mi ac pe rspe ctive s · predictive models are commonly based on historical experience/quantitative...

MI AC PE RSPE CTIVE SFall 2017

TABLE OF CONTENTS

CECL – Raising the Standards of SuccessDean C. Hurley, Managing Director, Structured Products Jeffrey Zuckerman, Vice President, Capital Markets Group

CECL is bringing fair value into the ALLL. We describe how the standards are impacting the financial industry, and how MIAC is partnering with organizations to ensure a smooth implementation.

03

Mortgage Benchmark Prices and Yields in a Post-LIBOR WorldK. Daniel Libby, CFA, Senior Vice President, Capital MarketsGroup

For purposes of lowering mortgagor borrowing costs, building a stronger banking system through more efficienthedging of mortgage risk, and encouraging additional capital to enter the mortgage sector, we believe there isample reason to include a mortgage current coupon yield (CCY) index as one of the alternative indices to LIBOR.

07

Residential MSR Market UpdateMike Carnes, Managing Director, MSR Valuations, Capital Markets Group

As measured by Bankrate, month-over-month primary market 30-Yr conventional mortgage rates increased bythree basis points to end the month of October at 3.83%. Quarter-over-quarter, 30-Yr conventional ratesexperienced a slight two basis point decline.

11

Whole Loan Execution UpdateBrendan Teeley, Senior Vice President, Whole Loan Sales & Trading

Buyers and Sellers of mortgage loans should insist on working with trading partners who are seasoned intransacting, performing valuation, and hedging as diverse a population of mortgage products as possible, withoptimal results.

13

Selling Seasoned Whole Loans to the GSEs or Ginnie MaeJason Eisendrath, Director, Loan Sale Strategies, MDS - Mortgage Delivery Specialists - Part of MIAC

If a bank or credit union decides to sell Agency-originated loans that are currently being held in their portfolio, what are their options?

17

Superior Integration to FNMA Cash Window in MarketShield®v5.0Tina Freeman, CFA, Managing Director, Secondary Solutions Group

MIAC’s secondary markets trading desks have been enjoying the recent addition of a new, advanced integrationto the FNMA Cash Window that is providing us with substantial time savings. We are pleased to announce thatthis enhancement to our MarketShield® platform is now available for client use.

19

03 CECL: Raising the Standards of Success

Overview of the RulesCECL overhauls the current impairment models forloans, leases and debt securities, and also impactscommitments. It removes the “probable” thresholdunder the “incurred loss model” for recognizing creditlosses.

Firms will be required to report the current estimate oflifetime loan losses, incorporated into the Allowancefor Loan and Lease Losses (ALLL). While a discountedcash flow (DCF) approach was considered by FASB inexposure drafts, the final standard allows anyapproach, as long as it is reasonable. Institutions,Auditors and Regulators will decide, so earlydiscussions are encouraged.

Both quantitative and qualitative methods are to beutilized jointly. Although there is a strong a bias to theuse of cash flow models with assumptions powered byquantitative data with “reasonable” scenarios, ahistorical loss-based result which could meet thestandards.

CECL requires that estimates for losses be based onrelevant information about past events, including bothqualitative and quantitative factors, such as historicalloss experience with similar assets, and then-currentconditions. Evaluations of the borrowers’creditworthiness through reasonable, supportableforecasts that demonstrate the expected collectabilityof the remaining assets contractual cash flows isencouraged.

Predictive models are commonly based on historicalexperience/quantitative data. Reasonable forecastsand conditional assessments are qualitative in nature,as they provide a forecast and estimates.

Other qualitative factors include changes in:

Lending policies and procedures, collections, etc.

Experience of management and staff

Quality of the loan review system

Financial assets carried at amortized cost less a lossallowance reflect the current expected cash flows to becollected, and income statements will reflect creditdeterioration or improvement.

For financial assets carried at fair value (FV) withfluctuations recognized through other comprehensiveincome (OCI), the balance sheet would reflect fairvalue, but the income statement would reflect creditdeterioration or improvement.

Under some circumstances, institutions can elect notto recognize expected credit losses on assets held atfair value. The conditions are that the FV equals orexceeds the amortized cost and if expected creditlosses are immaterial.

What Will it Take to Implement CECL?An array of new processes will be required. Policiesand procedures will need to be revamped inmanagement, governance, risk reporting, controls andfunctional integration.

Program Management will be the start: newfoundcoordination among functional areas such as finance,originations, credit, operations and technology, and arevised governance and risk management framework.

Segmentation of loan, lease and debt portfolios intoclearly identifiable portions with similar and discretecharacteristics is the next step.

CECL: Raising the Standards of Success

M I A C P E R S P E C T I V E S F a l l 2 0 1 7


This means the specific identification and descriptionof the characteristics of the assets to be modeled forprobability of default (POD) and loss given default(LGD) to make the ALLL processes Basel compliant.Development of specific credit risk modeling andforecasting models and tools will be the theme of theday.

