Current Mortgage Rates March 2026

So Freddie Mac dropped their weekly survey on March 12th and the 30-year fixed came in at 6.11%. Up from 6.00% the week before. Not a dramatic move but enough to notice—especially if you've been watching this thing inch around for months waiting for a dip that keeps not coming.

Here's the full snapshot as of mid-March 2026:

30-year fixed: 6.11% 15-year fixed: 5.50% 5/1 ARM: 5.56% (APR around 6.00%) 7/1 ARM: roughly 5.80-6.19% depending on lender Jumbo 30-year: 6.34% Jumbo 15-year: 5.85%

For context—a year ago the 30-year was sitting at 6.65%. So yeah, rates are down almost half a point year-over-year. Not the dramatic drop people were hoping for when the Fed started cutting rates in late 2024, but it's movement.

The 15-year at 5.50% is the sleeper play here, tbh. If you can stomach a higher monthly payment, the interest savings over the life of the loan are obscene. On a $400,000 loan, going 15 vs 30 saves you something like $180,000 in interest. That's a real number.

This is the part most rate articles skip, which is annoying because it's the part that matters.

Mortgage rates don't follow the Fed funds rate directly. They track the 10-year Treasury yield. And the 10-year moves based on what bond traders think inflation will look like over the next decade, plus any flight-to-safety dynamics, plus whatever weirdness is happening with fiscal deficits and Treasury supply.

Right now the tension is: the Fed has been cutting short-term rates, but the 10-year has been stubbornly high because bond markets aren't convinced inflation is fully contained. Core PCE came in hot in January. Labor market is still tighter than the Fed wants. So the long end of the curve isn't cooperating the way buyers were hoping.

And then there's the mortgage spread problem. Normally mortgages run about 150-175 basis points above the 10-year. Right now the spread is wider than that—closer to 200-250 bps. Banks are being extra cautious on pricing because prepayment risk is high (everyone would refi if rates drop meaningfully). Until that spread normalizes, mortgage rates will feel sticky even as Treasuries move.

The good news, sort of: most forecasters think we see 6.00% or slightly below on the 30-year by late 2026. The bad news: nobody thought rates would still be at 6%+ this deep into the cutting cycle, so take those forecasts with appropriate skepticism.

ARMs are interesting again. The 5/1 ARM at 5.56% vs the 30-year fixed at 6.11% is a 55 basis point spread. On a $500,000 loan that's roughly $275/month in payment difference.

So the math question is: will rates be lower in 5 years? If yes, you save money during the fixed period AND potentially benefit when it adjusts. If rates are higher in 5 years, you're exposed.

My honest read: ARMs make sense right now if you know you're moving or refinancing within 5-7 years. If you're buying your forever home and planning to stay 20+ years, the fixed rate certainty is probably worth the premium. But the 7/1 ARM in particular is worth a hard look—seven years is a long time, and you'd have seven years of savings plus the option to refi if rates drop.

One thing people forget about ARMs: the caps matter. Most have a 2/2/5 cap structure—meaning the first adjustment can't go more than 2% above your initial rate, subsequent adjustments can't go more than 2% at a time, and the lifetime cap is 5% above start. So a 5/1 ARM starting at 5.56% can't ever exceed 10.56%. That's uncomfortable territory but it's not unlimited exposure.

Jumbo ARMs are where this really gets interesting. Jumbo fixed at 6.34%, jumbo 5/1 ARM at roughly 5.75%. On a $1.2M loan that difference is $500+/month. A lot of high-balance buyers are going ARM for this reason.

Rates vary by lender—sometimes by more than you'd think. I've seen half-point differences on identical loan scenarios depending on who you call. So shopping matters.

Rocket Mortgage is still the volume leader. They're not always cheapest but their ONE+ program (1% down, Rocket covers 2% as a grant) is legitimately useful for first-timers. Closing time averages 22 days which is solid. App is genuinely good.

Better Mortgage is the one to check if you want fast and cheap. Their online-only model keeps costs down and they've been consistently competitive on rate. Their no-fee approach saves borrowers real money at closing.

Chase is worth it if you already bank there—they discount rates for existing customers and their DreaMaker product (3% down) has had income limits lifted in 15 major metros. The rate transparency on their website is better than most big banks too.

LoanDepot has been aggressive on VA loans in particular. Credit score minimum of 520 for VA is legitimately the lowest I've seen across mainstream lenders. And their policy of waiving lender fees on future refis is nice if you think you'll refi in a few years.

For jumbo specifically: consider local banks and credit unions. They often portfolio their jumbo loans (hold them on balance sheet) rather than selling to secondary markets, which means more flexibility on qualification and sometimes better rates. Worth a call to your regional bank before assuming the online lenders win.

Navy Federal Credit Union—if you're eligible (military, DoD, family)—is consistently among the best rates across the board. Their VA loan pricing in particular is hard to beat.

I hate wishy-washy takes on this so I'll be direct.

If you're closing in the next 30-45 days: lock now. You know what rate you're getting. The downside risk of floating (rates go up, you pay more) is worse than the upside (rates drop a bit, you save a little). The mortgage market is not a place to gamble when you're actually buying a house.

If you're closing in 60-90 days: harder call. You can get a float-down lock with some lenders—they lock your rate but let you float down once if rates drop more than 0.125%. Costs a little more (maybe 0.125 to 0.25 in points) but provides insurance. Worth asking about.

If you're just shopping and not in contract yet: don't stress about the rate right now. Focus on finding the house. The rate environment in 30-60 days when you're actually closing will be what it is. You can't perfectly time this.

One thing I've noticed: buyers who obsess over rates to the point of paralysis end up paying more in appreciation. A house that costs $450,000 today at 6.11% might be $465,000 in six months even if rates drop to 5.75%. The math doesn't always favor waiting.

The wildcard scenarios where you'd want to float: a significant deterioration in economic data (recession signals, weak jobs reports) could push the 10-year down fast and pull mortgage rates with it. If you're seeing labor market cracks, that's when floating makes more sense. Right now? The economic data is too mixed to bet on a big rate move in either direction in the near term.

Quick history for context.

September 2025: 30-year averaged around 6.20% October 2025: Crept up toward 6.40% briefly on hot CPI data November-December 2025: Pulled back to 6.00-6.10% range January 2026: Settled near 6.06% February 2026: Quiet, mostly 5.95-6.05% March 2026 (to date): Bumped back up to 6.00-6.11%

The pattern here is sideways with chop. Not crashing, not spiking. Just grinding around in a 5.90-6.40% range for months. Which is frustrating if you're waiting for something dramatic to happen. But it also means if you need to buy, you're not in a terrible environment by historical standards. The 50-year average on the 30-year fixed is about 7.7%. We're well below that.

The wild card going into the rest of 2026 is tariff-driven inflation. If trade policy pushes goods prices higher, that's inflationary—bad for rates. If it slows growth—potentially good for rates as a flight to safety. Economists are genuinely split on which effect dominates. That uncertainty is probably keeping the bond market, and by extension mortgage rates, in their current holding pattern.