Guide to Mortgages

The number most buyers shop last

You’re thirty, four months into house-hunting, and three Loan Estimates, the standardized form every lender must give you, are printed on the kitchen table in front of you. You spent weeks negotiating $8,000 off the asking price. The three rate offers sit within about half a percentage point of each other, and you’re leaning toward the one from the bank you already keep your checking account with. It’s familiar, and the numbers all look close enough.

They aren’t. On a $425,000 loan, half a percentage point is worth roughly $50,000 over the life of the loan — more than six times the price cut you spent weeks negotiating. And unlike the price, the lower rate shrinks the monthly payment instead of stretching it.

But the house has an address you can drive past, and the rate is a percentage a loan officer reads to you over the phone, so it gets the least of your attention. That asymmetry is the trap this guide is built around: the single number that sets the lifetime cost of a mortgage is the rate, and it’s the one most buyers shop last.

Why this guide exists

For most households, a mortgage is the largest single debt they will ever carry, and the longest. Thirty years of compounding works in your favor when the dollars are inside a Roth IRA; it works against you when the dollars are interest on a half-million-dollar loan. The arithmetic is the same; the direction is opposite.

That kitchen-table choice is the last easy moment to change any of it. By the closing table the structure finally gets explained — and by then the rate is locked, the points are baked in, the escrow account is opening, and the amortization schedule is set for the next three decades. The decisions that matter most are the ones already made before that table.

This guide covers the upstream ground: what the rate structure buys you, when points pay back, what triggers and removes PMI, what your servicer (the company you actually send the payment to — often not the lender that originated the loan) holds in escrow, and why every dollar of interest in year one is doing far more damage than every dollar of interest in year twenty-five. The mortgage payoff calculator handles the downstream strategy — extra payments, lump sums, bi-weekly schedules — once the loan is in place.

The two questions every mortgage answers

Every mortgage offer, from every lender, answers the same two questions in slightly different ways.

The rate-structure question: is the rate locked, or does it move? A fixed-rate mortgage holds the same interest rate for the full term. An adjustable-rate mortgage — usually written as ARM — holds a starting rate for a fixed introductory period (commonly five, seven, or ten years), then resets to a market-tied rate that can move up or down at each adjustment after that.

The cost-structure question: what does the loan cost you, all in? Rate is the headline, but the real cost over thirty years is rate + upfront points + monthly PMI (until it’s removed) + property taxes and homeowners insurance held in escrow + any HOA dues. Comparing two “6.5%” offers without comparing those five pieces is comparing two different products.

The rest of this guide walks through each piece in the order it appears on a typical disclosure.

Fixed vs ARM

A thirty-year fixed-rate mortgage is the American default for a reason: it shifts interest-rate risk from the borrower onto the lender, and the lender prices that risk into the rate. You pay a slightly higher rate in exchange for never having to think about the rate again.

An ARM goes the other direction. The starting rate is typically lower than the fixed rate for the same buyer — sometimes by half a point, sometimes by more — because the lender retains the right to reprice the loan at the first adjustment date. The savings are upfront; the risk is at the back end.

The terminology is worth knowing because it tells you exactly what you own:

• 5/1 ARM — fixed for five years, then adjusts once per year for the remaining twenty-five.
• 7/6 ARM — fixed for seven years, then adjusts every six months.
• 10/1 ARM — fixed for ten years, then annual adjustments.

Most ARMs cap each adjustment in three places: an initial cap on the first reset, a periodic cap on every reset after that, and a lifetime cap on how high the rate can go above the starting rate. Lender shorthand stacks the three as “5/2/5” — five points on the first reset, two on each one after, five points lifetime. Typical ranges are two points per reset and five points lifetime. The caps make a worst-case ARM payable; they do not make the difference small.

The chart shows the spread visually. The fixed line is a thirty-year promise — same payment every month. The ARM intro is a five-year promise at a slightly lower payment, followed by a reset that lands somewhere inside the cap-bounded range. Best case (rates fell) the payment drops by a few hundred dollars. Worst case (rates rose to the periodic cap) it jumps by about the same amount. The borrower carries both possibilities; the lender carries neither.

