MBS市场评论：没有浮动太大（全文完整）

发布时间:2022-07-03 20:45:02

下面是小编为大家整理的MBS市场评论：没有浮动太大（全文完整）,供大家参考。

　MBS

　Market

　Commentary:

　No float

　too

　big

　March

　27,

　2020

　North America Securitized Products Research 27

　March

　2020

　 The FOMC gave the New York Fed the mandate to conduct “unlimited” QE4 operations to purchase agency mortgages  Over the past two weeks, we estimate that around $250bn of agency MBS has been absorbed; this pace is unprecedented  We turn to past QE operations, QE1 and QE3 to gauge the potential purchase amounts going forward  We project $1.8tn of gross MBS issuance in 2020; the Fed could purchase up to $1.3tn of mortgages to relieve stress in the financial system  QE1 and QE3 brought mortgage spreads to around 20bp of Treasury OAS, this is

　what

　think

　QE4

　could

　ultimately

　achieve;

　this

　approximately

　50bp tighter than Thursday’s close  TBA dollar rolls have exploded higher with implied financing rates dropping by 100-200bp; much

　was due to the

　expected deliverable

　shift

　in TBA; mortgage cashflows should improve and so should TBA performance; be overweight MBS  We

　review

　the potential

　float

　improvements across

　the

　UMBS

　stack

　from

　the Fed’s

　purchases

　date in

　detail,

　and

　provide a model

　based

　approach

　for

　the scope of improvement additional purchases could generate  We sift through the bonds that the Fed was delivered through Wednesday this week; the collateral was largely non-spec and low payup  In

　this

　piece,

　walk

　through

　the

　risk

　that

　widespread

　forbearance

　poses

　to servicers,

　but

　reiterate

　that

　from

　the

　MBS

　investor’s

　perspective,

　the

　explicit Ginnie wrap and the Treasury lines of support to the GSEs are ample protection for timely payment of P&I

　 Securitized

　Products

　Research Brian

　(1-212)

　834-3128

　brian.ye@jpmorgan.com Nicholas

　Maciunas (1-212)

　834-5671

　nicholas.m.maciunas@jpmorgan.com

　Alex

　Kraus

　(1-212)

　834-5954

　alexander.d.kraus@jpmorgan.com

　J.P.

　Morgan

　Securities

　LLC

　See page 19 for analyst certification and important disclosures.

　www.jpmorganmarkets.com

　MBS Market Commentary

　  The FOMC gave the New York Fed the mandate to conduct “unlimited” QE4 operations to purchase agency mortgages Over the past two weeks, we estimate that around $250bn of agency MBS has been absorbed; this pace is unprecedented We turn to past QE operations, QE1 and QE3 to gauge the potential purchase amounts going forward We project $1.8tn of gross MBS issuance in 2020; the Fed could purchase up to $1.3tn of mortgages to relieve stress in the financial system QE1 and QE3 brought mortgage spreads to around 20bp of Treasury OAS, this is what we think QE4 could ultimately achieve; this is approximately 50bp tighter than Thursday’s close TBA dollar rolls have exploded higher with implied financing rates dropping by 100-200bp; much was due to the expected deliverable shift in TBA; mortgage cashflows should improve and so should TBA performance; be overweight MBS We review the potential float improvements across the UMBS stack from the Fed’s purchases to date in detail, and provide a model based approach for the scope of improvement additional purchases could generate We sift through the bonds that the Fed was delivered through Wednesday this week; the collateral was largely non-spec and low payup In this piece, we walk through the risk that widespread forbearance poses to servicers, but reiterate that from the MBS investor’s perspective, the explicit Ginnie wrap and the Treasury lines of support to the GSEs are ample protection for timely payment of P&IViews

　 Be overweight MBS Overweight 30-years versus 15s Overweight Ginnie production coupons

　Week two of QE4 finally brought some calm to the rate and mortgage markets. Instead of daily price swings measured in points, Treasury and mortgage price changes started to be measured in ticks toward the end of the week. We are optimistic that this is the beginning of the recovery process. Treasury and agency mortgages are the primary transmission mechanism for the Fed to relieve the broader stress points in the system.

　 Shock and awe: by the numbers

　Exhibit 1: Unprecedented: Mortgage QE versus mortgage supply Brief

　summary

　Fed

　MBS

　operations

　including

　QE4

　thus

　far:

　gross

　and

　net

　amounts

　compared

　MBS

　gross supply

　Gross Net Gross MBS Regime Date Range Purchase

　 Purchase Supply

　Addl. Info QE1 MBS QE Nov

　2008 - Aug 2010 $ 1,350 $1,119 $1650 in 2009 Nov 2008: $600 announced, M ar 2009: addl $750 announced QE2 Treasury

　QE Nov

　2010 - Aug 2011

　-$166

　Treasuries added: $589 QE3 MBS + Treasury QE Sept 2012 - Oct 2014 $1,400 $868 $1492 in 2013 QE3 tapering from Dec 2013 to Q4 2014 QE4 MBS + Treasury QE Mar 13 - tbd $250 $250 $1507 in 2019; $1800 in 2020 Figure during the first 2 weeks Source:

　J.P.

