Linkedin    Twitter   Facebook

Get Started
Log In

Linkedin

Articles Tagged with: Agency MBS
Home / Blog Archives

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

EDGE: An Update on GNMA Delinquencies

In this short post, we update the state of delinquencies for GNMA multi-lender cohorts, by vintage and coupon. As the Ginnie market has shifted away from bank servicers, non-bank servicers now account for more than 75% of GNMA servicing, and even higher percentages in recent-vintage cohorts.  

The table below summarizes delinquencies for GN2 cohorts where outstanding balance is greater than $10 billion. The table also highlights, in red, cohorts where delinquencies are more than 85% attributable to non-bank servicers. That non-banks are servicing so many delinquencies is not surprising given the historical reluctance (or inability)of these servicers to repurchase delinquent mortgages out of pools (see our recent analysis on this here). This is contributing to an extreme overhang of non-bank–serviced delinquencies in recent-vintage GNMA cohorts. 

The 60-day+ delinquencies for 2018 GN2 3.5s get honorable mention, with the non-bank delinquencies totaling 84% of all delinquencies, just below our 85% threshold. At the upper end, delinquencies in 2017 30yr 4s were 93% attributable to non-bank servicers, and they serviced nearly 90% of 2019 delinquencies across all coupons.

The delinquencies in this analysis are predominantly loans that are six-months or more delinquent and in COVID forbearance.[1] Current guidance from GNMA gives servicers the latitude to leave these loans in pools without exceeding their seriously delinquent threshold.[2] However, as noted in our previous research, several non-bank servicers have started to increase their buyout activity, driven by joint-ventures with GNMA EBO investors and combined with a premium bid for reperforming GNMA RG pools. While we saw a modest pullback in recent buyout activity from Lakeview,[3] which has been at the vanguard of the activity, the positive economics of the trade indicates that we will likely see continued increases in repurchases, with 2018-19 production premiums bearing the brunt of involuntary speed increases.

Contact us if you are interested in seeing variations on this theme. Using Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.

[1] Breakdown of delinquencies available on request.

[2] GNMA APM 2020-17 extended to July 31st the exemption of counting post-COVID delinquencies as part of the servicer’s Seriously Delinquent count.

[3] Lakeview repurchased 15% of seriously delinquent loans in January, down from 22% in December. Penny Mac and Carrington continued their repurchases at their recent pace.

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/ 

EDGE: An Update on GNMA Buyout Efficiency

In July, we examined buyouts of delinquent GNMA loans, with special focus on the buyout efficiency for bank servicers. At that time, several banks were 98% to 99% efficient at buying out delinquent loans, where “efficiency” is defined as the percentage of 90+ days delinquent loans that are repurchased. In this short note, we update the buyout efficiency of major bank and non-bank servicers. 

Buyout efficiency varies widely among banks. While the most efficient banks repurchase nearly 100% of eligible loans, others, including Flagstar and Citizens Bank, opt to leave virtually all the 90+ day delinquent loans they service in securities. In the table below, we show the dollar-weighted buyout efficiencies for top banks, as well as the UPB of each bank’s unpurchased 90+ day delinquent loans, as of the January 2021 factor date.

Buyout-EfficiencyBuyout efficiency for 90+ day delinquent loans, data as of January 2021. 

Servicers listed by total UPB serviced.

The overhang of seriously delinquent loans serviced by Flagstar and Citizens is spread across several GN2 Multi-lender sectors, with concentrations of delinquent loans rising to just 1% of the total current face of 2018 4% and 2018 4.5% cohorts. If Flagstar and Citizens were to repurchase all of their delinquent loans in a single month, it would add roughly 11-12 CPR to these cohorts. This represents the upper limit in involuntary speed, and actual speeds would likely be much slower with repurchases spread over several months.

The markedly lower buyout efficiency among GNMA non-bank servicers has created involuntary prepay overhang that is potentially much more daunting. The following table summarizes top non-bank servicers, their buyout efficiency over the past two quarters, and their current overhang of 90+ day delinquent loans.

Buyout-EfficiencyBuyout efficiency for 90+ day delinquent loans, data as of January 2021.

Servicers listed by total UPB serviced.

Both Penny Mac and Lakeview have improved their buyout efficiency over the last quarter and may continue to do so, as more investors begin to embrace the GNMA EBO trade. The multi-lender cohorts with the most exposure to 90+ day DQ loans serviced by Penny Mac or Lakeview include 2020 3.5s as well as 2017-19 production 3.5s and 4s, with each cohort ranging between 4% to 5% of its current face.