When the ARM math works

A 5/1 ARM is rational under a narrow set of conditions. All three usually need to be true at once:

• You expect to be gone before the first reset. A planned five-year ownership horizon — a starter home, a military rotation, a known job transfer — turns the back-end risk into someone else’s problem.
• The rate spread is large. Half a point on a five-year horizon matters; a tenth of a point does not. Run the dollar savings against the closing-cost spread before signing.
• You can absorb the reset if your plans change. Cap-rate math matters here. On a 6% start with a 2-point first cap, the reset worst case is an 8% rate — about a 25% increase in the principal-and-interest portion of your payment. If that breaks the budget, the ARM is not a fit even with a five-year horizon.

Outside that narrow case, the thirty-year fixed wins on simplicity alone. The rate certainty is worth the spread for buyers who plan to stay, for buyers who can’t easily refinance later (self-employed income, credit hiccups), and for buyers whose budget runs tight enough that a payment increase would hurt.

Points — paying upfront to lower the rate

A discount point (often just “point”) is a fee you pay the lender at closing in exchange for a lower interest rate over the life of the loan. One point typically equals 1% of the loan amount and lowers the rate by roughly 0.25 percentage points; the exact rate reduction per point varies by lender and by current market conditions, so confirm the trade on the loan estimate before signing.

The trade is a closing-cost dollar for a lower monthly payment. It’s worth it only if you keep the loan long enough for the monthly savings to repay the upfront cost.

The break-even calculation is concrete:

• Cost of the point. 1 point on a $425,000 loan is $4,250 at closing.
• Monthly savings. Roughly 0.25 percentage points off the rate. On the same $425,000 loan, that’s about $70 per month in the early years (the exact savings drift slightly because of how amortization works, but the rounded number is close enough for the decision).
• Break-even months. $4,250 ÷ $70 ≈ 61 months — a little over five years.

That’s the test. If you’ll keep the loan past the break-even month, points pay back. If you sell or refinance before then, you paid for a rate reduction you never collected.

A few practical wrinkles:

• Refinancing voids the math. Buying points on a 6.5% loan and refinancing to 5% two years later wastes the point cost. The break-even assumed you held this loan, not a future one. (Refinancing has a break-even test of its own — see the refinancing guide.)
• Lender credits run the other direction. Some lenders offer negative points — a higher rate in exchange for a credit toward closing costs. Same math, opposite direction. Useful when cash for closing is tight and you expect to refinance soon.
• Watch for points already baked in. The advertised rate sometimes assumes a half-point purchase. Compare loan estimates with the points-paid column lined up, not the headline rate alone.

The three levers a buyer can pull

Once you’ve decided to buy, three numbers on the offer remain yours to influence: the loan amount (price minus down payment), the interest rate (lender competition, credit score, points), and the loan term. Each one moves the lifetime cost of the loan in different amounts and, more importantly, in different directions. Most buyers underweight the rate lever relative to the price lever — the math says the opposite.

Reading the chart against the $425K @ 6.5% base:

• Lower rate is the only lever that pulls both numbers your way at once. A single percentage point off the rate saves $98K of lifetime interest AND lowers the monthly payment by $273. The star marks the asymmetry — no other lever can do this.
• Bigger house raises both. Going about 25% larger on the loan ($425K to $530K) adds $134K of lifetime interest and $664 to the monthly payment. Two numbers, same direction.
• Shorter term (20 years instead of 30) saves more interest than either rate move — $207K — but raises the monthly payment by $482. The savings come from forcing more principal into every payment from day one. A trade, not a win.

The practical takeaway: spend more time shopping the rate than shopping the listing. A few hours calling additional lenders, fixing a credit-score line item, or running the points break-even can buy you a half-point of rate. On a $425K loan, that half-point saves about $50K of lifetime interest — and lowers the monthly payment instead of raising it.

Federal law gives you the tool to do the shopping cleanly: every lender must provide a standardized Loan Estimate within three business days of a completed application. The format is identical across lenders, so line-item comparison is direct. Request three; take the best one — and use it as leverage with the others.

Buying with no credit score

The rate lever above assumes you have a credit score to shop with. Some buyers don’t, whether by youth, by circumstance, or on principle. A missing score doesn’t shut the door on a mortgage; it changes who opens it.

