There is a rule of thumb that says refinance when rates drop a full point. It is wrong often enough to be dangerous, in both directions.
The real test: divide your closing costs by your monthly savings. If you will still own the home well past that many months, refinancing is worth considering. If not, it is not.
The rule ignores the two variables that decide the answer: how much the refinance costs and how long you will keep the loan.
Put numbers on it and the rule collapses immediately. A half-point drop on a $600,000 balance saves $195 a month; with $4,000 in costs that is repaid in 21 months. A full point drop on a $120,000 balance saves $79 a month, and with $6,000 in costs it takes 76 months — six and a half years.
The smaller rate move on the larger loan is nearly four times faster to repay. The rule of thumb points the wrong way in both cases.
Here is the whole picture, assuming $6,000 in closing costs and a thirty-year term. Find your balance, then the drop you have been offered:
| Balance | −0.5% | −0.75% | −1.0% | −1.5% |
|---|---|---|---|---|
| $150,000 | 123 | 82 | 62 | 42 |
| $250,000 | 74 | 49 | 37 | 25 |
| $350,000 | 53 | 35 | 27 | 18 |
| $500,000 | 37 | 25 | 19 | 13 |
| $750,000 | 25 | 16 | 12 | 8 |
Read it as a single fact: the same rate drop is a different decision at different loan sizes. A full point is a poor trade on $150,000 and an obvious one on $750,000, because closing costs barely move while the monthly saving scales with the balance.
The green cells are break-evens under about two and a half years, which most owners will comfortably clear. The top-left corner is where refinances go to die — and it is exactly the corner the 1% rule sends people into.
Take your closing costs and divide by the monthly payment reduction. That is your break-even in months.
Then ask the harder question: will you still own this home then? People consistently overestimate how long they will stay. If the break-even is 52 months and you expect to move in four years, the deal is a wash at best.
Run your own numbers
See the break-even month and what a refinance costs across the whole loan.
A lower monthly payment does not mean lower cost, and the gap between those two things is larger than almost anyone expects.
Take a real case. You are eight years into a thirty-year loan: $260,000 left, 7.0%, twenty-two years to run, paying $1,933 a month. Rates have fallen a full point and you refinance at 6.0% with $6,000 in costs rolled in. Three versions of the same refinance:
| What you take | Monthly | Interest left | vs. staying put |
|---|---|---|---|
| Stay as you are | $1,933 | $250,286 | — |
| New 30-year | $1,595 | $308,130 | +$57,843 |
| New 22-year | $1,817 | $213,682 | −$36,604 |
| New 15-year | $2,245 | $138,039 | −$112,248 |
Look at the second row. A full point off the rate, $338 a month off the payment, and it costs $57,843 more than doing nothing at all. The rate genuinely improved. The term reset ate the improvement and then some.
That row is what most refinances look like, because thirty years is the default a lender quotes and a lower payment is what makes the offer feel like a win. Nothing about it is dishonest. It is simply a different transaction from the one the borrower thinks they are doing.
The third row is the same rate improvement without the reset — $36,604 saved instead of $57,843 lost, a swing of $94,000, for $222 a month more than the thirty-year version.
Match the term to what you have left. Twenty-two years remaining, ask for a twenty-year or a custom term. Fewer lenders advertise these but most will write them.
Or take the thirty and keep paying the old amount. This reproduces the shorter term almost exactly while leaving you a lower required payment if a bad month arrives. You get the safety net and the arithmetic at once — provided you actually keep paying the old amount, which is the part that fails in practice.
Whichever you choose, the number to compare between offers is total interest remaining, not the monthly payment. The monthly payment is the number designed to be compared, and it is the one that misleads.
If your rate is already low by current standards, refinancing to lower a payment almost never works — you would be raising the rate on the entire balance to buy a monthly discount. If you need cash flow relief and hold a low rate, look at a recast instead, which keeps your rate and costs a few hundred dollars.
Also leave it alone if you might move within the break-even window, or if your credit has weakened since you closed. The rate you are quoted depends on your profile today, not the one you had at purchase.
Dividing costs by monthly saving is the right first test, and it is not the whole test. Three things it leaves out:
A more complete test is two questions rather than one. Does it break even before I expect to leave? And does total interest go down? A refinance that passes both is worth doing. One that passes only the first is a cash-flow decision, which is legitimate but should be made knowingly.
If cash flow is the actual goal and your rate is decent, compare against a recast before refinancing — it lowers the payment without touching the rate or restarting the clock. The full comparison is in recast or refinance?