This final table, below, illustrates the impact of forbearance on buyout activity among non-banks. While forbearance status seems to pose no impediment to buyouts for banks — in fact, banks with the highest buyout efficiency seem to favor repurchasing loans that are in COVID-forbearance over loans that are “naturally” delinquent – non-bank behavior is more nuanced.

Of the top five non-bank servicers, only Lakeview has generated significant repurchases of loans in COVID forbearance, repurchasing 10% of eligible loans in Q4. In the table below, we separate the 90+ day delinquent loans by their forbearance status and then compute each servicer’s buyout efficiency across these sub-cohorts.

Buyout-EfficiencyBuyout efficiency for 90+ day delinquent loans, data as of January 2021.

Lakeview’s buyout behavior suggests that forbearance is not an impediment to non-bank repurchases. If we see continued improvements in buyout efficiency over the next few months, involuntary speeds in GNMA securities have the potential to rise significantly.

Contact us if you are interested in seeing variations on this theme. Using Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.

Nearly $8 Trillion in Senior Home Equity Pushes Reverse Mortgage Market Index Upward

The NRMLA/RiskSpan Reverse Mortgage Market Index (RMMI) rose to 280.99 during the third quarter of 2020, an all-time high. This reflects a 1.6% increase in senior home equity, which now stands at an estimated $7.82 trillion. Growth in senior homeowner’s wealth was largely attributable to an estimated 1.6% (or $149 billion) increase in senior housing value, offset by 1.6% (or $28 billion) increase of senior-held mortgage debt.

The National Reverse Mortgage Lenders Association (NRMLA) and RiskSpan have published the Reverse Mortgage Market Index (RMMI) since the beginning of 2000. The RMMI provides a trending measure of home equity among U.S. homeowners age 62 and older.

The RMMI defines senior home equity as the difference between the aggregate value of homes owned and occupied by seniors and the aggregate mortgage balance secured by those homes. This measure enables NRMLA to help gauge the potential market size of those who may be qualified for a reverse mortgage product. The chart above illustrates the steady increase in this index since the end of the 2008 recession.

Increasing house prices drive the index’s upward trend, mitigated to some extent by a corresponding modest increase in mortgage debt held by seniors. The most recent RMMI report (reflecting data as of the end of Q3 20202) was published last week on NRMLA’s website.

Note on the Limitations of RMMI

To calculate the RMMI, an econometric tool is developed to estimate senior housing value, senior mortgage level, and senior equity using data gathered from various public resources such as American Community Survey (ACS), Federal Reserve Flow of Funds (Z.1), and FHFA housing price indexes (HPI). The RMMI is simply the senior equity level at time of measure relative to that of the base quarter in 2000.[1]  The main limitation of RMMI is non-consecutive data, such as census population. We use a smoothing approach to estimate data in between the observable periods and continue to look for ways to improve our methodology and find more robust data to improve the precision of the results. Until then, the RMMI and its relative metrics (values, mortgages, home equities) are best analyzed at a trending macro level, rather than at more granular levels, such as MSA.

[1] There was a change in RMMI methodology in Q3 2015 mainly to calibrate senior homeowner population and senior housing values observed in 2013 American Community Survey (ACS).

Cash-out Refis, Investment Properties Contribute to Uptick in Agency Mortgage Risk Profile

RiskSpan’s Vintage Quality Index is a monthly measure of the relative risk profile of Agency mortgages. Higher VQI levels are associated with mortgage vintages containing higher-than-average percentages of loans with one or more “risk layers.”

These risk layers, summarized below, reflect the percentage of loans with low FICO scores (below 660), high loan-to-value ratios (above 80%), high debt-to-income ratios (above 45%), adjustable rate features, subordinate financing, cash-out refis, investment properties, multi-unit properties, and loans with only one borrower.

The RiskSpan VQI rose 4.2 points at the end of 2020, reflecting a modest increase in the risk profile of loans originated during the fourth quarter relative to the early stages of the pandemic.

The first rise in the index since February was driven by modest increases across several risk layers. These included cash-out refinances (up 2.5% to a 20.2% share in December), single borrower loans (up 1.8% to 52.0%) and investor loans (up 1.4% to 6.0%). Still, the December VQI sits more than 13 points below its local high in February 2020, and more than 28 points below a peak seen in January 2019.

