Out-of-consensus calls in MBS for 2025

Chris Helwig and Brian Landy, CFA | November 22, 2024

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors. This material does not constitute research.

After years of relatively weak excess returns, fundamental and technical drivers appear poised to push MBS spreads tighter next year. The likely prospects for a smaller GSE footprint should lift MBS convexity as TBA quality improves and certain specified pool stories become more difficult to refinance. Demand from both depositories and asset managers should remain strong, especially if the yield curve continues to steepen and attractive relative value to other spread products continues to hold. A steep curve should keep a cap on agency supply, but look for an uptick in private label issuance as more loans flow from the GSE and government balance sheets to private capital.

A smaller footprint improves MBS convexity

The prospect of a smaller footprint for government-sponsored mortgage lending should improve convexity and value in affected pools. There are several plausible ways the next administration could do that:

• Investment properties and second homes

Pay-ups for investor and non-owner-occupied specified pools should increase if the GSEs stop acquiring these loans. The last Trump administration attempted to limit acquisitions of those loans in early 2021, but the Biden administration ended the effort. Investment properties and second homes accounted for about 7% of GSE lending in 2024; shifting this production to the private market would be an easy way to shrink the GSEs without adversely affecting the typical borrower. Existing investor and non-owner-occupied pools would have better convexity and higher pay-ups, as would existing non-agency deals backed by agency-eligible investor loans.

• Jumbo conforming loans

Pay-ups for jumbo conforming pools (Fannie Mae CKs, Freddie Mac T6/3P) should increase if the GSEs were to limit, or fully stop, acquiring jumbo conforming loans. These pools are not TBA-deliverable and typically trade at prices below TBA, but eliminating agency execution for those loans would improve convexity so they should trade at a smaller discount to TBA. The GSEs could outright suspend purchases of these loans, or sharply increase the jumbo conforming LLPA to a point that non-agency execution is better.

Ginnie Mae MJM pools would similarly benefit if the Federal Housing Administration and the US Department of Veterans Affairs no longer originated jumbo conforming loans. However, changing the VA loan limit would require Congressional action, which might be difficult to obtain. A simpler, although less extreme, solution would be for Ginnie Mae to no longer allow jumbo conforming loans to be placed in TBA-deliverable multiple issuer pools. Forcing those loans to custom pools would hurt execution for jumbo loans, slow production, and improve the convexity of existing MJM pools.

Jumbo conforming loans accounted for about 6% of GSE issuance and 5% of Ginnie Mae issuance in 2024. Scaling back these acquisitions would improve the convexity of new production generic pools, which should lower mortgage rates to new borrowers. That should be an added benefit to policymakers.

Removing jumbo conforming loans would also encourage innovations in the specified pool market. Many originators will not pool low pay-up stories separately since at times that production is used to offset delivery of jumbo conforming loans into the TBA.

• Low FICO pools

Convexity of low FICO pools should increase if the GSEs adjust LLPAs to reduce subsidization of low credit loans by high credit loans. This would slow prepayments of low FICO borrowers that would need to pay a higher LLPA on refinancing.

• FHA mortgage insurance premiums

Convexity of existing Ginnie Mae pools would improve if the administration were to raise FHA mortgage insurance premiums. The FHA’s insurance fund is currently well-capitalized, but higher premiums could be used as a blunt tool to shrink the FHA and normalize insurance pricing relative to private insurers. Higher insurance premiums would slow refinancing and turnover, but that more stable prepayment profile has better convexity and would improve the value of existing Ginnie Mae MBS.

Mortgage credit looks more levered to policy than fundamentals

Similar to MBS convexity, mortgage credit appears more likely to be influenced by policy rather than fundamentals next year. A smaller GSE footprint and subsequent growth in the supply of previously GSE-eligible loans should increase competition for existing capital supporting the private-label market. Investors, particularly those at the top of the capital structure, may skew allocations towards previously GSE-eligible loans as they may more closely track potential MBS tightening next year and, in the case of investor loans, may offer better convexity than other private-label exposures. Private-label investors will likely maintain a healthy allocation to non-QM as well as they may be viewed more so as a surrogate for investment grade corporate exposures, which are currently sitting near five-year tights than MBS. More novel, less liquid cohorts may underperform relative to more scalable exposures.

