When Does It Make Sense to Refinance?

The honest answer is a single number: break-even months = total closing costs divided by your monthly principal-and-interest savings. If you plan to keep the loan longer than that number of months, the refi pays for itself. Two important qualifiers: (1) rate-and-term and cash-out are different products with different rules, and (2) a refi can be worth doing even at a higher rate if it removes mortgage insurance (MI) — that's a separate calculation. "No-cost" refis are real, but the closing costs don't disappear; they're paid by accepting a higher rate. The decision is always: break-even months vs. how long you'll keep the loan.

The handbook view (what the rules actually say)

Refinances split into two product families with different agency rules. Knowing which one you're doing matters for pricing, LTV (loan-to-value) limits, and disclosure timing:

• Rate-and-term (also called "limited cash-out" on conventional): payoff of the existing first mortgage, financing of closing costs, and limited cash back (the Fannie cap is generally the lesser of 2% of the new loan or $2,000). Typical LTV up to 95–97% on conventional, 97.75% on FHA, 100% on VA Interest Rate Reduction Refinance Loans (IRRRL). (Source: Fannie Mae Selling Guide B2-1.3-02, Limited Cash-Out Refinance Transactions; Freddie Mac Single-Family Seller/Servicer Guide Section 4301.4; HUD Handbook 4000.1, II.A.8; VA Lender's Handbook Chapter 6.)
• Cash-out: any refi where the new loan exceeds the payoff + costs + the rate-and-term cash-back cap. Standard conventional cap is 80% LTV on a primary residence; FHA caps at 80% LTV; VA can go to 100% on cash-out (lender-overlay dependent). Conventional cash-out requires a minimum 6-month seasoning from the prior note date in most cases. (Source: Fannie Mae Selling Guide B2-1.3-03, Cash-Out Refinance Transactions; Freddie Mac Section 4301.5; HUD Handbook 4000.1, II.A.8.d.)
• Disclosure timing under TRID (12 CFR § 1026.19): a Loan Estimate must be delivered within 3 business days of application, and the Closing Disclosure must be received by the borrower at least 3 business days before consummation. On a refi of a primary residence, you also have a 3-business-day right of rescission after closing (12 CFR § 1026.23) — the loan doesn't fund until the rescission window closes.
• MI removal as its own refi reason: under HOPA (12 USC §§ 4901–4910), conventional PMI auto-terminates at 78% LTV on the original amortization. FHA annual MIP, by contrast, stays for the life of the loan when the original down payment was less than 10% (Source: HUD Handbook 4000.1, II.A.1.b.iv). Refinancing from FHA into conventional once you have 20% equity removes that monthly MIP permanently — sometimes worth doing even if the new rate is higher than the old one.

The plain-English translation

Strip out the jargon and the question becomes a single piece of arithmetic:

• Take the total closing costs of the refi (lender fees + title + recording + any prepaids that aren't already escrowed). Divide by the monthly P&I savings. That's your break-even in months. If you plan to keep the loan longer than that, the refi is profitable. Shorter, and you lose money on the trade.
• The often-repeated "you need 1% lower rate to refinance" is a rule of thumb, not a rule. On a small loan, even a half-point drop might not break even inside 5 years. On a large loan, an eighth might. The break-even formula is the actual answer.
• Rate-and-term is the "just lower my payment" refi. Cash-out is taking equity out as a check at closing — different product, tighter LTV cap (usually 80% on conventional and FHA), and almost always a slightly higher rate than rate-and-term on the same day.
• MI removal is its own reason to refi even if the rate is similar. If you're paying FHA MIP for life and you now have 20% equity, refinancing into a conventional loan with no PMI can save more per month than a rate change would — that's a real refi reason, separate from rate.
• A "no-cost refi" means the lender pays your closing costs by giving you a higher rate. The costs still exist; you're just paying them monthly forever instead of once at the table. That can be the right move if you're going to sell or refi again in a few years, and the wrong move if you'll keep the loan a long time.

Lender overlays — where the rules get tighter

The handbook minimums above are the program floor. Individual lenders impose "overlays" — tighter rules on top of the program rule — that can push the decision around even when the math says go:

• Seasoning overlays: agency rules generally require 6 months from the prior note for conventional cash-out; many retail lenders overlay 12 months on cash-out and rate-and-term both. Brokers can shop for an investor that holds to the agency rule.
• FICO floor on cash-out: Fannie's minimum FICO for cash-out at high LTVs is generally 680 in the Selling Guide grid, but many lenders won't quote cash-out under 700–720. The same loan repriced through a different wholesale investor can clear at a lower score.
• Appraisal-waiver eligibility: Fannie's value-acceptance and Freddie's ACE programs can waive a full appraisal on some rate-and-term refis, which knocks $600–$900 of closing cost off and changes the break-even meaningfully. Not every lender runs the file through the AUS in the order that triggers the waiver offer.
• Net-tangible-benefit policies: some lenders refuse to do a refi that doesn't hit an internal "benefit" threshold (rate drop of 0.5%+, specific savings amount). Net-tangible-benefit rules exist at the agency level for FHA streamlines and VA IRRRLs (HUD Handbook 4000.1 II.A.8.d.v; 38 CFR § 36.4307), but retail overlays often add more. Brokers see when the overlay, not the agency, is blocking the deal.
• Cash-out LTV overlays: the agency cash-out cap on conventional is generally 80% LTV, but some lenders quietly overlay 75% (or restrict 80% to owner-occupied 1-unit only). A second-home or 2–4 unit cash-out can require shopping to find the lender that prices it cleanly.