FASB will require institutions to develop well-documented data management and quality processes.Validated Extract-Transform-&-Load (ETL) systems,data quality reporting and procedures must be created.Leading examples of these include MIAC’sDataRaptor®, DR-SurveillanceTM, and the processesembodied respectively in both, facilitate the creation ofreliable data warehouses for the segmentation ofportfolios.

Next, firms will identify control points in the originationand operations areas to insure adequate support anddata capture, and integrate new tools within financialand regulatory reporting procedures.

Importantly, dual jurisdiction reporting institutions -CECL and IFRS 9 - firms will need to:

Handle 12 month (IFRS 9) versus lifetime (CECL)credit loss projections and reporting in models andreport generation

Estimate credit losses for future draws oncommitments for IFRS and for commitments thatcannot be unconditionally canceled for CECL

FASB’s CECL Implementation Process

MIAC: Planning for CECLThere are definite steps that organizations need totake, and MIAC suggests the following. Start withdefining a revised governance standard for CECL, andestablish a steering committee or task force withmembers currently in high-level positions in finance,originations, credit and operations who havemanagement backing. This team needs to be givenbudget and action authority to implement prescribedprocedures.

Then, as required, firms will need to identifyappropriate external consultants and partners who canparticipate and contribute to the process, and shouldseek ways to integrate them into the process early.These firms should be able to create and/or evaluatemodels for conformance with the new standards.

Next, these teams will determine the resource needsinvolved in each area and inventory what exists today,beginning with an initial portfolio segmentation of allloan, lease and debt assets held. Determine what datais needed, what models are needed, as well astechnology needs by portfolio.

The committee will need to examine existing modelsand methods used in the ALLL process to determinewhich have the potential to meet the newrequirements. Generally, cash flow forecasting modelsoffer potential, and static models do not. Firms willperform pilot evaluations of the potential impact ofCECL.



These findings will be used to build the “how to”document or “roadmap”. Key objectives andmilestones should be along these lines:

Portfolio SegmentationSegment the loan, lease and debt portfolio intomeaningful segments, so that then, firms canappropriately define data elements needed for each.This leads to the identification of the critical statisticaldrivers of performance, which will be used in themodeling and reporting, and defining data models.

Data ManagementThe CECL plan should include the data warehousingand capture infrastructure, tools and data modelsrequired, which will require implementation of powerfuldata capture methods, in monthly snapshots, includingcredit data and performance data. Firms will need toobtain and reconstruct as much historical data aspossible, plus review existing and alternative models,research credit modeling, loss modeling and voluntaryprepayment modeling.

Firms must identify models that can be used orrepurposed for different products, internal andexternal, and identify leverage opportunities andefficiency gains from current models, processes,workgroups, modeling approaches (ALLL, DFAST, orinternal vs. vendor models).

Develop Metrics and AssumptionsMIAC helps firms use these data models to developkey reporting metrics and descriptions of the driversused in the credit, loss and prepayment models,determine assumptions, and drive the building of adescriptive narrative of all the models selected and theprocesses to be used.

Model Design and DevelopmentInstitutions will utilize the warehouses and the tools todevelop historical analysis and internal models and willintegrate results with existing systems. The task willbe to expand these systems to handle scenario inputs,and develop financial and managerial reportingformats, as FASB will require lenders to define andexplain the use of any credit grading systems and otherqualitative inputs to the ALLL process.

Scenario DesignCECL urges banks to develop a narrative to explain thebasis of the scenarios and why they are deemed to bereasonable. Regulators will review the scenariorationales with senior management, so institutions arepreparing currently with external consultants andauditors.

Firms will begin parallel test runs of the new CECLprocess, and the existing ALLL process. The challengeis to identify: volatility in results and capital impacts,and evaluate strategic considerations across productlines. Naturally, product pricing, origination standards,processing, collections and operations, will be of theessence.

The CECL Committee is encouraged to work closelywith auditors and external partners to finalize thenarratives and reporting.

01 SEGMENTATION

02 MANAGEMENT

03 METRICS

04 MODELING

05 SCENARIOS



Disclosure Format Examples and SampleSizeFASB provides examples as to what reportingdisclosures look like; which are not required formats,but guidance. Internal reports can be tailored to satisfyreporting needs such as allowance aggregation. FASBhas also set a requirement (FAS Topic 326) fordisaggregation of receivables by credit metric and byvintage year where receivables more than 5 years oldare aggregated. See figure 1 as an example.

CECL PartnersMIAC’s core competencies qualify us as a collaboratorfor CECL planning at financial institutions in the USAand Europe.

Our seasoned executive team leads relationships withthe most established entities in consumer lending,servicing, and regulation is what sets us apart. Ourasset-specific knowledge in whole loan valuation, ALLL,trading and securitization is unrivaled.

The FASB’s prescription essentially mirrors MIAC’sbest practices: our methodology and software suitefor data warehousing and analytics, due diligence, andborrower behavior modeling is validated and accepted.

