In the July prepayment report, many cohorts of GN2 multi-lender pools saw a substantial jump in speeds. These speeds were driven by large delinquency buyouts from banks, mostly Wells Fargo, which we summarized in our most recent analysis. Speeds on moderately seasoned GN2 3% through 4% were especially hard-hit, with increases in involuntary prepayments as high as 25 CBR.
The upcoming August prepayment report, due out August 7th, should be substantially better. Delinquencies for banks with the highest buyout efficiency are significant lower than they were last month, which will contribute to a decrease in involuntary speeds by 5 to 15 CBR, depending on the cohort. In the table below, we show potential bank buyout speeds for some large GN2 multi-lender cohorts. These speeds assume an 80% roll-rate from 60DQ to 90DQ and 100% buyouts from the banks mentioned above. The analysis does not include buyouts from non-banks, whose delinquencies continue to build.
We have details on other coupon and vintage cohorts as well as buyout analysis at an individual pool level. Please ask for details.
——————————————————
If you are interested in seeing variations on this theme, contact us. Using Edge, we can examine any loan characteristic and generate an S-curve, aging curve, or time series.
Ginnie Mae prepayment speeds saw a substantial uptick in July, with speeds in some cohorts more than doubling. Much of this uptick was due to repurchases of delinquent loans. In this short post, we examine those buyouts for bank and non-bank servicers. We also offer a method for quantifying buyout risk going forward.
For background, GNMA servicers have the right (but not the obligation) to buy delinquent loans out of a pool if they have missed three or more payments. The servicer buys these loans at par and can later re-securitize them if they start reperforming. Re-securitization rules vary based on whether the loan is naturally delinquent or in a forbearance program. But the reperforming loan will be delivered into a pool with its original coupon, which almost always results in a premium-priced pool. This delivery option provides a substantial profit for the servicer that purchased the loan at par.
To purchase the loan out of the pool, the servicer must have both cash and sufficient balance sheet liquidity. Differences in access to funding can drive substantial differences buyout behavior between well-capitalized bank servicers and more thinly capitalized non-bank servicers. Below, we compare recent buyout speeds between banks and non-banks and highlight some entities whose behavior differs substantially from that of their peer group.[1]
In July, Wells Fargo’s GNMA buyouts had an outsized impact on total CPR in GNMA securities. Wells, the largest GNMA bank servicer, exhibits extraordinary buyout efficiency relative to other servicers, buying out 99 percent of eligible loans. Wells’ size and efficiency, coupled with a large 60-day delinquency in June (8.6%), caused a large increase in “involuntary prepayments” and drove total overall CPR substantially higher in July. This effect was especially apparent in some moderately seasoned multi-lender pools. For example, speeds on 2012-13 GN2 3.5 multi-lender pools accelerated from low 20s CPR in June to mid-40s in July, nearly converging to the cheapest-to-deliver 2018-19 production GN2 3.5 and wiping out any carry advantage in the sector.
This CPR acceleration in 2012-13 GN2 3.5s was due entirely to buyouts, with the sector buyouts rising for 5 CBR to 29 CBR.[2] In turn, this increase was driven almost entirely by Wells, which accounted for 25% of the servicing in some pools.
In the next table, we summarize performance for the top ten GNMA bank servicers. The table shows loan-level roll rates from June to July for loans that started June 60-days delinquent. Loans that rolled to the DQ90+ bucket were not bought out of the pool by the servicer, despite being eligible for it. We use this 90+ delinquency bucket to calculate each servicer’s buyout efficiency, defined as the percentage of delinquent loans eligible for buyout that a servicer actually repurchases.
Roll Rates for Bank Servicers, for July 2020 Reporting Date
Surprisingly, many banks exhibit very low buyout efficiencies, including Flagstar, Citizens, and Fifth Third. Navy Federal and USAA (next table) show muted buyout performance due to their high VA concentration.
Next, we summarize roll rates and buyout efficiency for the top ten GNMA non-bank servicers.
Roll Rates for Ginnie Mae Non-bank Servicers, for July 2020 Reporting Date
Not surprisingly, non-banks as a group are much less efficient at buying out eligible loans, but Carrington stands out.
Looking forward, how can investors quantify the potential CBR exposure in a sector? Investors can use Edge to estimate the upcoming August buyouts within a sector by running a servicer query to separate a set of pools or cohort into its servicer-specific delinquencies.[3] Investors can then apply that servicer’s 60DQ->90DQ roll rate plus the servicer’s buyout efficiency to estimate a CBR.[4] This CBR will contribute to the overall CBR for a pool or set of pools.
