The Perverse Cycle Driving Mortgage Rates

The headline screams: mortgage rates hit 6.75%, bonds are in a bloodbath, inflation is surging, and the Fed is asleep. The mainstream narrative blames the usual suspects—inflation, a lax Fed, a flood of new Treasury debt. That story is not wrong, but it is incomplete. It misses the quiet, self-defeating mechanism buried inside the government’s own attempt to help the housing market.

The Overlooked Angle

Everyone sees the trifecta of inflation, Fed inaction, and supply glut. But the real operational farce is the unintended consequence of Fannie Mae and Freddie Mac buying back their own mortgage-backed securities (MBS) to narrow the spread between mortgage rates and Treasury yields. The idea sounds clever: increase demand for MBS, lower MBS yields, and thereby lower mortgage rates. What the architects forgot to ask is where the cash comes from.

Fannie and Freddie fund these MBS purchases by selling down their substantial Treasury holdings and diverting operational cash flow away from Treasuries. They become net sellers of Treasuries. Two of the largest institutional holders of U.S. government debt suddenly turn into sellers. The result? The 10-year Treasury yield gets pushed higher. Mortgage rates track the 10-year yield. So the intervention that aims to lower mortgage rates actually raises the very benchmark that mortgage rates are priced off. It is a loop of self-inflicted pain.

Why This Small Detail Matters

This is not a footnote. The spread between mortgage rates and the 10-year Treasury yield is a narrow technical line. Shaving 10 or 20 basis points off that spread by buying MBS is a victory if the 10-year yield stays flat. But if the buying of MBS forces the 10-year yield up by 30 basis points, the net effect is a higher mortgage rate. The mechanism is not merely neutral; it is counterproductive.

The Treasury market is the deepest in the world, but it is not infinitely elastic. When two enormous GSEs start selling billions of dollars of Treasuries to fund MBS purchases, they add supply to a market already straining under record deficit financing. The bond market’s response is mechanical: more supply demands higher yields to attract buyers. The very act of trying to compress the mortgage spread widens the base rate.

The Economic Mechanism

Let’s walk through the cash flows. Fannie and Freddie have two sources of funds for MBS buybacks: (1) their retained earnings and cash flows from existing MBS portfolios, and (2) the liquidation of their Treasury holdings. These GSEs held roughly $200 billion in Treasuries at the start of 2025. As they sell, those Treasuries enter the secondary market. The Treasury Department is simultaneously auctioning $30–$50 billion of new debt per week. The combined supply overwhelms demand at current yield levels.

The 10-year yield is the marginal price of money for the U.S. government. It sets the floor for all risk-free rates. Mortgage lenders add a spread on top for servicing costs, prepayment risk, and credit risk. If the GSE action reduces that spread by, say, 15 basis points but pushes the 10-year yield up by 25 basis points, the net mortgage rate change is +10 basis points. The housing market gets higher rates, not lower.

But the problem does not stop there. The MBS buyback itself also reduces the outstanding supply of MBS, which should lower MBS yields. However, the largest buyers of MBS—pension funds, insurers, foreign central banks—are also sensitive to Treasury yields. When Treasuries become more attractive due to higher yields, some capital that would have gone into MBS shifts to Treasuries. This substitution effect further reduces the demand for MBS, offsetting the GSE purchases. The net effect on MBS yields is ambiguous at best.

Meanwhile, the Treasury’s own borrowing costs rise. Higher 10-year yields mean higher interest expense on the $35 trillion national debt. That adds to the deficit, which requires even more debt issuance next quarter. The spiral feeds itself.

The Strategic Consequence

Who benefits from this scheme? Short-term, the MBS market gets an artificial bid that allows some mortgage originators to offload risk at slightly better prices. But those gains are fleeting. The real beneficiaries are the sellers of Treasuries—namely, Fannie and Freddie themselves—who get to book cash that they then use to buy MBS. For the GSEs’ shareholders (the Treasury, since they are under conservatorship), the net effect on portfolio value is near zero after accounting for the yield increase.

The losers are obvious: anyone taking out a mortgage. The 6.75% rate is not purely a function of inflation fears. It is partly a function of a policy that inadvertently raises the risk-free rate. Also losing are holders of long-term Treasury bonds who see prices fall as yields rise. The GSE selling amplifies that loss.

The Fed is also a loser in credibility. As it signals delay or cuts, the bond market does the opposite. The GSE program adds one more layer of confusion. The market sees a government that cannot coordinate its own agencies: the Treasury issues debt, the Fed talks dovish, and the GSEs sell Treasuries to buy MBS. The result is a cacophony that destroys trust in the anchoring of yields.

What Most Commentary Gets Wrong

Financial media and even some economists praise the GSE MBS buyback as a clever way to lower mortgage rates without explicit Fed intervention. They point to the narrowing of the spread as evidence of success. They miss the denominator. The spread can narrow while the base rate explodes. The commentary focuses on the output (mortgage rate relative to Treasuries) and ignores the input (absolute level of Treasuries). It is like celebrating a smaller gap between your car and the guardrail while your speedometer climbs to 100 mph.

Another lazy take is that the bond selloff is solely about inflation and Fed inaction. That is the easy story. But the mechanics of GSE funding are a material, incremental force. The Treasury market is a giant bathtub: every seller adds a drop. Two institutional sellers adding billions every month is not a drop; it is a bucket. The market prices in that flow.

Some argue that the GSEs are not large enough to move the 10-year yield. This is naïve. The GSEs hold hundreds of billions of Treasuries. The daily volume in the 10-year futures is about $200 billion. A few billion of net selling from the GSEs each month matters at the margin, especially when the Treasury is simultaneously auctioning $150 billion a month. Marginal supply determines price in a liquid market.

The Hard Business Lesson

There is a universal principle here: every intervention creates a funding signature. If you try to fix a price in one market, you must borrow from another market. The borrowed force will work against you unless you neutralize it. The Fed can lower mortgage rates by buying MBS directly and simultaneously sterilizing the effect on reserves or Treasuries. But the GSEs have no such tool. They sell Treasuries, which is the opposite of sterilization.

The lesson for any business strategist: before you launch a subsidy or intervention, trace the cash flow. If your funding source destabilizes a correlated input, you are running on a treadmill. The housing market is learning this the hard way. Mortgage rates are 6.75% today not just because inflation is 4% and the Fed is complacent. They are that high because the government’s own agencies are pushing up the cost of money while pretending to push it down.

The bond market is a brutal teacher. It punishes incoherence. The trifecta of inflation, soft Fed, and debt supply is bad enough. Adding a self-defeating GSE program on top is a masterclass in how not to manage a crisis. The only path to lower mortgage rates is not more clever gimmicks. It is a credible hawkish Fed, a credible fiscal consolidation, and a halt to the GSE Treasury sales. Until then, every basis point of spread compression will be stolen back by a rising base rate.

The real story is not the inflation number. It is the hidden feedback loop that turns a well-intentioned intervention into a rate spike.