Dean C. HurleyManaging Director, Structured Products Group

Jeffrey ZuckermanVice President, Capital Markets Group


Figure 1.

P L A N R E S E A R C H E X E C U T E

Governance: Task Force Create Data Models Set or Obtain Data Tools

Evaluate Existing Infrastructure Segment Portfolios

Build Data Warehouse

Obtain/Repurpose Credit Models

Identify Consultants and External Partners

Research Credit Modeling

Create Scenarios

Identify Qualitative Factors

Create Plan with Milestones and

Procedures

Test, Parallel Runs

Strategic Concerns

Develop Narrative

Finalize Reporting

E X E C U T I V E S U M M A R Y

• For purposes of lowering mortgagor borrowing costs, building a stronger banking systemthrough more efficient hedging of mortgage risk and encouraging additional capital toenter the mortgage sector, we believe there is ample reason to include a mortgagecurrent coupon yield index as one of the alternative indices to LIBOR.

• Now that protective guidelines by self-regulating bodies have been put into place, it isreasonable to assume that financial and civil/criminal penalties for the next crisisinvolving financial index reporting and misalignment of interest (real or perceived) will befar harsher.

• MIAC has developed an IOSCO compliant TBA Pricing Service providing Level 1, 2 and 3prices for the entire TBA market including all contract months as well as their associatedCurrent Coupon Yields (CCY) fixings. These prices and yields are fully transparent andprovided hourly.

Mortgage Benchmark Prices and Yields in a Post-LIBOR World

08 Mortgage Benchmark Prices and Yields in a Post-LIBOR World

Recent headlines returned a sharp focus to the specterof shadow banking, opaque valuations andmisalignments of interest that result when marketagents are left unchecked. Just as homeowners andsavers had begun to see their personal balance sheetsheal from the subprime crisis and its associated equitymarket contagion, they learned in July 2012 ofsomething known as The LIBOR Scandal. On July 27,2017, things became even more interesting as the UK’sFinancial Conduct Authority (FCA) announced that itwould no longer guarantee the existence of LIBOR after2021 due to lack of trading activity in certainmaturities. That’s when homeowners and savers alikeawoke to the story.

It’s worthwhile to get a sense of the stakes at hand.Misstatements of LIBOR may have amounted to nomore than approximately a single basis point onany given fixing and over time must have beenoverwhelmed by larger factors moving the markets.Nevertheless, this was enough for one prominentfinance academic to characterize it as the “largestfinancial scam in the history of markets by an order ofmagnitude” affecting perhaps as much as $800 trilliondollars of investment contracts1. There have been morethan $9 billion of fines against global banking titanssuch as Deutsche Bank, Barclays and UBS for their rolein this fiasco with more than a dozen convictions andpending legal actions2 3.

In response to the LIBOR Scandal, The Board of theInternational Organization of SecuritiesCommissioners (IOSCO) published the Principles forFinancial Benchmarks (“the Principles”) in July 20134.This is a roadmap of best practices intended for allmarkets, i.e. equities, fixed income, currencies andcommodities. The Principles are a detailed 45-pagedocument that lays out 19 guidelines coveringGovernance, Methodology and Transparency.

Quite appropriately, the first Principles discussed in thedocument are good governance practices. The mostimportant take-aways are the following. The indexmust have an Administrator with overall responsibilityfor the index itself and oversight of third-party productsand services used in index construction.

The Administrator cannot be tainted by conflicts ofinterest or the appearance of conflicts of interest. Andthe use of opaque valuation techniques, oftentimescalled "expert judgement”, is strongly discouraged.

The other overarching theme of the Principlesis transparency. The gold-standard for publication offinancial indices/benchmarks is to use transaction datawhenever available. Certainly, the Principles recognizethere are times when recent transaction data is notavailable for a financial asset held within an index. Inthose instances, various methodological prescriptionsare available. However, clearly the best solution is touse either a recent transaction for a close surrogateasset or a clear and robust valuation formula that isitself directly tied to transaction data.

How does all this relate to LIBOR and ultimately to themortgage market? We turn to those questions next.Arising out of The LIBOR Scandal and in response torecommendations from the Treasury’s FinancialStability Oversight Council (FSOC) and the globalFinancial Stability Board (FSB), the Federal ReserveBoard together with the New York Fed convened theAlternative Reference Rates Committee (ARRC) onNovember 17, 2014. The mandate of ARRC is “…topromptly identify alternative interest rate benchmarksanchored in observable transactions and supported byappropriate governance structures…” and again later inthe same section “…to identify a set of alternativereference interest rates that are more firmly based ontransactions from a robust underlying market and thatcomply with emerging standards such as the IOSCOPrinciples for Financial Benchmarks..." 5. It’s clear fromboth passages that the mandate calls for theidentification of more than one benchmark as areplacement to LIBOR. This is because applications ofLIBOR developed over many years until they becamealmost ubiquitous, used in a myriad of ways. It isunlikely that a single replacement index will serve allpurposes for which LIBOR has grown to be used.1 http://www.accountingdegree.net/numbers/libor.php “The LIBOR Scandal Explained” The quote is attributed to Andrew Lo, Professor MIT2 https://www.cfr.org/backgrounder/understanding-libor-scandal 3 https://qz.com/723127/after-10-years-and-billions-in-fines-the-uk-has-convicted-precisely-five-people-for-rigging-interest-rates/4 https://www.iosco.org/library/pubdocs/pdf/IOSCOPD415.pdf 5 https://www.newyorkfed.org/arrc See section under “About Us”.