Given the significant premium at which GNMA passthroughs are trading, the profits from repurchase and re-securitization are substantial. While we expect repurchases will continue to play an outsized role in GNMA speeds, this analysis illustrates the extent to which this behavior can vary from servicer to servicer, even within the bank and non-bank sectors. Investors can mitigate this risk by quantifying the servicer-specific 60-day delinquency within their portfolio to get a clearer view of the potential impact from buyouts.
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] This post builds on our March 24 write-up on bank versus non-bank delinquencies, link here. For this analysis, we limited our analysis to loans in 3% pools and higher, issued 2010-2020. Please see RiskSpan for other data cohorts.
[2] CBR is the Conditional Buyout Rate, the buyout analogue of CPR.
[3] In Edge, select the “Expanded Output” to generate servicer-by-servicer delinquencies.
[4] RiskSpan now offers loan-level delinquency transition matrices. Please email techsupport@riskspan.com for details.
Inspection waivers have been available on agency-backed mortgages since 2017, but in this era of social distancing, the convenience of forgoing an inspection looks set to become an important feature in mortgage origination. In this post, we compare prepayments on loans with and without inspections.
Broadly, FNMA allows inspection waivers on purchase single-family mortgages up to 80% LTV, and no cash-out refi with up to 90% LTV (75% if the refi is an investment property). Inspection waivers are available on cash-out refis for primary residences with LTV up to 70%, and investment properties with LTV up to 60%.
Inspection waivers were first introduced in mid-2017. In 2018, the proportion of loans with inspection waivers held steady around 6% but started a steady uptick in the middle of 2019, long before the pandemic made social distancing a must.[1]
In the current environment, market participants should expect a further uptick in loans with waivers as refis increase and as the GSEs consider relaxing restrictions around qualifying loans. In short, PIW will start to become a key factor in loan origination. Given this, we examine the different behavior between loans with waivers and loans with inspections.
In the chart below, we show prepayment speeds on 30yr borrowers with “generic” mortgages,[2] with and without waivers. When 100bp in the money, “generic” loans with a waiver paid a full 15 CPR faster than loans with an inspection appraisal. Additionally, the waiver S-curve is steeper. Waiver loans that are 50-75bp in the money outpaced appraised houses by 20 CPR.
Next, we look at PIW by origination channel. For retail origination, loans with waivers paid only 10-15 CPR faster than loans with inspections (first graph). In contrast, correspondent loans with a waiver paid 15-20 CPR faster versus loans with an inspection (second graph).
We also looked at loan purpose. Purchase loans with a waiver paid only 10 CPR faster than comparable loans purchase loans with an inspection (first graph), whereas refi loans paid 25 CPR faster when 50-75bp in the money.
We also examined servicer-specific behavior for PIW. We saw both a difference in the proportional volume of waivers, with some originators producing a heavy concentration of waivers, as well as a difference in speeds. The details are lengthy, please contact us on how to run this query in the Edge platform.
In summary, loans with inspection waivers pay faster than loans without waivers, but the differentials vary greatly by channel and loan purpose. With property inspection waivers rising as a percentage of overall origination, these differences will begin to play a larger role in forming overall prepayment expectations.
If you interested in seeing variations on this theme, contact us. Using RS Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.
[1] Refi loans almost entirely drove this uptick in waivers, see RiskSpan for a breakdown of refi loans with waivers.
[2] For this query, we searched for loans delivered to 30yr deliverable pools with loan balance greater than $225k, FICO greater than 700, and LTV below 80%.
On Tuesday, the market received a modicum of clarity around Agency prepayments amid the uncertainty of COVID-19, when the FHFA released new guidelines for mortgage borrowers currently in forbearance or on repayment plans who wish to refinance or buy a new home.
Borrowers that use forbearance will most likely opt for a forbearance deferment, which delays the missed P&I until the loan matures. The FHFA announcement temporarily declares that borrowers are eligible to refinance three months after their forbearance ends and they have made three consecutive payments under their repayment plan, payment deferral option, or loan modification.”
With the share of mortgage loans in forbearance accelerating to over 8 percent, according to the MBA, and retail mortgage interest rates remaining at historically low levels, the FHFA’s announcement potentially expands the universe of mortgages in Agency securities eligible for refi. However, mortgage rates must be sufficiently low as to make economic sense to refinance both the unpaid principal balance of the loan and the deferred payments, which accrue at 0%. We estimate that a 6-month forbearance means that rates must be an additional 25bp lower to match the same payment savings as a borrower who doesn’t need to refinance the deferred payments. In turn, this will slow refinancing on loans with a forbearance deferment versus loans without forbearance, when faced with the same refinancing incentive. This attenuated refi activity is on top of the three-payment delay after forbearance is over, which pushes the exercise of the call option out three months and lowers the probability of exercise. In total, loans in forbearance will both be slower and have better convexity than loans not in forbearance.