　Morgan

　 This calming effect came on the back of extraordinary measures by the Fed. QE4 was announced barely two weeks ago and during this short period, the Fed has already bought over $250bn agency MBS (along with roughly double that in Treasuries). In Exhibit 1, we compare QE4 thus far to the entireties of the 2008-2009 QE1 and the 2012-2014 QE3. QE2 was mainly Treasuries only, except for some paydown reinvestments, thus we are leaving it out of the table. During both of the aforementioned prior QE operations, the Fed bought around $1.4tn of mortgage assets, each lasting from one to two years in duration. Presently, it has already accumulated roughly 20% of those purchases in two weeks. The scale of Fed’s purchases is unprecedented. Over the past week, the Fed conveyed to the market its intent to purchase $50bn MBS on a daily basis (practically unlimited). The amount acquired has fallen short of the maximum. We estimate that the Fed has purchased around $250bn of mortgages over the past two weeks since the March 13 th announcement. This massive scale has relieved pressure points in the system.

　Gross supply forecast: $1.8tn

　Faced with a mandate to purchase an “unlimited” amount of mortgages, how much can or should Fed purchase in the weeks and months ahead? We turn to past mortgage QEs for guidance and compare them to mortgage gross production. Our idea is to compare purchases to gross supply to calibrate the potential amount of total Fed purchases. The gross supply spanning the time period of peak QE1 and QE3 episodes was approximately $1.5tn in 2009 and again in 2013. During these two refi waves, mortgages paydowns averaged 19c in 2009 and 20c in 2013. Mortgage paydowns averaged 14c in 2019. To be fair, the refi wave didn’t start in earnest until last summer. The average prepayment speed in 2019 H2 was 18c. This is gives us a starting point to project 2020 paydowns and gross supply.

　The gross supply of agency MBS was $1.5tn in 2019. How much higher will paydowns in 2020 be compared to H2’2019? Assuming that average paydowns in 2020 are 10-20% higher than in H2’2019, the 2020 gross supply would end up in excess of $2tn. In our view, the changing landscape in mortgage origination will moderate prepayments and slow the pace of mortgage supply. Social distancing and widespread lockdowns are slowing down new home purchases and housing turnover. Cash-out refis are harder to obtain and the recent pace of home price appreciation may not be sustainable. Therefore, our base case forecast for this year’s gross supply is $1.8tn of supply.

　 Demand shock: $1.3tn of MBS purchases?

　The Fed purchased between 60%-80% of gross production during the peaks of QE1 and QE3. Matching these prior episodes, the Fed may end well end up buying $1.3tn mortgages, basically matching the amounts purchased in QE1 and QE3. These prior episodes took well over a year to execute. The recent pace of QE is unsustainable, in our view.

　Excess supply in the system has been relieved. We wouldn’t be surprised if the NY Fed scales back its purchase allowances in the weeks ahead. Its goal is to stabilize markets and to let private investors allocate capital, instead of crowding them out. The key take away from the past two weeks is that the Fed is prepared to do whatever it takes to stabilize markets, first in Treasuries and mortgages, and then onto credit markets. In our view, given its unlimited mandate, the NY Fed will be dynamically calibrate its response to reduce purchases when necessary and increase its support should market stress comes back.

　Changing tack

　It is well known that during the later QE1 stage and then again in QE3, the Fed purchased mostly production coupons, which in today’s rate environment are 30-year 2.5s and 3s. The goal was to support the TBA market in primary production coupons, thus ensuring that the primary mortgage origination market can function. This time, the Fed changed tack. In response to the balance sheet pressure as a result of de- levering by mortgage investors, who mostly hold pools and whose positions largely mirror those of the Mortgage Index, the Fed expanded the purchase range to the

　entire tradable coupon stack. Just as importantly, these purchases are for T+2 and

　T+3 settle. As Exhibit 2 shows, approximately 40% of Fed purchases are for T+2/T+3 settle, translating to $105bn of pools off of investor and dealer balance sheets. Across mortgage sectors, roughly 3/4 th

　 have been in conventional 30-years with the rest about evenly split between Ginnies and 15-year conventionals (Exhibit 3). The seemingly low Ginnie share reflects the fact that most of Ginnie holdings are in the hands of US banks and overseas investors, the investor sectors that are the least under strain. It’s clear that the NY Fed has been nimble given its experience in agency MBS markets over the past 10 year. It also reflects where the stress points in the mortgage sector have been. Ginnie Maes and 15-years are held in stronger hands.

　Exhibit 2: 40% of the Feds QE4 purchases have been for short settle Share

　Fed

　purchases

　QE4

　for

　short

　and

　reg

　settle

　Exhibit 3: 30yr conventionals comprise the bulk of purchases Share

　Fed

　purchases