As one replacement index for LIBOR, we believe themarkets would benefit from the creation of mortgagebenchmark TBA Prices and associated Current CouponYield (CCY) Indices created under IOSCO Principles.We offer below three brief examples for considerationaddressing needs of homeowners, mortgage servicersand institutional investors. We offer three briefexamples for consideration addressing needs ofhomeowners, mortgage servicers and institutionalinvestors.

But aren’t there already TBA prices and CCY indicespublished on Bloomberg and elsewhere? Yes, howeverthese prices and indices suffer from all the issues thatIOSCO was created to address. We have written aboutthis elsewhere and provided data supporting thesethoughts6. We summarize here that TBA prices and theCCYs published on Bloomberg and other sources facethe following challenges:

Nevertheless, haven’t there been some OTC contractswritten on these CCY indices called Constant MaturityMortgage (CMM) Swaps to hedge the mortgage basis?Yes, however there has also been some history oflitigation over disputes of the CCY settings used invaluing them. Not surprisingly, the adoption of thesevaluable hedging tools has been muted for the samereasons that afflict CCY. This is another testament forthe need for a more standardized TBA Price and Yieldmethodology.

As the ARRC undertakes to evaluate “a set of”alternative indices to LIBOR, now is an opportune timeto take the long view and prepare for market cyclesthat lie ahead in the growth of the mortgage market.For purposes of lowering mortgagor borrowing costs,building a stronger banking system through moreefficient hedging of mortgage risk and encouragingadditional capital to enter the mortgage sector, webelieve there is ample reason to include a mortgagecurrent coupon yield index as one of the alternativeindices to LIBOR.

6 “The MIAC TBA Pricing Product” Published by MIAC Analytics

01 The use of LIBOR as an index in adjustable ratemortgages (ARMs) might be considered one suchexample. LIBOR is a short money market rate; butwhen packaged with a large margin into ARMs it isused to reflect borrowing alternatives at a muchlonger point on the yield curve. Creating ARMsindexed to the mortgage current coupon yield indexwould reduce the measurement risk in hedgingthese mortgage portfolios. This would have realworld implications for efficiently hedging ARMportfolios, in fact for hedging all high-grademortgage portfolios where mortgage basis riskrepresents the largest risk factor. Better hedgedreturns for lenders will translate into lowerborrowing costs for homeowners over time.

02 The mortgage servicing rights (MSR) market is anexample where the creation of an investible CCYunder best practices could have significant benefitsto market participants. Perhaps more than any otherasset, MSR prices are heavily tied to the movementof CCY. Current practices use imperfect hedgessuch as LIBOR-based swaps, swaptions, Treasuriesand TBAs to hedge servicing portfolios. Creation ofmore liquid hedge instruments closely tied to theirprimary risk exposure could allow servicers anddepositories to better manage their balance sheet.

03 Lastly, creating investible index products tieddirectly to CCY would allow institutional and retailinvestors alike to gain exposure to the mortgagemarkets via ETFs or index funds that would nothave the prepayment risk (aka “fat tail risk”)associated with traditional actively managedmortgage portfolios. This could attract additionalcapital to enter the mortgage market and allowinstitutional investors to have a more completeconstruction of the market portfolio.

A More Transparent and Robust Mortgage Current Coupon Yield (CCY): Application Examples

Current TBA Pricing Issues Location of/Comment on Issue

Wide Bid-Offer Spreads Key coupons & contractsthat impact CCY

Lack of Trading Activity Key coupons & contractsthat impact CCY

Opaque Pricing Methodologies

During periods with no trading activity

Non-standardization of Pricing Sources & Methodologies

Different pricing sources apply idiosyncratic & oftentimes opaque methodologies



7 ibid8 “The MIAC TBA Pricing Product: Case Study #1 – Bid/Offer Spreads as a Function of MIAC Prices” published by MIAC Analytics is the first in this series. “The MIAC TBA Pricing Product Case Study #2 – A Comparison of MIAC & Bloomberg Current Coupon Yields”

MIAC has developed an IOSCO compliant TBA PricingService providing Level 1, 2 and 3 prices for the entireTBA market including all contract months as well astheir associated CCY fixings. These prices and yieldsare fully transparent and provided hourly. Level 1prices are based on TRACE transactions. Level 2 and3 prices are based on robust spread relationships andrules that are also directly transaction-based and havebeen market tested. MIAC has published a brochureon its TBA product available upon request7.