While the share of cash-out refinances has risen some from these highs, the risk layers that have driven most of the downward trend in the overall VQI – percentage of loans with low FICO scores and high LTV and DTI ratios – remain relatively low. These layers have been trending downward for a number of years now, reflecting a tighter credit box, and the pandemic has only exacerbated tightening.

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 These loans do not represent credit availability in the market as they likely would not have been originated today but for the existence of HARP.

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.

This analysis is developed using RiskSpan’s Edge Platform. To learn more or see a free, no-obligation demo of Edge’s unique data and modeling capabilities, please contact us.

EDGE: GNMA Forbearance End Date Distribution

With 2021 underway and the first wave of pandemic-related FHA forbearances set to begin hitting their 12-month caps as early as March, now seems like a good time to summarize where things stand. Forbearance in mortgages backing GNMA securities continues to significantly outpace forbearance in GSE-backed loans, with 7.6% of GNMA loans in forbearance compared to 3.5% for Fannie and Freddie borrowers.[1] Both statistics have slowly declined over the past few months.

Notably, the share of forbearance varies greatly amongst GNMA cohorts, with some cohorts having more than 15% of their loans in forbearance. In the table below, we show the percentage of loans in forbearance for significant cohorts of GN2 30yr Multi-lender pools.

Percent of Loans in Forbearance for GNMA2 30yr Multi-lender Pools:

Cohorts larger than $25 billion. Forbearance as of December 2020 factor date.

Not surprisingly, newer production tends to experience much lower levels of forbearance. Those cohorts are dominated by newly refinanced loans and are comprised mostly of borrowers that have not struggled to make mortgage payments. Conversely, 2017-2019 vintage 3s through 4.5s show much higher forbearance, most likely due to survivor bias – loans in forbearance tend not to refinance and are left behind in the pool. The survivor bias also becomes apparent when you move up the coupon stack within a vintage. Higher coupons tend to see more refinancing activity, and that activity leaves behind a higher proportion of borrowers who cannot refinance due to the very same economic hardships that are requiring their loans to be in forbearance.

GNMA also reports the forbearance end date and length of the forbearance period for each loan. The table below summarizes the distribution of forbearance end dates across all GNMA production. This date is the last month of the currently requested forbearance period.[2]

For loans with forbearance ending in December 2020 (last month), half have taken a total of 9 months of forbearance, with most of the remaining loans taking either three or six months of forbearance.

 

 

For loans whose forbearance period rolls in January and February 2021, the total months of forbearance is evenly distributed between 3, 6, and 9 months. Among loans with a forbearance end date of March 2021, more than half will have taken their maximum twelve months of forbearance.[3]

In the chart below, we illustrate how things would look if every Ginnie Mae loan currently in forbearance extended to its full twelve-month maximum. As this analysis shows, a plurality of these mortgages – more than 25 percent — would have a forbearance end date of March 2021, with the remaining forbearance periods expiring later in 2021.

A successful vaccination program is expected to stabilize the economy and (hopefully) end the need for wide-scale forbearance programs. The timing of this economic normalization is unclear, however, and the distribution of current end dates, as illustrated above, suggests that the existing forbearance period may need to be extended for some borrowers in order to forestall a potentially catastrophic credit-driven prepayment spike in GNMA securities.

Contact us if you interested in seeing variations on this theme. Using Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.

[1] As of the December 2020 factor date, using the data reported by the GSEs and GNMA. This data may differ marginally from the Mortgage Bankers Association survey, which is a weekly survey of mortgage servicers.

[2] Data as of the December 2020 factor date.

[3] Charts of January, February and March 2021 rolls are omitted for brevity. See RiskSpan for a copy of these charts.

EDGE: COVID Forbearance and Non-Bank Buyouts

November saw a significant jump in GNMA buyouts for loans serviced by Lakeview. Initially, we suspected that Lakeview was catching up from nearly zero buyout activity in the prior months, and that perhaps the servicer was doing this to keep in front of GNMA’s requirement to keep seriously delinquent loans below the5% of UPB threshold. [1]

 

Buyout rates for some major non-bank servicers.

Using EDGE to dig further, we noticed that Lakeview’s buyouts affected both multi-lender and custom pools in similar proportions and were evenly split between loans with an active COVID forbearance and loans that were “naturally” delinquent.

The month-on-month jump in Lakeview buyouts on forborne loans is notable. The graph below plots Lakeview’s buyout rate (CBR) for loans that are 90-days+ delinquent.

Further, the buyouts were skewed towards premium coupons. Given this, it is plausible that the buyouts are economically driven [2] and that Lakeview is now starting to repurchase and warehouse delinquent loans, something that non-banks have struggled with due to balance sheet and funding constraints.