Changes to the GSE pricing grid for cash out refinance loans would likely be a tailwind to supply for closed-end second liens, especially against the backdrop of relatively elevated mortgage rates. An additional increase in LLPAs on these loans would have the combined effect of maintaining the lock-in effect on existing, lower-rate GSE first liens and driving up volumes of second lien originations as the combined rate on the two loans would look even more attractive than a single consolidation loan as LLPAs increase.

Absent barring delivery of high balance loans completely, originators would likely still deliver some jumbo conforming loans to the enterprises to the extent that they are below the TBA de-minimis pooling amount of 10%. Originations above those amounts would likely flow to private-label execution especially given the negligible differences in price drops relative to TBA for high balance agency pools relative to private-label pass-throughs with the same coupon.

A future decline in fees collected by the GSEs on higher margin loans may not be particularly disruptive to mission-based affordable housing mandates. Away from a fairly uniform guaranty fee for all originators irrespective of prior credit performance, higher LLPAs provide a valuable cross-subsidy to the enterprises to fund mission-centric initiatives such as affordable housing. While difficult to quantify, the existing affordable housing infrastructure has been taxed to some extent by the increased migrant population. Potential relief of existing stresses on that stock may alleviate some pressure to grow that stock further and afford a new FHFA director substantial latitude to let higher LLPA loans flow to private channels.

Other policy wildcards with regards to mortgage credit are the potential impact of policies implemented under the newly formed Department of Government Efficiency (DOGE). To date the Department’s leaders have focused on cuts across both expired appropriations and various Federal departments and functions. Per a CBO survey released in July over $500 billion was spent this year on programs whose Congressional authorizations have expired. Looking at the breakdown of those expenditures, it appears any cuts here would do little to impact housing or mortgage credit. In fact, less than 10% of expired or expiring appropriations are earmarked for the Department of Transportation, HUD and related agencies. Conversely, a material reduction in the Federal workforce could have implications for regional housing supply and home prices, particularly in the Mid-Atlantic region and other areas of the country with large concentrations of federal employees.

Net supply faces headwinds

Agency MBS net supply is likely to stay near historically low levels next year and could fall if the government acts to reduce the size of Fannie Mae, Freddie Mac, and Ginnie Mae. Lower net supply would likely push spreads tighter. Net supply is typically driven by a few factors:

• Flows of capital between agency and non-agency MBS
• Existing home sales, if the original and new loan have difference sizes
• New home sales

A smaller government role in mortgage finance could push net supply of agency MBS lower next year. For many years the first item has favored agency MBS supply since non-agency lending was relatively small. But if the government does act to shrink Fannie Mae, Freddie Mac, and Ginnie Mae, then loans that are no longer eligible for agency execution would be replaced with non-agency loans when the borrower sells the property or refinances. And a larger share of new production would no longer qualify for agency execution.

Higher projected home sales next year should have only a modest effect on net supply, increasing roughly 7%, all else equal. The Mortgage Bankers Association expects existing home sales to increase 5% next year, but the average loan size is expected to increase only 1%. New home sales are expected to increase about 10%, also with a roughly unchanged average loan size. But new home sales are a smaller portion of overall home sales, so that translates to only a 2% increase in net supply.

Agency net supply is currently near the lowest level since 2015 (Exhibit 1). A 7% increase from higher existing and new home sales would do little to alter that picture, and a shrinking GSE footprint could more than outweigh the effect on supply from home sales.

Exhibit 1. Net issuance of agency MBS (monthly average by year).

Source: Fannie Mae, Freddie Mac, Ginnie Mae, Santander US Capital Markets.

MBS demand likely to surprise to the upside

The demand side of the equation for MBS appears poised to push spreads tighter over the course of next year. A further steepening of the yield curve should spur greater net demand from depositories. Improving convexity and attractive value relative to the Treasury curve should signal strong demand from asset managers and hedge funds as well. Conversely, foreign investors may be net sellers of MBS if the new administration implements broad-based tariffs on imports.

