Agency mortgage-backed securities have long been treated as the safe, predictable corner of the fixed income universe – government-backed, liquid, and reliably bid. That reputation is getting stress-tested right now, as the Federal Reserve’s gradual retreat from the MBS market leaves a pricing gap that private buyers are only slowly filling.
How the Fed Became the Market – and What Happens When It Steps Back
During the post-2008 era and again through the pandemic period, the Federal Reserve accumulated an enormous portfolio of agency MBS as part of its quantitative easing programs. At its peak, the Fed held well over $2.7 trillion in agency MBS – a position so large it effectively set the floor for pricing across the entire sector. When the biggest buyer in any market is also largely price-insensitive, spreads compress. For years, that’s exactly what happened.
Quantitative tightening changed the calculus. As the Fed allowed its MBS holdings to run off without reinvesting principal payments, the volume of securities flowing back into the open market increased steadily. The Fed has not been actively selling, but passive runoff alone has been enough to shift the supply-demand dynamic in a market that had grown accustomed to a captive, permanent buyer absorbing a significant share of new issuance and prepayment proceeds.
The mechanics matter here. When agency MBS borrowers prepay – whether through refinancing or home sales – the principal returns to MBS holders. For years, the Fed recycled those payments back into new agency MBS purchases, effectively keeping that capital in the sector. Once reinvestment stopped, that capital started looking for a home elsewhere. For private investors, this is both an opportunity and a source of uncertainty: spreads are wider, but the price discovery process is messier than it used to be.
Current spread levels on 30-year current coupon agency MBS relative to Treasuries have widened noticeably from the tighter ranges that prevailed when Fed reinvestment was active. The widening isn’t catastrophic by historical standards, but it represents a meaningful repricing of what was previously one of the most compressed spread sectors in investment-grade fixed income. For investors who bought into the sector expecting Fed support to persist indefinitely, the adjustment has been uncomfortable.
Who Is Buying – and Why the Replacement Demand Hasn’t Fully Arrived
The traditional buyers of agency MBS outside the Fed are commercial banks, insurance companies, and dedicated fixed income funds. Each group has its own constraints, and right now, several of those constraints are working against aggressive MBS accumulation. Banks, for instance, took significant unrealized losses on their MBS portfolios when rates rose sharply, and many are managing those positions cautiously rather than adding exposure. The accounting pain from mark-to-market losses has made MBS a politically and operationally sensitive asset class inside bank treasury departments.
Insurance companies remain active, but their appetite is calibrated to liability-matching needs rather than opportunistic spread capture. They buy what they need based on duration requirements and regulatory capital frameworks – not because spreads look attractive in isolation. That kind of demand is steady but not elastic. It doesn’t surge when spreads widen, which means it can’t fill the void left by a price-insensitive central bank buyer.
Real money fund managers – the mutual funds and separately managed accounts that buy on behalf of pension funds, endowments, and retail investors – are more spread-sensitive, and there’s genuine interest from that cohort as current coupon MBS yields push into more attractive territory. The problem is that these investors also face competition from other spread products. Callable bond ladders and investment-grade corporate credit have been absorbing investor attention in an environment where rate-cut expectations keep getting pushed out, leaving MBS to compete for a finite pool of fixed income capital.
The prepayment uncertainty layered on top of spread volatility makes the MBS valuation exercise genuinely difficult. Unlike a corporate bond with a fixed maturity, an MBS pool’s duration shifts as interest rates move – extending when rates rise and shortening when rates fall. This convexity risk requires sophisticated hedging, and in a volatile rate environment, the cost of that hedging eats into the spread advantage that wider levels theoretically offer. For generalist fixed income investors without dedicated MBS analytical infrastructure, the sector looks cheap on paper but complicated in practice.
Overseas central banks, historically a meaningful source of agency MBS demand, have also been less active buyers in recent years. Some have trimmed their U.S. fixed income holdings as dollar reserve management strategies have evolved. Others simply have fewer dollars to recycle given current account dynamics. Whatever the reason, that source of external demand is thinner than it was during the peak years of global reserve accumulation.
What Wider Spreads Actually Mean for Investors Watching the Sector
For investors with the duration tolerance and the analytical capability to navigate prepayment modeling, the current spread environment in agency MBS is more interesting than it has been in several years. The yield pickup relative to comparable-duration Treasuries is real, and the government guarantee on principal removes credit risk entirely from the equation. The question isn’t whether the bonds are safe – it’s whether the compensation for convexity risk and duration uncertainty is adequate given where rates might go from here.
The more uncomfortable question for the broader market is structural: if the Fed isn’t reinvesting, banks are cautious, overseas demand is soft, and real money accounts are only selectively interested, who absorbs MBS supply at scale on an ongoing basis? Mortgage originators can’t hold everything they produce. The market needs a clearing mechanism, and right now, that mechanism is price – spreads wide enough to attract the marginal buyer. How wide is wide enough is exactly the debate playing out in fixed income desks right now, with no settled answer in sight.