Where do the current exposures lie? The table below summarizes Lakeview’s 60-day+ delinquencies for loans in GN2 multi-lender pools, for coupons and vintages where Lakeview services a significant portion of the cohort. Not surprisingly, the greatest exposure lies in recent-vintage 4s through 5s.

To lend some perspective, in June 2020 Wells serviced around one-third of 2012-13 vintage 3.5s and approximately 8% of its loans were 60-days delinquent, all non-COVID related.

This analysis does not include other non-bank servicers. As a group, non-bank servicers now service more than 80% of recent-vintage GN2 loans in multi-lender pools. The Lakeview example reflects mounting evidence that COVID forbearance is not an impediment to repurchasing delinquent loans.

If you interested in seeing variations on this theme, contact us. Using Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.

 

 

[1] Large servicers are required to keep 90-day+ delinquencies below 5% of their overall UPB. GNMA has exempted loans that are in COVID forbearance from this tally.

[2] Servicers can repurchase GNMA loans that have missed 3 or more payments at par. If these loans cure, either naturally or due to modification, the servicer can deliver them into a new security. Given that nearly all GNMA passthroughs trade at a significant premium to par, this redelivery can create a substantial arbitrage opportunity, even after accounting for the trial period for the modification.

EDGE: Unexplained Prepayments on HFAs — An Update

In early October, we highlighted a large buyout event for FNMA pools serviced by Idaho HFA, the largest servicer of HFA loans. On October 28, FNMA officially announced that there were 544 base-pools with erroneous prepayments due to servicer reporting error. The announcement doesn’t mention the servicer of the affected pools, but when we look at pools that are single-servicer, every one of those pools is serviced by Idaho HFA.

FNMA reports the “September 2020 Impacted Principal Paydown” at $133MM. The September reported prepayment for FNMA Idaho HFA pools was 43 CPR on a total of just over $6B UPB. If we add back the principal from the impacted paydown, the speed should have been 26 CPR, which is closer to the Freddie-reported 25 CPR.

FNMA provides an announcement here and list of pools here. According to the announcement, FNMA will not be reversing the buyout but instead recommends that affected investors start a claims process. We note that Idaho HFA prepayment speeds will continue to show these erroneous buyouts in the October factor date.

Contact us to try Edge for free.

RiskSpan VQI: Current Underwriting Standards Q3 2020

Sept 2020 Vintage Quality Index

Riskspan VQI Historical Trend

Riskspan VQI Historical Trend

RiskSpan’s Vintage Quality Index, which had declined sharply in the first half of the year, leveled off somewhat in the third quarter, falling just 2.8 points between June and September, in contrast to its 12 point drop in Q2.

This change, which reflects a relative slowdown in the tightening of underwriting standards reflects something of a return to stability in the Agency origination market.

Driven by a drop in cash-out refinances (down 2.3% in the quarter), the VQI’s gradual decline left the standard credit-related risk attributes (FICO, LTV, and DTI) largely unchanged.

The share of High-LTV loans (loans with loan-to-value ratios over 80%) which fell 1.3% in Q3, has fallen dramatically over the last year–1.7% in total. More than half of this drop (6.1%) occurred before the start of the COVID-19 crisis. This suggests that, even though the Q3 VQI reflects tightening underwriting standards, the stability of the credit-related components, coupled with huge volumes from the GSEs, reflects a measure of stability in credit availability.

Risk Layers Historical Trend

Risk Layers – September 20 – All Issued Loans By Count

FICO < 660 - Share Issued Loans

Loan to Value > 80 - Share of Issued Loans

Debt-to-Income > 45 - Share of Issued Loans

Ajustable-Rate-Share-of-Issued-Loans

Loans-w-Subordinate-Financing-Sept-2020

Cashout-Refinance

Risk Layers – September 20 – All Issued Loans By Count

Loan-Occupancy

Multi-Unit-Share-of-Issued-Loans

One-Borrower-Loans

Analytical And Data Assumptions

Population assumptions:

  • Monthly data for Fannie Mae and Freddie Mac.

  • Loans originated more than three months prior to issuance are excluded because the index is meant to reflect current market conditions.

  • 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 HARP.                                                                                                                          

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.                                                

Logo

Company

Products

Security & Compliance

Amazon Web Services
Microsoft

Linkedin

SOC2 Type II Certified

Get Started
Log in

Linkedin   

risktech2024