MIAC has also begun publishing Case Studies of theirTBA Price and Yield indices8. Among other topics,these studies will address well-known weaknessesfrom other pricing sources that are addressed in theMIAC pricing methodology. They will also show tradingand liquidity characteristics for premium/discountcoupons and cash/forward contracts important tovarious market participants.

K. Daniel Libby, CFASenior Vice President, Capital Markets Group


11 Residential MSR Market Update

As measured by Bankrate, month-over-month primarymarket 30-Yr conventional mortgage rates increasedby three basis points to end the month of October at3.83%. Quarter-over-quarter, 30-Yr conventional ratesexperienced a slight two basis point decline. While thisamount of month-over-month mortgage ratemovement may produce only a minimal upward shift invalue, there was also an increase in benchmarkearnings rates. In comparison to the month-over-monthincrease in base mortgage rate, the 5-Yr swap ratewhich is MIAC’s benchmark for earnings ratesincreased in a non-parallel fashion to end the month at2.08% or approximately eight basis points higher thanSeptember’s closing rate. Likewise, quarter-over-quarter the 18 basis point increase in the 5-Yr swaprate was in stark contrast to the quarterly change inprimary mortgage rate.

Residential MSR Market UpdateConsidering its influence on mortgage rates and theimpact that a volatile CMS 10-Year Rate can have onMSR’s, it too needs to be a closely watchedbenchmark. Month-over-month the CMS 10-Yearincreased by just over 6 basis points which forhistorians is nearly 100 basis points higher than the all-time low set in July 2016.

What was even more impactful to MSR values duringthe quarter were the trading values witnessed onmostly larger MSR offerings which month-over-monthvaulted MIAC’s Conventional 30-Yr 3.50% 2017 GSAIndex north of a 4.0 multiple.

Period-over-Period Price Change by Product


12 Residential MSR Market Update

Looking to take advantage of a vastly improvedsecondary bulk market for Fannie Mae, Freddie Mac,and Ginnie Mae mortgage servicing rights, bulktransfers approaching $300 billion took place in thefirst nine months of 2017. The fourth quarter will be noexception as numerous offerings hit the market inOctober from firms wanting to get one moretransaction completed before end-of-year.

Counterparty net worth and the size of the offeringcontinue to impact who bids and at what price butduring October, “Conventional at-market multi-billionofferings” mostly experienced very respectable bidsranging from a 4.0 to 4.25 multiple. Low delinquencyconventional 30-Yr product can trade in the mid-4multiple range although sellers continue to experiencebuyer resistance at asking levels in excess of a 4.5multiple.

The smaller conventional packages categorized as$500 million or less are seeing an uptick in tradingvolume as well but at prices that “on average” can be 5to 15 basis points lower than the bid prices obtained onlarger offerings.

Similarly, trading levels on mostly newer “at-marketGinnie Mae offerings” ranged in the low to mid 3multiple range, while “stressed Ginnie Mae offerings”are subject to a much wider range that is heavilyinfluenced by, delinquencies, geography, counterpartyand potential exposure to recent storms.

MIAC’s MSR Valuation department provides MSRvaluation advisory services to over 200 institutionstotaling nearly $2 Trillion in residential MSR valuationsevery month. MIAC is an industry leader in brokeringMSR bulk and flow offerings.

Conv 30-YR 3.50% 2017 GSA Index

Mike CarnesManaging Director, MSR Valuations, Capital Markets Group


13 Whole Loan Execution Update

Market SizeResidential mortgages generally fall into one of twocategories: Agency — eligible for programs offered byFannie Mae, Freddie Mac, or Ginnie Mae (FHA/VA) —and non-Agency. As of Q12017, Agency loans areapproximately 85% of new originations. According tothe Federal Reserve Board, mortgage debt outstandingin that year totaled approximately $10.3 trillion1. Bycontrast, the private (non-Agency) market consists ofover $1 trillion of this balance. However, theimportance of the non-Agency market isdisproportionate to its market share, because non-Agency programs, typically sold through whole-loanexecution, are often where lenders can find a niche todifferentiate themselves from the competition.

Agency vs. Non-AgencyFannie Mae, Freddie Mac, and FHA/VA issue guidelinesdefining the loans that that they will purchase. Banks,mortgage companies, and other originators, generateloans that meet those guidelines, allowing consumersaccess to multiple sources of home lending.Originators compete on rate, price, and service, whileproviding consistent access to standardized financingvehicles offered by the Agencies. This creates liquidityin the market by providing a known source of fundingfor mortgage lending.

The non-Agency market, comprising loans that are notpurchased by the Agencies, serves a different tier ofconsumer, many of whom have been left out of thehousing recovery due to the drastic reduction in capitalfor non-Agency mortgages since the market down-turnbegan in 2007. These customers may not qualify foran Agency loan for any number of reasons, includinglack of credit history, self-employed income,inconsistent employment history, excluded propertytype, or loan balance. non-Agency loans are typicallyoriginated by banks to be held on their balance sheets.This risk is typically priced at an interest rate of 50-150basis points (bps) over Agency paper.

