Mortgage rates move to highest level in 5 weeks, but homebuyers shake it off

Mortgage applications rose 1.7% last week as purchase applications increased 4% week over week and 7% from a year earlier, suggesting buyers are adapting to higher rates. The average 30-year fixed mortgage rate edged up to 6.46% from 6.45%, while refinance applications fell 1% and the refinance share slipped to just over 40%, the lowest since July 2025. Rates have since moved another 14 bps higher this week amid renewed Iran-war uncertainty and a hotter-than-expected CPI print.

Analysis

The key signal is not the weekly uptick itself, but that demand is holding even as affordability remains strained and rates have started to re-accelerate. That implies a near-term release of pent-up household formation rather than a durable easing in financing conditions, which usually means transaction volume can stay resilient for a few more weeks even if rates move another 25-50 bps higher. The second-order effect is that sellers and builders may become more confident on pricing and starts into the spring selling season, but that confidence can reverse quickly if the rate move persists through month-end.

The more interesting winner is not necessarily homebuilders, but rate-sensitive housing intermediaries and suppliers with operating leverage to transaction volume: mortgage originators, title/escrow, and home-improvement chains tied to move-in activity. If purchase apps keep outpacing refis, mix shifts away from refinance-heavy lenders and toward purchase-focused platforms, which tends to favor better credit quality and lower rate-locked pipeline risk. On the flip side, refi-dependent servicers and lenders with longer-duration hedges may see margin pressure if rates grind up while origination pull-through weakens.

The contrarian point is that the market may be overestimating how much a modest demand bounce can offset the affordability shock. At roughly current levels, each additional 25 bps in mortgage rates typically trims borrowing power enough to force either smaller loan sizes or more concessions, so the real test is whether applications remain positive after the next CPI and rate reset. If geopolitical headlines or inflation keep rates elevated for 2-4 more weeks, the recent improvement likely becomes a front-loaded spring-pull-forward rather than a true demand inflection.

For macro positioning, the move argues for caution on broad housing beta until we see whether rates stabilize; the data support a tactical trade, not a structural one. In equities, the cleaner expression is a relative-value long in purchase-volume beneficiaries versus rate-sensitive homebuilders, with options used to define downside if rates reprice higher again. The risk/reward skews best over a 1-3 month horizon because that is when pipeline conversion, seller behavior, and builder guidance will reveal whether the demand recovery is real or just rate-resilient noise.

Key Decisions for Investors

• Long RKT / short XHB for 4-8 weeks: express a view that purchase-driven mortgage activity and housing transaction services outperform builders if rates stay elevated; stop if the 30-year mortgage rate falls back below the recent range and refi volume re-accelerates.
• Buy LEN or DHI on a 2-3 month pullback only if weekly mortgage apps remain positive despite rising rates; otherwise fade rallies in homebuilders because affordability deterioration will show up in cancellations and incentives before it hits reported sales.
• Long Z or PGR? No direct housing proxy in data; instead consider long RDFN calls vs short regional bank exposure tied to mortgage banking for 1-2 months, betting that purchase demand improves sentiment for transaction platforms while refi mix compresses lender economics.
• For macro hedging, buy short-dated puts on XHB into any further 15-25 bps rate spike over the next 2 weeks; the housing complex is vulnerable to a fast sentiment reversal if CPI or geopolitics push rates higher again.
• Avoid chasing mortgage originators with heavy refinance mix until refi share stabilizes for at least 2 consecutive weekly prints; the current setup favors volume, not margin durability.