What are today's mortgage interest rates: March 26, 2026?

Average 30-year mortgage purchase rate is 6.25% and the 15-year purchase rate is 5.75% as of Mar 26, 2026; average 30-year refinance rate is 6.87% and 15-year refi is 6.02%. Rates have ticked up from the mid-5% range seen in mid-February after the Fed paused rate cuts following disappointing unemployment and inflation releases, though they remain below year-ago levels. Borrowers are advised to shop lenders, consider mortgage points, larger down payments or alternative terms (e.g., 20-year) to secure lower effective rates; best pricing is reserved for highest credit scores.

Analysis

Higher mortgage rates are not just a borrower pain point — they re-price the economics across the housing ecosystem and the fixed-income curve. Expect a near-term drop in transaction volumes (mortgage originations and home sales) that compresses builder revenues and realtor fee flow, while simultaneously pushing mortgage product mix toward higher-down-payment purchase loans and away from rate-sensitive cash-out refis. On the balance sheet side, bank NIMs can widen if deposit betas stay low, but origination pipelines and servicing fee income will shrink, creating a mixed earnings profile over the next 2-6 quarters.

Second-order effects concentrate in MBS mechanics and prepayment uncertainty: slower prepayments lengthen duration and increase convexity losses for long-duration holders if rates later resume downward momentum, but they also make newly issued coupons more attractive to reinvestors. The key macro catalysts are CPI/employment prints and Fed messaging over the next 3 months — either could re-price the front end by 50–75bp and flip the trade. Also watch deposit competition (regional banks vs fintech funding), which can reverse any NIM gain within 1–2 quarters if deposit betas rise unexpectedly.

Consensus is underweighting the asymmetry between purchase and refinance demand: purchase mortgage spreads are stickier and less rate-sensitive, so select mortgage-credit exposures (servicers with durable seller/servicing agreements, or REITs that can re-deploy cash into higher coupons) can re-rate before broad housing recovery. Conversely, homebuilder equities appear to still price in normalization that requires sub-5.5% 30yr rates; that creates a concentrated downside if rates remain >6% for several quarters.

Actionable posture: favor capital-lite financials that benefit from wider yield curves and selectively harvest MBS carry while hedging duration. Short rate-sensitive real-economy names (builders, certain consumer lenders with high prepayment optionality) and use curve trades and TBA structures to express directional or relative-value views with defined loss thresholds.

Key Decisions for Investors

• Short DHI and PHM via 3-month put spreads (e.g., buy 1x 3m 15% OTM put / sell 1x 3m 25% OTM put) — target 20–35% downside in equity value if mortgage rates stay >6% for 3–6 months; stop-loss: exit if 30yr mortgage rate falls below 5.8% for two consecutive weekly prints.
• Pair trade: long JPM (6–12 month horizon) / short DHI (6–12 month) — long JPM to capture NIM expansion and fee diversification, short DHI to capture volume risk; size 1:1 sector-dollar neutral, target 15–25% relative outperformance, stop if broad rates curve flattens by >40bp.
• Long agency MBS 30yr TBA (selected coupons) + short equivalent-duration 10y futures (steepener basis) — horizon 3–9 months to capture spread widening and carry if prepayments slow; risk: sudden rate drop causing prepayment surge—use a 25–50% hedge via interest-rate swaptions.
• Long AGNC or NLY with downside protection (buy equity + 6-month ~20% OTM put) — play reinvestment into higher coupons and dividend carry while limiting tail risk from NAV shocks; target 8–12% YTD total return net of hedges, stop-loss on NAV decline >25%.
• Directional curve trade: buy 10y futures / sell 2y futures (6–12 month) — tactical steepener to capture market pricing of eventual Fed easing; target 40–80bp steepening profit, stop if 2y yield falls relative to 10y by >30bp in one month (signals front-end capitulation).