Non-Agency Loans in Secondary MarketIn the mid-2010s, mortgage lenders began venturinginto new products that are further down the creditcurve and/or allowing for impairments that previouslymay have resulted in prohibitive risk-based pricingadjustments. Specifically, larger community bankswere underwriting portfolio mortgages to higher loan-to-value ratios (LTVs), lower FICOs, and/or alternativedoc types. Generally, only one of the three criteriadeviated from what makes a loan Agency-eligible;another way to say this is that portfolio lenders seek toavoid excessive risk layering. Banks still need to beable to justify to regulators why they make a loan, and“the relationship” is not an acceptable answer. As thisstyle of community-bank lending expands, thesecondary market for these loans is currentlyoutpacing origination. Other banks are willing to pay apremium for these loans, because they can avoid theoverhead associated with retail loan origination, andthe higher yields on the loans are adequatecompensation for the (perceived or actual) incrementalcredit risk.

Whole Loan Execution Update

First Lien Residential Originations per MBA2

2017 Est. $1.627 T - 67% Purchase Money



1 https://www.federalreserve.gov/data/mortoutstand/current.htm 2 http://www.mortgagedaily.com/ResidentialStatistics.asp



Figure 1,

CRAMid 2017 has been a busy season for CRA transactionsfor MIAC. Historically, the CRA buying season does notpick up until the fall. In 2017 MIAC released a softwaretool that allows originators to parse their product byCRA credit eligibility. MIAC is uniquely suited to workwith depositories to solve their CRA credit need. Thisinvestment in technology has allowed MIAC to moreefficiently serve our Bank clients’ needs for CommunityReinvestment Act credit.

Pricing for CRA loans has historically centered around50-150 bps over TBA prices. Recent transactions havecentered around 75bps over, which we feel reflects adiscount to the true market price prior to demandincreasing. This is due to the relatively fewer Buyers inthe market, as well as efficiencies in transacting thatMIAC has generated by relying on existing Agencyclients to provide them and increase in price receivedwhile at the same time allowing Buyers to pay less thanother Sellers are asking.

Non-QM LendingThe market has nearly doubled in 2017 from 2016numbers, mainly because there are more lenderswilling to expand their credit box. In 2016 there wererelatively few lenders willing to lend to a challengedborrower at reasonable rate. This allowed for a fewmajor lenders to capture the majority of the market. In2017, we are seeing more liquidity for these borrowersfrom banks, funds and increasingly, non-bank lenders.This has the dual effect of reducing rates throughcompetition, which consequently increases demand forthis product.

MIAC has transacted a substantial volume of theseloans through our depository and fund clients. Usingour robust suite of modeling and marketing tools wehave been able to help Buyers understand the productand the Sellers to price the product correctly for asuccessful execution.

The growth we areseeing in Non-QMderives from bothcredit as well ascollateral expansion.Borrowers with moresevere credit defectsas well are obtainingfinancing at higherLTVs than previously.Additionally, higherLTVs for strongborrowers who maynot have incomedocumentation orcitizenship to qualifyfor Agency financing.See figure 1.



Pricing of Whole Loans in 2017Non-PerformingIn early 2017 there was a 3-5% bump in pricing thatBuyers were willing to pay for Non-Performing WholeLoans. With bench mark pricing centering aroundupper 50s to low 60s through 2016, there were anumber of larger transactions in the market that werewell though this pricing. MIAC executed several tradesin NPLs at the mid-60s+ level for Banks and Funds.NPLs are a very diverse market and it is difficult todraw direct correlations from one transaction to thenext, however it widely agreed that demandoutstripped supply earlier this year.

More recently, there is increasing demand and supply,although robust, appears to broadly represent the“tails” of the market. There is currently a large ~$1btrade in the market from a bank that represents onethe largest, non-government, transactions this year.When this clears the market, we will have additionalcolor, and would be glad to discuss.

Non-AgencyWith the increased demand for yield by depositoriesthere is a greater money supply for borrowers outsideof traditional Agency lending. MIAC sees this manifestin multiple ways; both an Increase in the number oflenders/banks that are willing originate a portfolioproduct as well as expansion of the credit box that isavailable. 90 LTV loans are much more common thanlast year, when 80% was generally the ceiling and 65%in many cases.

Pricing has improved for the consumer, this is due to anumber of reasons. In addition to the increase inwillingness to lend, as there is a higher level of comfortwith regulations in the Non-QM lending space. It wasnot too long ago that Non-Agency was nearlynonexistent, yet it is fairly common today withincreased access to more consumers. Rates aretypically 50-150 bps through Agency, with a greaterspread as you move further from Agency guides. MIACcurrently has two large transactions under contractwith 3% yields for short term paper.

AgencyAgency lending is very transparent and is broadlyunderstood. Included in this pricing is the expansion oflending. Agencies are willing purchase higher LTVsand lower credit scores than in 2016. There has been apush to increase home ownership, the easiest way toaccomplish this is to lend to borrowers who previouslywere not credit worthy. The MIAC Agency desk Sells$3-4 billion of mortgages each month, providingsignificant insight into the market on a daily basis.

