Why Mortgage Rates Went Up After the Fed's Big Cut

Episode Summary

Executive Summary: This Odd Lots episode explains why mortgage rates do not move one-for-one with Fed cuts. Guest Tom Graff argues mortgages embed prepayment risk and negative convexity, so investors demand a spread over Treasuries. The conversation covers who buys mortgage bonds, how G-fees and loan pooling shape rates, why spreads stayed wide, and what could finally pull mortgage rates lower.

Main Topics: Why Fed cuts don’t mechanically lower mortgage rates (Priority: 5/5): Tracy and Joe open by noting mortgage rates rose even after the Fed cut rates, emphasizing that mortgage borrowing costs are not directly tied to the policy rate and often move on expectations rather than the actual cut date. Mortgage bonds vs. Treasuries: prepayment risk and negative convexity (Priority: 5/5): Tom Graff explains that mortgage investors face prepayment risk: when rates fall, homeowners refinance and investors get principal back early, while rising rates offer little offset. That embedded option requires extra yield over Treasuries. Who buys mortgage-backed securities (Priority: 4/5): The discussion outlines the main buyers of mortgage bonds: banks, financial institutions, money managers, and mortgage REITs. Banks still matter because mortgages have favorable capital treatment, though the flat yield curve makes the trade less attractive. How mortgage rates are set in practice (Priority: 5/5): Graff describes the components of a quoted mortgage rate: the Fannie/Freddie guarantee fee, market execution value, borrower credit profile, down payment, and mortgage type. He also explains conforming vs. non-conforming loan limits. Why certain mortgage pools command premiums (Priority: 4/5): The episode explains that banks pool loans for best execution and can get higher prices for loans with slower prepayment behavior, such as New York mortgages that are less likely to refinance because of transfer taxes. Why the spread over Treasuries has stayed wide (Priority: 5/5): Graff attributes the elevated mortgage-Treasury spread to negative convexity, a flat yield curve, higher interest-rate volatility, and a frozen housing market that traps borrowers in low-rate or high-rate mortgages. Market anticipation, refinancing behavior, and future rates (Priority: 5/5): Rates move on anticipated Fed policy and macro conditions. The guest says mortgage rates could fall if Treasury yields drop due to weaker growth and more Fed cuts, but less volatility and less negative convexity are needed for a durable narrowing of spreads.

Key Arguments: Mortgage rates are not mechanically linked to Fed funds cuts; they respond more to expectations for future rates, economic growth, and volatility. Mortgage-backed securities must compensate investors for prepayment risk because borrowers can refinance without penalty when rates fall. Banks, money managers, and mortgage REITs still buy mortgage bonds, but a flat yield curve reduces banks’ incentive to hold them. Quoted mortgage rates reflect more than benchmark yields: guarantee fees, borrower credit, down payment, loan conforming status, and market demand all matter. Loan pools with slower refinancing behavior can fetch premiums, as shown by New York mortgages with transfer-tax-driven refi resistance. The unusually wide spread between mortgage rates and Treasuries reflects negative convexity, high rate volatility, and market concern about refinance behavior. Mortgage rates generally decline when the 10-year Treasury falls, which usually requires weaker economic data and/or more Fed easing priced into markets.

Data Points: Fed rate cut: 50 basis points - The September 18 cut discussed at the start of the episode Fed policy rate upper bound after cut: about 5.0% - Tracy and Joe reference the post-cut level after the September meeting Previous Fed policy rate upper bound: about 5.5% - Referenced as the level before the 50 bp cut 30-year mortgage rate: about 6.6% to 6.9% - Joe and Tracy note mortgage rates rose after the cut, nearing 7% 30-year Treasury yield: 4.32% - Joe uses this to ask why mortgage rates can’t simply track the government’s borrowing cost Facet AUM: $4 billion - Tom Graff’s current firm scale Mortgage market share in Bloomberg Aggregate: about 30% - Graff says mortgages are a major component of the benchmark bond index Rate-cut date: September 18 - Referenced as the meeting after which mortgage rates still rose Mortgage rate spread comment: about 150 basis points historically - Graff says historical Treasury-to-mortgage spreads are often around this level Potential spread improvement from lower volatility: 20-25 basis points - Graff estimates reduced rate volatility could compress mortgage rates by this amount Mortgage app activity: down in the last couple of weeks - Opening segment cites lower applications and refis amid rising mortgage rates Live event date: November 4 - Promo for Odd Lots live recording the night before the U.S. election

Pivotal Quotes: "If interest rates fall, everyone just refinances. I just get all my money back at par. I’m no better off. And so you got to get paid for that." — Tom Graff: Explaining prepayment risk and why mortgage bonds yield more than Treasuries "It’s a little like doing a covered call strategy in a stock." — Tom Graff: Analogy used to explain the mortgage investor’s payoff profile "Nothing mechanical happened on September 18th that just like made cost borrowing. And everyone knew September 18th or that a Fed cut was eventually coming." — Joe Weisenthal: Summarizing why the Fed cut did not instantly reduce mortgage rates

Implications: Listeners should not expect Fed cuts to instantly translate into cheaper mortgages. Future mortgage rates depend more on Treasury yields, rate volatility, refinancing incentives, and housing-market dynamics than on the policy rate alone.