Jumbo refi. Different game, different rules.

It seems counterintuitive — bigger loan, lower rate? — but it happens regularly for prime-tier borrowers. The reason: banks compete fiercely for high-balance, high-FICO portfolio customers. A $1.5M jumbo loan generates more interest income than a $400K conforming loan, and the borrower (typically) has higher net worth and lower default risk.

Banks especially want this customer because the relationship can extend to investment management, private banking, business lending, and other revenue streams. That's why depository banks offer "relationship pricing" — they're willing to discount the mortgage rate to win the broader relationship.

This effect is strongest at top-tier (740+ FICO, 65%-or-lower LTV, 12+ months reserves). At standard tier, conforming and jumbo rates roughly match. Below standard tier, jumbo costs more — sometimes much more — because there's no agency backstop to spread the risk.

Less than conforming, generally. Most jumbo lenders cap cash-out LTV at 70-75%, vs the 80% you can typically get on conforming cash-out. Some lenders go to 80% on jumbo cash-out for prime-tier borrowers, but this is the exception.

The math: on a $2M home, conforming cash-out at 80% LTV (if it qualified) would let you pull up to $1.6M total loan size. Jumbo cash-out at 75% caps you at $1.5M total loan size. That $100K cap difference matters.

Multi-unit jumbo cash-out is even more restricted — typically 65-70% LTV cap. And some lenders won't do jumbo cash-out at all on second homes or investment properties.

If you need cash, also evaluate cash-out refinance options versus a HELOC layered onto your existing low-rate first lien — sometimes that's the better structure.

Higher in absolute dollars, similar as a percentage. Title insurance, appraisal, recording fees, and lender origination scale roughly with loan size. A $500K conforming refi might run $4,000-$6,000 in closing costs; a $1.5M jumbo refi might run $8,000-$15,000.

Two real differences to know about:

1. Appraisal complexity. Jumbo loans on luxury properties often require two independent appraisals, especially when loan amount exceeds $2M. This adds $800-$1,500 vs a single appraisal.

2. Title-insurance pricing tiers. Title insurance scales with loan amount, but the rate per thousand drops as loans get larger. So a $2M loan's title insurance is much less than 4× a $500K loan's title insurance — closer to 2.5×.

The break-even math (closing costs ÷ monthly savings) is what actually matters. Higher closing costs are fine if monthly savings are correspondingly higher. A 1% rate drop on a $1.5M loan generates ~$1,000/month in savings — meaningful even against $12K closing costs (12-month break-even).

A meaningful difference in the jumbo market — moreso than in conforming.

30-year fixed jumbo is the most common product but tends to price at a premium of 0.25-0.50% over equivalent ARMs. The premium pays for rate certainty across 30 years.

7/1 ARM jumbo (fixed for 7 years, then adjusts annually) often prices below conforming 30-year-fixed for prime-tier borrowers. This is the rate-arbitrage opportunity that gets sophisticated jumbo borrowers excited.

10/1 ARM jumbo (fixed for 10 years) splits the difference — slight rate premium over 7/1 but more years of certainty.

The right choice depends on how long you'll keep the home. If you're confident you'll move or refi within 7 years, the 7/1 ARM is mathematically dominant for most prime borrowers. If you might keep the loan 15+ years, the 30-year-fixed protects against rate volatility you can't predict.

Common mistake: assuming "ARM = risky" without doing the math. For loans you'll have less than the fixed period, ARMs are strictly better than fixed in expectation. They become risky only if you keep them past the fixed window.

RSU and equity-compensation income is one of the trickiest underwriting situations in mortgage banking. Jumbo lenders' approaches vary dramatically.

Conservative approach (typical big-bank): use only base salary plus a 2-year average of cash bonus. RSUs and stock comp are ignored entirely or heavily discounted. This often leads to DTI failure even for borrowers who've made $500K+/year for years.

Standard approach (most non-banks): use base salary + 2-year average of vested RSU income, with documentation showing past vesting. Future-vesting RSUs not counted unless within 12 months and tied to a current employment commitment.

Aggressive approach (private banks, some non-banks): structure the loan against the borrower's overall asset picture rather than just income. Use brokerage assets as compensating factor, leverage existing investment portfolios, accept past-vest history as predictive.

The lesson: find a loan officer who has experience with equity comp. The first quote from a generic jumbo lender is often artificially low because it ignores half your real income. This is one of the most important questions to ask before committing to any jumbo lender.

Often yes — it's worth running the math, especially if you have liquidity sitting in low-yield accounts.

The setup: let's say your current balance is $880,000, just $47,250 over the 2026 baseline conforming limit of $832,750 (you're not in a high-cost county). Your jumbo rate is 6.5%; conforming would be 6.0%. The 0.5% rate difference on $832,750 is roughly $260/month, or $93,600 over 30 years.

To get into conforming territory, you'd pay down $47,250. If that money is sitting in a 4.5% money market, you're giving up $2,126/year in interest income. The mortgage savings are $260/month or $3,120/year — net win of $994/year, plus you're now 47K closer to fully owning your home. Plus the $47K is "earning" the mortgage rate by reducing it, which is risk-free 6.5% on that capital.

The math gets more compelling at higher rate-spread differences. At 1% spread (7% jumbo vs 6% conforming), the same paydown saves over $5,000/year vs $2,126 forgone interest — clear win.

Does it backfire? Only if the rate spread is small (under 0.25%) or you have higher-yield investments earning materially more than the mortgage rate. The right answer depends on your full balance sheet — not just the mortgage in isolation.