Execution of Whole Loan SalesBuyers and Sellers of mortgage loans should insist onworking with trading partners who are seasoned intransacting, performing valuation, and hedging asdiverse a population of mortgage products as possible,with optimal results. This includes Agency-eligiblepricing execution, seasoned mortgage performing,Prime Jumbo, Hybrids, RPL’s, NPL’s, HELOC, Reversemortgage portfolio valuations, and hedging of a diversepopulation of mortgage products. It is important towork with a partner that sees as much trading andanalytic activity as MIAC, which, by virtue of thebreadth of its business activities, has frequent contactwith most of the largest originators, funds, banks, andportfolio companies.

The most efficient traders must gather the currentmarket intelligence needed to identify and engage thebest execution for a given trade. There are timingconsiderations, external events that affect a buyer’sability to focus on a trade, as well as other marketconsiderations. This includes an intimateunderstanding of firms’ business models on both thebuy and sell side, as well as their pricing requirementsand investors’ “appetites.” Some dealmakers have, atbest, a cursory level of understanding of investors’objectives. The lack of depth of understanding oftenyields a sale that has been put out so widely (shown totoo many unqualified buyers) that it becomes a tradethat many serious funds will avoid.



Best execution comes from a complete understandingof the seller’s needs, both in terms of price and timing,and the credit and liquidity nuances of the portfolio.The dealmaker you work with must understand thesenuances and match the trade with a proper number ofserious buyers who will focus on the trade with a no-fade bid that will close with a very high degree ofsurety.

Brendan TeeleySenior Vice President, Whole Loan Sales, Trading


17 Selling Seasoned Residential Whole Loans to the GSEs or Ginnie Mae

Who are the GSEs and Ginnie Mae?Fannie Mae, the Federal National MortgageAssociation (FNMA) and Freddie Mac, the FederalHome Loan Mortgage Corporation (FHLMC) are theestablished secondary market lenders responsible forthe liquidity of the majority of conventional, non-Government, conforming residential whole loans thatare originated today and for the past number ofdecades.

Ginnie Mae, or the Government National MortgageAssociation (GNMA) provides liquidity for Government-insured residential mortgages. Those include FederalHousing Administration (FHA), VeteransAdministration (VA) and the Rural HousingAdministration (RHS/USDA).

For the purpose of this article Fannie Mae, Freddie Macand Ginnie Mae will be referred to collectively as“Agency” or “Agencies”.

The normal process for new Agency-underwrittenloans is to originate them and sell or securitize themdirectly with the Agencies or sell them to an aggregatorthat would buy the loans and then deliver them to theAgencies themselves. This often happens if theoriginator does not have their Agency Seller/Servicerauthority. There are also instances where banks andcredit unions originate loans and hold them in theirportfolios. In some cases, this is because thedepository institution does not need the cash that theywould receive by selling the loans or the securities inthe secondary market and they prefer to hold the loanson their books to capture the yield.

If a bank or credit union decides to sellAgency-originated loans that are currentlybeing held in their portfolio, what are theiroptions?

There are circumstances where these depositoryinstitutions decide that they would like to sell loanswhich were originated under Agency Guidelines, and toput that cash to work in other ways or hold thesecurities in their portfolio, such as when lenderschoose to focus on originating different products ordivesting of residential whole loans.

If residential loans were originated under Agencyguidelines, an institution can start the process of aseasoned or negotiated sale with the Agencies if theyare an approved Seller and Servicer with them.

The Ginnie Mae securitization process with seasonedloans is rather simple. For Sellers with loans that wereoriginated under FHA, VA or RHS/USDA guidelines andare currently insured, they can be securitized withGinnie Mae. All of the same procedures would apply asif the Seller was securitizing loans when newlyoriginated. There are a few caveats for modified loansor those purchased out of a pool to be re-securitized.However, in general it’s very similar to delivering loansas if newly originated.

Selling seasoned whole loans to Fannie Mae andFreddie Mac is a much different process than selling toGinnie Mae. These transactions are commonlyreferred to as “Seasoned/Negotiated” or ‘Bulk’ GSEsales and are normally completed while the Seller isalso the Servicer (a/k/a servicing retained). There are,however, occasions where Sellers can sell or securitizeloans that are serviced by others (SBO). An example ofthis would be when banks purchase CRA loans fromother lending institutions on a servicing retained basis(where the selling institution retains or “holds” theservicing after the loan is sold) and they decide to sellthem at a later date.

Selling Seasoned Residential Whole Loans to the GSEs or Ginnie Mae


18 Selling Seasoned Residential Whole Loans to the GSEs or Ginnie Mae

The Seasoned GSE Process is as follows:

Bid data file of clean data needs to be put togetherto show to the GSE’s for eligibility and pricing. Thisfile would contain all of the data necessary for theGSE to run through eligibility and pricing models

Data accuracy is of utmost importance since thedata that the GSEs use to bid is what the Seller isproviding representations/warrants on.