Today’s FHFA release also extends Fannie’s and Freddie’s ability to purchase single-family mortgages currently in forbearance until at least August 31, 2020.
With the rapid rise of social distancing and a looming recession, investor thoughts are turning towards mortgage delinquencies and defaults. In GNMAs, loans that are 90+ days delinquent may be bought out of the pool by the servicer. When a servicer does this, the repurchase shows up as an involuntary prepay for the investor. GNMA servicers may buy the loan out a pool when it turns 90 days delinquent or more but must do so using their own capital. Given this, we may start to see a separation in buyout behavior between well-capitalized bank servicers and more thinly capitalized non-bank servicers, with longer liquidation timelines for some entities over others.¹
The GNMA loan–level data shows each loan’s delinquency status, listed from current to 180+ days delinquent.² In Edge, users can run either a single pool or a portfolio and separate the loans into buckets by individual servicer. Users can simultaneously overlay other filters such as loan guarantor, geography, mark-to-market LTV, and other. Doing this at portfolio level can help quantify a portfolio’s exposure to various bank and non-bank servicers segregated by different loan characteristics.
It is difficult to predict the exact repurchase differential we will see between bank and non-bank servicers, but for MBS investors, it will certainly be important to quantify the exposure at both pool and portfolio level as a first step. Most market participants expect a substantial uptick in the number of involuntary prepayments in the GNMA space. Edge lets users rapidly assess delinquency exposure across many different loan characteristics for an entire portfolio, which may matter more now than it ever has in recent history.
Loan-level delinquency by servicer for 2019-vintage GN Multi-lender pools, broken out by FHA/VA. This search is simple to execute in the Edge platform. Contact us for details.
¹ In 2011-12, the market saw significant differences in buyout behavior, for example Bank of America was slow to buy out delinquent loans.
² On Bloomberg, the delinquency states 90 days onward are compressed into a single 90+ state.
Over the last six months, TBA speeds have progressively accelerated and are poised to print even faster given the recent lows in primary rates and near doubling in the Refi Index. But how do these speeds compare against previous refi cycles? In this short piece, we compare today’s S-curve against the 2012-13 cycle and the massive refi wave in 2002-03.
First, we start by running an S-curve on loans in recently issued Majors¹, filtering for loans that were 2019 vintage and at least 6 months seasoned. Below, we plot prepay speeds against refi incentive². In aggregate, fully in-the-money mortgages pay around 55 CPR, with actual speeds varying from pool to pool.
Next, we overlay a TBA S-curve during the 2012-13 refi wave, covering the period both pre– and post–QE3 (September 2012). Traders during that time will remember top speeds in the mid to upper 30s. Clearly the 2019-20 prepay experience has exceeded these speeds for similar refi incentive.
Finally, we compare the current refi environment against the 2002-03 environment—the gold standard for refinancing waves. Traders active during that period will remember this as a time of high cash-out refis, which drove both in– and out-of-the money prepayments higher. Out-of-the-money mortgages paid in the mid-teens as homeowners tapped their homes like an ATM, while in-the-money mortgages paid in the high 60s to low 70s in aggregate, with some sectors paying in the 80s for several months.
The 2002-03 refi wave also featured a surge driven by the “media effect.” With nearly the entire mortgage market in–the–money, a combination of aggressive advertising plus frequent news reporting reminded homeowners at every turn that they could save hundreds of dollars per month if they refinanced their loans. In early 2003, 95% of the mortgage market was 50bp or more in the money, compared with 80% today.
We conclude that with Fed easing and recession fears, 2020 could see a renewed media effect, which may help drive prepayments higher at every point on the S-curve.
If you are interested in seeing variations on this theme, contact us. Using RS Edge, we can examine any loan characteristic and generate a S-curve, aging curve, or time series.
¹ We did a similar analysis on multi-lender Giants. Please contact us for details.
² To be included, the loan had to be at least 6 months old, higher than $225k balance, LTV < 80, and FICO > 700.
In 2019, the non-bank servicing sector continued to grow faster than traditional bank-servicers. As a group, non-bank servicers now represent nearly half of the agency MBS market, with outsized representation in newer-production mortgages. Their aggressive refinancing has driven speeds on in-the-money mortgages to post-crisis highs, and we believe this behavior will continue into 2020.
