What Actually Sets Your Mortgage Rate, and Why Your Neighbor's Rate Tells You Nothing

Somebody at work got 5.75%. You were quoted 6.375%. Same town, same week, same size loan. One of you is being taken advantage of, obviously.

Almost certainly not. A mortgage rate is not a price on a shelf. It is the output of eight or nine inputs, and you and your coworker differ on at least four of them. Unless you know all four, the comparison tells you nothing at all.

Here is the full list, in rough order of how much leverage each one has.

Credit score

The largest single borrower-controlled input. Pricing is set in score bands, and the bands are steps, not a slope. Moving from 719 to 720 can change your price; moving from 700 to 718 may change nothing.

Two practical consequences. First, if you are near the top of a band, ask what the next band up is worth in dollars. Sometimes paying down one revolving balance before you lock is worth more than anything else you can do in thirty days. Second, the score that matters is the mortgage credit score pulled by the lender, which uses a different scoring model than most free consumer apps. The number in your banking app is not the number being priced.

Loan-to-value

How much you are borrowing against the value of the home. Like credit, it prices in bands, commonly at 80%, 75%, 70% and 60%. Crossing below a threshold can improve pricing.

The 80% line does double duty, because it is also where mortgage insurance generally comes off the table on a conventional loan. That is a payment effect rather than a rate effect, but it lands in the same place: what you pay each month. Our breakdown of what a Montana mortgage payment actually includes separates the pieces.

Note that credit and LTV interact. A price adjustment is usually drawn from a grid with score on one axis and LTV on the other, so the cost of a lower score is larger at higher LTV.

Occupancy

Primary residence prices best. A second home prices worse. An investment property prices worse still, and by a meaningful margin. This is one of the largest single adjustments on the grid. If your coworker's rate is on their own house and you are buying a rental, you are not comparing the same product.

Property type

A single-family detached home is the baseline. A condominium typically carries an adjustment, more at higher LTV, and a project that fails agency warrantability leaves the conventional world entirely. Manufactured housing has its own pricing. Two- to four-unit properties price above single-family.

Loan amount, against your county's limit

Below the conforming limit you are in agency pricing. Above it you are in jumbo, which is priced by whoever is buying the loan rather than by the agencies, sometimes better, sometimes worse, always underwritten more tightly. In 2026 every Montana county sits at the same baseline conforming limit, which is why the jumbo line surprises buyers in the higher-priced markets; the jumbo page covers where it falls and what changes above it.

Very small loan amounts also price worse, because the fixed cost of originating a $90,000 loan is nearly the same as originating a $400,000 one.

Term

A 15-year fixed generally prices below a 30-year fixed. You pay far less total interest and a much larger monthly payment. That is not a better deal or a worse one, it is a different one, and it belongs in a calculator rather than an argument.

Points

Discount points are prepaid interest. You pay money at closing to buy a lower rate for the life of the loan. One point is 1% of the loan amount. Lender credits run the same machinery backwards: you accept a higher rate and the lender pays some of your closing costs.

The math is a break-even, and it is simple enough to do on a napkin.

[keyfact] Break-even = cost of the points ÷ monthly payment savings.

Pay $6,000 to save $95 a month, and you break even at 63 months, a bit over five years. Keep the loan longer than that and the points paid off. Sell or refinance sooner and you lost money. [/keyfact]

So: buying points is usually worth it if you are confident you will hold the loan well past break-even, if you have cash beyond a healthy reserve, and if rates are not likely to fall far enough that you refinance out of it. It is usually not worth it if the cash would be better spent on a larger down payment that drops you below an LTV threshold, if you might move in a few years, or if paying points empties your reserves. On a starter home in a market where people move in five or six years, points are frequently the wrong purchase.

Run yours on the payment calculators with and without.

Lock length, and the bond market that day

Mortgage rates track mortgage-backed securities, not the federal funds rate. That is the single most useful thing to know about rate news. The Federal Reserve can cut and mortgage rates can rise the same afternoon, because the bond market prices expectations ahead of the announcement.

A rate lock fixes your rate for a set number of days while your loan is processed. It protects you from the market moving up. It also means you do not benefit if the market moves down, unless your lock includes a float-down provision, which typically requires the market to improve by some minimum amount and can be used once.

Longer locks cost more, because the lender is carrying the risk for longer. A 60-day lock prices worse than a 30-day. Extensions cost money too (usually a fraction of a point per week) and they are most often needed because something in the file was late, which is an argument for getting your documents in early rather than for shopping harder.

And your rate is set on the day you lock. Two identical borrowers who locked eleven days apart can hold different rates permanently. Comparing your rate to one somebody locked in a different week is comparing two different markets. Freddie Mac's weekly survey is a fine way to see the trend, but it is a national average of surveyed lenders, not a quote you can get.

APR versus interest rate

The interest rate determines your monthly principal and interest payment. The APR expresses the rate plus most of the lender's costs as a single annualized percentage, so it is designed to make two offers comparable.

APR is genuinely useful and it is not a complete answer. It assumes you hold the loan for the full term, which most people do not, and lenders can differ in which third-party fees they include. Use it as a first filter, then compare the actual Loan Estimates line by line. The standardized form exists precisely so that you can.

One more input that is not on the pricing grid: how your income is documented. A borrower using a bank statement or asset-based program is buying access to a smaller pool of investors, which shows up as a higher rate for the same credit and equity: how underwriters average self-employed and seasonal income covers when that trade is worth making.

The honest ending

Look back at that list. Credit band, LTV, occupancy, property type, loan amount against the limit, term, points, lock length, and the market on the day. Your neighbor's rate is one specific answer to nine specific questions, and you do not know eight of their answers.

Which leads to the practical rule: a rate quoted before someone has seen your file is a guess. Anybody handing you a number before they know your score, your down payment, your occupancy and your property type is not pricing your loan, they are advertising. The number will change later, and it will change in one direction.

What you should ask for instead is a Loan Estimate. It is a standardized form, it shows the rate and the costs together, and it can be compared side by side with another lender's. That comparison is real. The one at the barbecue is not.

If you want the underlying mechanics rather than the pricing, the learning center covers how the pieces of a loan fit together, and the loan programs page lays out which product families exist and who each is for.

Send us the actual details and we will send back an actual number, with the assumptions written down next to it.