GSE will review the data for eligibility and pricingand provide preliminary eligibility and pricing to theSeller: Generally, the minimum bid size with theGSEs is $10MM (this is negotiable)

If the eligible population and bid are acceptable to theSeller, the following steps would then be followed tocomplete the sale of the loans to the GSE:

Obtain all data necessary for selling to the GSE

Work with the document custodian to ensurecertification will be timely

Work through contracting and variances with theGSE. For a negotiated transaction, a separatecontract will be prepared for the sale.

Load the data into the appropriate delivery systemand resolve any data-related issues

Submit the data in the delivery system

Work with the document custodian to ensure thepools are all certified

Settlement occurs on chosen settlement date

When selling to the GSEs, Sellers need to understandthat they are providing representations and warrants tothe GSE Selling Guides1,2. The difference is that GSEsare looking at seasoned loans and not newproduction. Seasoned transactions typically consist ofloans seasoned anything beyond twelve months.However, 4+ month production loans can be inseasoned transactions as well.

The Agencies have assembled teams that are ready toexamine seasoned loans and determine what they canand cannot work with in these types of negotiatedtransactions and have an efficient process that hasbeen in place for decades.

MDS / MIAC professionals maintain longstandingrelationships with the Agencies (Fannie Mae, FreddieMac and Ginnie Mae) in addition to national mortgageloan servicers and mortgage loan custodians. Ourclients are able to leverage these relationships in orderto maximize their mortgage loan securitizationexecution. We utilize the MDS proprietary analyticsystem to audit and produce data uploads for thedelivery of mortgage loan data and MSR data to givebuyers and sellers the confidence in the underlyingproduct quality on which they are transacting.

1 https://www.fanniemae.com/content/guide/selling/2 http://www.freddiemac.com/singlefamily/guide/

Jason EisendrathDirector, Loan Sale Strategies, MDS - Mortgage Delivery

Specialists - Part of MIAC


19 Superior Integration to FNMA Cash Window in MarketShield® v5.0

MIAC’s secondary markets trading desks have beenenjoying the recent addition of a new, advancedintegration to the FNMA Cash Window that is providingus with substantial time savings. We are pleased toannounce that this enhancement to our MarketShield®platform is now available for client use.

Our latest release of MarketShield® v5.0 not onlyincludes fully automated pool formation andtransmission to the FNMA cash window, it alsoincludes new functionality to manage bulk bids forboth buyers and sellers. BuySide Trader will helpclients manage their bulk purchase programs, bymapping seller data imports and providing data qualitycontrols, pricing bulk pools directly to the buyer’stakeouts, allowing multiple levels of marginmanagement as well as loan-level spiffs, and emailintegration for communication of bids.

SellSide Trader provides equally robust functionality tobulk sellers, providing easy integration of bulk pool bidsinto best execution, market movement normalization,as well as email integration for communication ofpools to potential buyers.

Danielle Roper, Vice President of our New York tradingdesk, had this to say:


Superior Integration to FNMA Cash Window in MarketShield® v5.0

“The new FNMA integration feature is a real gamechanger when it comes to taking down specifiedcommitments. MS Trader automatically queues upthe pools according to the best execution and withone click the commitments are executed and finalpricing and commitment numbers are loaded backinto the system. No more back and forth betweenthe FNMA screen and internal spreadsheets ormanually keying in UPB and pass-through rates, andre-confirming the password on each take-down. Thenew process is much more efficient.”

Tina Freeman, CFAManaging Director, Secondary Solutions Group

TWI TTE R@MIACAnalytics

LINKE DI NSearch: MIAC

WEBSI TEwww.MIACAnalytics.com

ABOUT MIAC ANALYTICSFor 28 years, Mortgage Industry Advisory Corporation (MIAC) has been the preferred provider of mortgage and asset-backed valuation and hedging software, MIAC Analytics™, MSR and whole loan brokerage services, secondary market risk management, and a complete CECL (Current Expected Credit Loss) solutions.

MIAC Analytics™ is the most sophisticated mortgage pricing and risk management software suite available. The MIAC Analytics™ suite includes VeriFi™, DR-Surveillance™, MIAC CORE™, and Vision™ to address FASB’s new Current Expected Credit Loss (“CECL”) requirements with the industries best modeling practices. VeriFi™ is used to support and manage the data quality auditing and review process. DR-Surveillance™ will measure a client’s collateral behavior including historical transition roll rates and Time_in_FCL exit curves; and these client specific behaviors are integrated into MIAC CORE™, our loan level credit loss model embedded in our Vision™ cash flow engine and balance sheet model.

CONTACT

MIAC Analytics521 Fifth Avenue, 9th FloorNew York, NY 10175

(212) 233 – [email protected]

CONNECT

http://www.MIACAnalytics.com/

mi ac pe rspe ctive s · predictive models are commonly based on historical experience/quantitative...

Documents