But within the non-bank sector, prepayment behavior varies widely. In this short post, we measure the fastest non-bank servicers against their cohorts and against the wider market.
We used the Edge platform to generate WALA ramps for the top 25 non-bank servicers for 30yr “generic” mortgages.¹ In the first graph, we show WALA ramps for bank-serviced and non-bank–serviced loans that were 75-125bp in the money over the last calendar year. At the peak, non-bank servicers outstripped bank servicers by roughly 8 CPR.
In the next chart, we break out performance for the two fastest non-bank servicers: United Shore and Provident Funding.² United Shore clocked in at a blazing 83 CPR for the 7-8 WALA bucket with Provident printing in the high 70s.
Switching to SMM—the right way to examine such fast speeds—we see that loans serviced by United Shore paid at 13.7 SMM, more than twice the unscheduled principal per month than the cohort of non-bank servicers in months 7 and 8.
In closing, we note that newer vintage Freddie Mac Supers consistently contain more United Shore and Provident product than similarly aged Fannie Mae Majors. Together, United Shore and Provident account for 14-18% of newer–production Freddie Supers, such as FR SD8016, SD8005, SD8001, and SD8006, but only 4-6% of Fannie Majors, such as FN MA3774 or MA3745. Most of the fast-payer Freddie Supers are 3s and 3.5s and may not show fast speeds at current rates, but in a 25-50bp rally we may see separation between Fannie and Freddie TBA speeds. As a consequence, Freddie Supers may have worse convexity than similar vintage Fannie Majors.
If you are interested in seeing variations on this theme, contact us. Using RS Edge, we can examine any loan characteristic and generate a S-curve, WALA curve, or time series. [startapp_separator border_width=”1″ opacity=”25″ animation=””]¹For a loan to be included, it had to be securitized into a deliverable 30yr Fannie or Freddie pool and have a loan balance greater than $225,000, FICO > 700, LTV <= 80, and not in NY state. All analysis was done at loan level.
²New Residential and Home Point Financial receive an honorable mention for fast speeds. Their speeds showed more response for loans 50-100bp in the money but started to converge to average non-bank speeds when 75-125bp in the money. See RiskSpan for details.
In the past few months, recent-vintage FNMA Major pools have shown significant acceleration in prepay speeds, significantly impacting TBA prices and dollar rolls. In our August report, we showed a progression of ever faster WALA ramps on FNMA Major pools1. In this installment, we update that behavior using data from Edge, the online prepayment graphing tool.
We start with a population of recent FNMA Majors and generate WALA ramps at loan level, to capture the precise behavior of the WALA ramp. In the first chart, we show loans from Majors that are 75-125bp in the money, approximately TBA 4s, over three different time periods:
- August 2018 to July 2019 (“baseline”)
- August-September 2019
- October 2019
In October, aggregate speeds on Majors hit a new high of 60 CPR for loans in the 9-10 WALA range. More troubling: the tail of the WALA ramp moved higher by roughly 5 CPR. This acceleration impacts carry in the 12mo+ seasoning range and is a potential negative for valuations in the TBA sector.
Graph: Speeds on loans from FN Major pools, holding refi incentive 75-125bp over three different periods.
In the next graph, we use Edge to isolate loans in Major pools that are 25-75bp in the money (approximately 3.5s). Similar to 4s, the progression in the aging curve shows the same story: a faster tail for loans 10+ months seasoned.
Graph: Speeds on loans from FN Major pools, holding refinancing incentive 25-75bp over three different periods.
We next look at the change in prepayment speeds from the Aug-Sep period to October and attribute that change to the origination channel. On average, FNMA Major pools are 50:50 Retail origination versus TPO, and we break down the speed contribution into these two groups. In the analysis below, we look at the speed change in each WALA bucket.
For Major 3.5s, the TPO loans accelerated more than the Retail origination loans. But in Major 4.0s, the speeds increased almost equally across each bucket.
In summary, the WALA ramp for TPO is more sensitive than Retail loans when refinancing incentive is small. But when loans are far enough in the money the increase in the WALA ramps are evenly distributed across origination channel.
We continue to monitor the ever-accelerating speeds on FNMA Majors and Freddie Giants, but the trend is clear – the fastest, cheapest to deliver TBA continues to be faster for longer. This makes the ongoing analysis of prepays, whether specified pools or non-spec deliverables, more important that it has been in previous rate cycles.
If you interested in seeing variations on this theme, contact us. Using Edge, we can examine any loan characteristic and generate a S-curve, WALA ramp, or time series.
1See RiskSpan for a similar analysis on newer WALA multi-lender Giants