Most applications run through an

system — software that scores the file in seconds against the lender’s rules. With no score to read, the software can’t decide, so a person does instead: manual underwriting. A human underwriter reads your full file by hand and builds a credit picture from nontraditional credit: the on-time rent, utility, insurance, and phone payments that never reach a bureau but prove the same thing a score does. Expect to hand over more paper than a standard file: a year or so of documented on-time payments (canceled checks, a landlord letter, utility bills), plus a longer wait while a person works through it.

Not every loan program allows it, and not every lender offers it:

• VA loans set no minimum credit score and lean on residual income (the cash left each month after the mortgage and major bills) as the core test. Manual underwriting is routine here.
• FHA loans accept nontraditional credit for a buyer with no score, and lenders commonly route a low-score FHA file to manual review.
• USDA and some conventional loans allow it too, under tighter conditions.

Because a manual underwrite is hands-on work, fewer lenders bother with it, so the first step is finding one that does. A mortgage broker, or a free HUD-approved housing counselor, can point you to lenders that will. From there, approval turns on compensating factors: the strengths that offset a missing score. Verified cash reserves (a few months of payments saved), a low debt-to-income ratio, years of steady employment, and a larger down payment all carry weight; lenders weight them differently (the VA leans hardest on residual income), so ask which matter most for your file.

Be honest about the tradeoffs. A manual file asks for more documentation, holds you to tighter debt-to-income limits, and moves slower than a few clicks of software. But a no-score, debt-free buyer is not locked out. The same boring habits that build a score in the first place (paying rent and bills on time, keeping reserves) are exactly what a manual underwriter looks for. If you’d rather build a score before you apply, the credit-score guide shows how.

PMI — what triggers it, what removes it

Private mortgage insurance (PMI) protects the lender, not the borrower. When you put down less than 20% on a conventional loan, the loan-to-value ratio — the LTV — is above 80%, and the lender requires PMI as a hedge against default. The borrower pays the premium; the lender collects the payout if the loan goes bad.

PMI typically costs between 0.3% and 1.5% of the loan amount per year, billed monthly as part of the PITI line on your mortgage statement. For borrower-paid PMI (BPMI, the standard flavor below), the premium is assessed on the original loan balance, so the monthly charge stays flat year over year until it drops off. On a $425,000 loan at 1%, that’s $4,250 per year, or about $354 per month — paid until the LTV drops below the cancellation threshold.

Two thresholds matter under the Homeowners Protection Act of 1998, a federal law (often shortened to HPA) that gives borrowers cancellation rights once specific equity milestones are met:

• 80% LTV — borrower-requested cancellation. When the loan balance reaches 80% of the original purchase price, you can request PMI cancellation in writing. The lender may require evidence the home hasn’t lost value (sometimes an appraisal at your cost) and that you’re current on payments.
• 78% LTV — automatic termination. When the scheduled amortization drives the loan balance to 78% of the original price, the servicer must remove PMI automatically. No request needed. This is the line that protects borrowers who don’t know about the 80% rule.

Two readings of the chart worth sitting with:

• PMI is finite. Most low-down buyers worry the premium will stretch on indefinitely. The math says otherwise: on a typical 5%-down loan, scheduled amortization clears the 78% line in about eleven years even without extra principal payments.
• Knowing the 80% rule is worth about $3K. Servicers won’t cancel early without a written request. The roughly $3K of premium saved by requesting cancellation at 80% (instead of waiting for the 78% auto-trigger) is about an hour of paperwork.

Both thresholds use the original purchase price, not current market value — which means PMI lingers on its scheduled timeline even if your neighborhood appreciated. The exception: if you’ve held the loan at least two years and can document significant appreciation, you can request reappraisal-based cancellation. The mechanics: order a new appraisal at your cost (typically $400–700), submit it with a written cancellation request, and the servicer applies its appreciation policy (commonly 75% LTV against the new value). Policies vary; read the original loan disclosure first.

The two PMI flavors worth knowing

• Borrower-paid (BPMI). The monthly premium described above. Removable on the HPA schedule.
• Lender-paid (LPMI). The lender pays the premium in exchange for a slightly higher rate, baked into the loan for its full life — including past 78% LTV. Because the cost lives in the rate rather than a separate line, the only way out is to refinance; default to BPMI unless you expect to refinance within about five years anyway.

