trending upward in July and August, reaching 10 bps across all classes. CRT CDRs spiked to 50 bps in September and October before retreating to 40 bps, while prime and non-QM CDRs ramped up to the 30 bps and 40 bps, respectively, by year-end. 
  • CPRs among non-QM loans held steady throughout the year at around 25%. Prime and CRT loans experienced a steady increase in CPR, rising from 20%-25% earlier in the year to approximately 45% by the end of the year. Record refinancing activity appeared to drive much of this. 

Partially in response to Covid, rating agencies have implemented changes in their requirements for structural features, including the following: 

  • Waterfall changed from pro–rata allocation to full sequential pay 
  • Triggers eliminated based on DQ, loss and credit enhancement 
  • DQ P&I advancing assumptions reduced from 4-6 months to 0-3 months 
  • Credit enhancement increased, especially among lower–rated tranches — excess spread and reserve account requirements are now required 
  • Rating scenario analyses and stress assumptions increased to reflect Covid forbearance assumptions 

The pandemic created a significant disruption for non-QM origination and acquisition. Some market participants struggled to adapt during the pandemic’s early months as funding and margin calls impacted mortgage buyers.  Aggregators who maintained strong relationships with their originators had more success maintaining funding commitments. These relationships were critical to maintaining overall market health and liquidity. An ability to adapt, allow forbearance and modifications, and work with borrowers was viewed as equally important.  

How is underwriting likely to evolve? 

The panelists agreed that a revamped underwriting process must consider different sources of income as well as borrower assets and reserves. 

Additional borrower requirements that have been proposed include the following: 

  • Consistency of income in the borrower’s work industry 
  • Additional weight accorded to other assets held by a borrower 
  • Consideration of PPP loans as borrower assets but not income 
  • Stricter underwriting for DSCR loans including a required reserve 
  • Additional scrutiny of loans made to foreign nationals 

The pandemic had an outsized effect on non–QM origination volume, which continues to experience headwinds from Agency-eligible production. There’s just no getting around the fact that, all else equal, brokers find it more profitable (not to mention easier) to focus on Agency production. The importance of specialization continues to be felt, as non–QM aggregators tend to focus more of their attention and efforts on “pure” Non–QM origination shops, as opposed to full-service mortgage bankers, which originate a mix of Agency and non-Agency mortgages. Non-QM underwriting standards will likely need to take this reality into account. 

What role will technology play? 

While not yet as ubiquitous as in Agency lending, front-end automated underwriting systems continue to make strides in the non-QM world. This growth in AUS consistency and efficiency is a critical component to creating a digital environment for mortgages and accelerating the approval process while maintaining strong risk management and compliance. 

The industry is crying out for a clean end-to-end loan acquisition solution for aggregators and other whole–loan portfolio investors. Investors are increasingly looking to get into whole loans, but the secondary whole–loan acquisition process is extremely demanding from an operational perspective. RiskSpan’s Edge Platform enables residential whole–loan buyers to outsource many of these functions. 

Beyond whole loans, non-QM securitization data and analytics continues to be a source of angst, due in part to inconsistent forbearance and modification reporting. RiskSpan is seeking to alleviate these pain points by working with clients to standardize and normalize reporting inconsistencies, particularly in the non-QM space. The goal is to provide a way for investors and other market participants to benchmark deals against one another, confident that delinquency, forbearance, and modification comparisons are truly apples-to-apples.  

Finally, the panel discussed industry efforts to improve clarity around mapping (or bucketizing) loan types. Doing this is challenging in the non–QM sector because there are so many (literally hundreds) different types of loan documentation requirements. Understanding these is vital to modeling credit risk. Mapping time series data based on the loan type is hard, RiskSpan is at the forefront of developing methodologies to speed and simplify analysis by logically mapping many different loan types into fewer buckets. 

Contact us for a free demo and to discuss how RiskSpan can combine its powerful Edge Platform with expert services to help you tackle your thorniest underwriting data and modeling challenges. 

RiskSpan’s Edge Platform Wins 2021 Buy-Side Market Risk Management Product of the Year

RiskSpan, a leading SaaS provider of risk management, data and analytics has been awarded Buy-Side Market Risk Management Product of the Year for its Edge Platform at Risk.net’s 2021 Risk Markets Technology Awards. The honor marks Edge’s second major industry award in 2021, having also been named the winner of Chartis Research’s Risk-as-a-Service category.

RMTA21-BSMRMPOTYLicensed by some of the largest asset managers and Insurance companies in the U.S., a significant component of the Edge Platform’s value is derived from its ability to serve as a one-stop shop for research, pre-trade analytics, pricing and risk quantification, and reporting. Edge’s cloud-native infrastructure allows RiskSpan clients to scale as needs change and is supported by RiskSpan’s unparalleled team of domain experts — seasoned practitioners who know the needs and pain points of the industry firsthand. 

Adjudicators cited the platform’s “strong data management and overall technology” and “best-practice quant design for MBS, structured products and loans” as key factors in the designation.

GET A DEMO

Edge’s flexible configurability enables users to create custom views of their portfolio or potential trades at any level of granularity and down to the loan level. The platform enables researchers and analysts to integrate conventional and alternative data from an impressive array of sources to identify impacts that might otherwise go overlooked.

For clients requiring a fully supported risk-analytics-as-a-service offering, the Edge Platform provides a comprehensive data analysis, predictive modeling, portfolio benchmarking and reporting solution tailored to individual client needs.

An optional studio-level tier incorporates machine learning and data scientist support in order to leverage unstructured and alternative datasets in the analysis.

Contact us to learn how Edge’s capabilities can transform your mortgage and structured product analytics. 

Learn more about Edge at https://riskspan.com/edge-platform/ 

Company

Products

Security & Compliance

Amazon Web Services
Microsoft

Linkedin

SOC2 Type II Certified

Get Started
Log in

Linkedin   

risktech2024