US depositories have net added roughly $100 billion in MBS this year and may be poised to add substantially more in 2025, fueled by tepid loan demand, a further steepening of the yield curve and a less stringent regulatory regime. At their recent peak, MBS comprised roughly 13% of assets on depository balance sheets and fell to 10.8%, a fairly seismic change given a total asset base of just over $23 trillion. While it seems implausible that banks will restock their balance sheets with the same quantity of MBS they held in the past, there is likely still some more room to grow. A retracement back to the five-year average level of holdings implies banks could add an additional $150 billion in MBS absent any additional deposit growth.

Net demand from banks appears poised to surprise to the upside for a handful of reasons. First and foremost, residential mortgage loan demand remains weak, and banks are doing little to expand the credit box to increase volumes. Recent readings of the Federal Reserve Senior Loan Officer Opinion Survey (SLOOS) show, while modestly improved in recent quarters, banks continue to report weak demand for both QM jumbo and non-QM loans (Exhibit 1). Additionally, after a slight uptick in credit availability for non-QM loans last quarter, banks subsequently reversed course and tightened standards for non-QM loans while credit availability for jumbo loans remained broadly unchanged.

Exhibit 2: Demand for jumbo, non-QM loans remains weak at US depositories

Source: Santander US Capital Markets, Federal Reserve Senior Loan Officer Opinion Survey.

Banks will likely have incentives to add MBS in the face of weak loan demand if the yield curve continues to steepen. Both near- and longer-term observations suggest bank demand will continue to increase if the curve continues to steepen out. Over a 5-year horizon, MBS holdings across US depositories surged when the curve steepened and waned as the curve flattened, generally with some lagged effect.

The final piece of the puzzle when it comes to sizing up bank demand is the potential relaxation of the regulatory framework for US depositories. Further relaxation of the Basel III Endgame rules could generate 9% in surplus capital for G-SIBs and up to 4% for smaller banks.

In comments made in September, Federal Reserve Vice Chairman for Supervision Michael Barr provided guidance that under a revised Basel III Endgame proposal, Tier 1 capital requirements for global systemically important banks (G-SIBs) would still rise by roughly 9%. Smaller banks, broadly defined as between $100 billion and $700 billion in total assets could expect capital requirements to rise by an additional 3% to 4% over the intermediate term.  At the end of the third quarter of this year US G-SIBs held just over $1 trillion in Tier 1 common equity, meaning an additional $90 billion of surplus capital could be released if the Basel 3 Endgame rules were to be further curtailed. Non-G-SIB banks held slightly less than $550 billion in capital at the end of the third quarter, translating to roughly $20 billion in surplus capital that could be redeployed into loans or securities.

Asset managers look likely to add MBS next year as well. The complex of domestic mutual funds benchmarked to the Bloomberg US Aggregate index collectively hold roughly $2.5 trillion in total assets. They currently maintain a substantial MBS overweight with roughly 38% of total assets held in MBS versus an index weighting of 27% per the October release of the Bloomberg US Aggregate Fixed Income Indices Factsheet. The MBS overweight comes at the expense of a substantial underweight in US Treasuries.

With concerns underpinning Treasury performance given the prospects for increased supply given the extension of corporate tax cuts under the incoming Trump administration, MBS overweights across the fund complex appear poised to hold over the near term. Future fund inflows should be disproportionately allocated to MBS and there’s potential for managers to increase their overweight

One potential source of negative net demand for MBS may be from foreign investors. If the incoming administration were to engage in broad-based tariffs on imports it would likely spur a further rally in the dollar versus foreign currencies which, in turn, could spur selling of US dollar assets by foreign sovereign portfolios to defend the local currency. The decision amongst foreign sovereign portfolios to sell MBS or Treasuries will likely come down to whether increases in local inflation associated with a stronger dollar may be offset in part or whole by the rising value of exports sold to the US.

While lagged, the best source of data on foreign holding of MBS is the Treasury International Capital System (TIC). Holdings of MBS across portfolios domiciled in Asia peaked in 2020 at just under $1 trillion, with mainland China making up nearly a quarter of total holdings. As of last year, total holdings fell to roughly $875 billion while holdings across mainland China remained relatively flat (Exhibit 4). While still substantial and potentially a source of supply that the MBS market would have to absorb, foreign portfolios may choose to sell Treasuries rather than MBS as a first order move to defend their currencies if that is the preferred course of action.

Exhibit 3: MBS holdings across China, broader Asia

Source: Santander US Capital Markets, Treasury International Capital System