Escrow and PITI

The monthly figure on your mortgage statement is usually larger than the loan’s principal-and-interest payment alone. Most servicers bundle four costs into one number, which is where the acronym PITI comes from:

• Principal — the portion of the payment that reduces the loan balance.
• Interest — the portion that goes to the lender as the cost of the loan.
• Taxes — annual property taxes, held in escrow and remitted to the county on your behalf.
• Insurance — annual homeowners insurance, held in escrow and remitted to the carrier on your behalf.

The escrow account is the servicer’s way of making sure the taxes and insurance get paid on time, because a tax lien or a lapsed insurance policy creates risk for the lender as well as the borrower. Each month, roughly one-twelfth of the annual tax and insurance bills lands in escrow; once per year, the servicer pays the bills from the account.

A few realities buyers discover only after the first statement:

• The payment can change every year. When property taxes rise — or your homeowners insurance carrier raises premiums — the escrow piece of PITI rises with them. The principal-and-interest piece stays flat (on a fixed-rate loan); the T and the I do not.
• Escrow shortages get spread over twelve months. If taxes went up more than the prior year’s escrow assumed, the next year’s monthly payment increases to make up the gap and rebuild the cushion (under the Real Estate Settlement Procedures Act, servicers may hold up to two months of payments as a cushion, which is why first-year “shortage” letters land even on stable bills). A 10% property-tax assessment increase can show up as a 1–2% jump in your monthly PITI.
• Property tax is the largest swing factor in most states. Insurance premiums vary, but property taxes vary far more — by state, by county, and year over year. The Tax Foundation tracks median effective property-tax rates by state; the range runs from under 0.3% to over 2% of home value annually. The same $425,000 house in two different states can produce monthly PITI figures that differ by hundreds of dollars on the tax line alone.

Amortization — the curve everyone gets wrong

The defining feature of a mortgage is amortization: the way a fixed-rate loan splits each monthly payment between interest and principal. The total payment stays flat for thirty years, but the split inside it changes every month — and the change is dramatic.

In month one, almost the entire payment is interest. In month three-sixty, almost the entire payment is principal. Between those two points, the curve crosses over — usually somewhere around year sixteen to year twenty on a thirty-year loan, depending on the rate.

The mechanism is straightforward: interest each month is calculated on the remaining balance. When the balance is high, interest is high; when the balance is low, interest is low. The fixed monthly payment makes up the difference with principal, which means the principal portion grows every single month, automatically.

The same math that makes early dollars worth more also makes a fifteen-year mortgage a radically different product from a thirty-year one. Shorter terms force more principal into every payment from day one, so less of the balance gets time to compound into interest on the lender’s books. The trade is a higher monthly payment for a dramatically smaller total cost over the life of the loan.

The chart shows the size of the trade. At a $425,000 loan and a 6.5% fixed rate, the fifteen-year option costs about $1,016 more a month than the thirty-year — and saves roughly $301,000 of interest over the life of the loan. The math is the same lever the consequence bullets below describe; the chart is what it looks like when you pull it all the way at signing instead of in pieces over the next thirty years.

The consequence shapes every payoff decision:

• An extra dollar in year one shaves more than an extra dollar in year fifteen. The year-one dollar would have spent twenty-nine years compounding into interest on the lender’s books. The year- fifteen dollar has only fifteen years left to do damage.
• Lump sums hit hardest early. $10,000 applied at month thirty-six (year three) saves more total interest than the same $10,000 applied at month one-twenty (year ten).
• Bi-weekly schedules add one full payment per year. Twenty-six half-payments equal thirteen monthly payments instead of twelve. The extra payment goes straight to principal, every year.

The mortgage payoff calculator shows the exact dollar effect on your loan, with extra payments, lump sums, and bi-weekly schedules modeled side by side. The calculator includes a chart that shows the interest-to-principal crossover for any rate you punch in.

The mortgage interest deduction, briefly

The home mortgage interest deduction is the most-talked-about, least- relevant tax break in personal finance. It was relevant before 2018. For most filers, it isn’t anymore.