The break-even decision rule

A 30-second filter that beats most refi advice:

1. 1Calculate break-even months = total closing costs ÷ monthly P&I savings. Use principal-and-interest only; ignore taxes and insurance (they're the same on either loan).
2. 2Estimate how long you'll realistically keep the loan. Plan to sell in 3 years? 36 months. Plan to stay 10+? 120 months. Be honest, not optimistic.
3. 3If your stay-in-loan horizon is comfortably longer than the break-even (say 2x or more), the refi makes sense. If it's shorter, it doesn't — even if the monthly drop looks attractive.
4. 4If you're removing FHA MIP (or PMI you can't get auto-canceled), redo the math with the MI savings added to the P&I savings. The break-even often shifts dramatically.
5. 5For a no-cost refi, the break-even formula doesn't apply (there's no upfront cost to recover). Instead, compare the rate you're accepting to the rate on the same refi with costs paid — and decide if the rate premium is worth not writing the check.

Which lenders we actually use for this scenario

For a clean rate-and-term refi with strong credit and a W-2 borrower, I'm usually shopping the big agency-direct wholesale shops that price aggressively on conforming product. They give us pricing inside an eighth of a point of one another on most days, and the win is matching the borrower to whichever one has the lowest LLPA (loan-level price adjustment) hit for that file's specific FICO-and-LTV bucket.

For MI-removal refis where the borrower's credit has improved since the original FHA loan, I lean toward lenders that price conventional 720+ aggressively and don't pad the rate sheet with overlays. For borrowers who are tight on cash to close, I'll run “lender credit” pricing — accepting a slightly higher rate so the lender funds the closing costs out of yield spread — but only when the break-even math still pencils with the higher rate baked in.

For self-employed borrowers refinancing, I'm steering toward shops with experienced underwriters on bank-statement and P&L product. The rate is rarely the issue on those files; the issue is which underwriter actually understands the income calculation.

Real-world cases

I've seen this pattern: borrower calls because their retail bank's loan officer told them “rates are down, you should refinance.” We run the numbers — they're saving $84 a month, closing costs are around $5,800, break-even is roughly 69 months. They're planning to sell in three years to upsize. That's a money-losing transaction dressed up as savings. We told them to stand pat. The retail LO didn't.

Another pattern: FHA borrower from 2021, bought with 3.5% down, now has roughly 22% equity from appreciation plus paydown. Their FHA MIP is running them $180-something a month and won't drop off (FHA MIP for the life of the loan when you put less than 10% down). Refinancing into a conventional rate-and-term at a rate similar to their current rate eliminates the MI entirely. Even with closing costs, the break-even is inside two years and the borrower is keeping the house long-term. That's a real refi.

Third pattern I see a lot: “no-cost refi.” The borrower is told there are no closing costs. There are. They're embedded in the rate — the lender is taking a higher rate so the yield premium covers the costs. That's a legitimate structure when the borrower is short on cash or planning to refi again soon, but the rate is higher than market, and the comparison to “what could I have gotten if I paid the costs out of pocket” almost never gets shown. Ask for it shown both ways.

How the big retail lenders typically handle this

The retail-channel pitch on refinancing is, in my experience, almost always rate-first and break-even-quiet. You'll hear “rates dropped, let's lock you in” before you hear anything about how many months it takes to recover the costs. That's not always wrong, but the incentive structure inside a retail call center rewards the refi closing, not the borrower's net-present-value over the next five years.

The “no-cost” framing is the one I'd push back on hardest. It's structurally legitimate — yield-spread-funded closing costs are how the math works — but it's marketed as a free lunch and it isn't. You're paying for the closing costs every month for as long as you hold the loan, through the rate. On a long hold that's a worse deal than paying the costs up front; on a short hold it's a better deal. The retail conversation rarely makes that distinction.

The big retail shops will quote you a refinance. They're usually competitive on rate sheet alone. Where the broker channel earns its keep is showing you the same loan three ways — pay the costs, embed the costs, pay points to buy down further — and letting the break-even math decide which version fits your actual hold period. That conversation takes about twenty minutes and it's the conversation that actually answers the question of whether to refinance at all.