Two changes from the Tax Cuts and Jobs Act of 2017 broke it for the majority of households:

• The standard deduction roughly doubled. Pre-2018, a typical married-filing-jointly household itemizing $13,000 of deductions cleared the (then much lower) standard deduction handily. The 2026 MFJ standard deduction is $32,200 — so that same $13,000 of itemized deductions doesn’t come close. The mortgage interest deduction only matters if you itemize, and you only itemize if your total itemized deductions exceed the standard deduction.
• The deductible loan balance was capped at $750,000. Down from $1,000,000 for loans originated after December 15, 2017. Existing loans were grandfathered at the old cap.

The practical result: roughly nine in ten filers now take the standard deduction, according to Treasury data tracked since the TCJA took effect. For those filers, the mortgage interest deduction is irrelevant — a 6.5% mortgage rate stays a 6.5% rate, not the lower “after-tax” rate older guides used to highlight (the rough math: rate × (1 − marginal bracket), only meaningful if every interest dollar actually shows up on Schedule A, the IRS form for itemized deductions).

When does it still matter? High-balance loans in high-property-tax states, where the combination of mortgage interest, the state and local tax (SALT) deduction (capped at $40,400 for 2026, phasing down toward a $10,000 floor at higher incomes and reverting to a flat $10,000 in 2030), and large charitable giving pushes itemized deductions above the standard threshold. That set of households is a minority, and they already have a CPA running the numbers — they don’t need this guide.

For everyone else, the takeaway is simple: don’t model the deduction into your decision unless your CPA tells you it changes the math. It mostly doesn’t.

See IRS Publication 936 (Home Mortgage Interest Deduction) for the current rules, and Publication 530 (Tax Information for Homeowners) for the broader homeowner-tax picture.

Where this fits in the order of operations

Before any sizing decision: the Money Guy housing rule caps PITI (principal, interest, taxes, insurance) at 25% of gross monthly income. That ceiling is the precondition for the order below; it’s why mortgage payoff lands so far down the list.

The Money Order of Operations places mortgage payoff late — well after the employer match, after high-interest debt, after the emergency fund, after the Roth and HSA, and usually after the fifteen-percent retirement target. A thirty-year fixed mortgage at a typical historical rate (roughly 7–8% nominal) sits below the threshold at which paying it down beats investing the same dollar; once the rate climbs above that band, the math tightens and a payoff preference becomes more defensible.

The sequencing is deliberate:

• Capture the match first, always. A 50–100% one-time return from the employer outpaces any mortgage rate. Don’t divert match dollars to extra principal.
• Kill high-APR debt next. A 22% credit card costs more than a 6.5% mortgage by a wide margin. The order is rate-driven, not emotion-driven.
• Fund the emergency fund. Mortgage prepayment is the opposite of liquid. Once a dollar is inside the house, getting it back out requires selling or refinancing — neither is fast.
• Hit the tax-advantaged accounts. Roth IRA, HSA, and the fifteen-percent retirement target generally beat the mortgage rate over a long horizon. Both returns are uncertain; the gap is wide enough that the asymmetry favors investing for most households.
• Consider extra principal only after the above. Even then, the decision is partly behavioral — a paid-off house in retirement shrinks the income you need to draw, which is its own form of risk reduction. The math says invest; the math doesn’t price peace of mind.

For the full sequence, see the order of operations guide and the walkthrough. For the strategy mechanics once you’ve decided to add extra principal, the mortgage payoff calculator models the dollar effect month by month.

Which loops back to the kitchen table. The sequence above governs the loan once you hold it; the rate you shopped before signing governs what it will cost for as long as you hold the loan. The hour spent lining up those three Loan Estimates outranks every payoff decision that follows it.

Vocabulary you’ll meet at the table

Realtors and loan officers move through these terms every day and rarely stop to define them. This closing section is lookup material, not part of the guide’s main argument: it collects the highest-confusion ones, grouped by where in the process you’ll hear them. Each links to its full glossary entry — skim now, or come back when one shows up on a form.

Shopping and qualifying

• Pre-qualification vs Pre-approval. A back-of-envelope estimate from numbers you say out loud, versus a verified letter after the lender checks your income, assets, and credit. Sellers respect the second; the first is informal interest.
• DTI (debt-to-income ratio). Total monthly debt payments over gross monthly income; conventional lenders cap it around 43%. The biggest single number setting how much loan you qualify for — it drives the no-score case in the section above too.
• Conventional loan vs FHA / VA / USDA. Conventional means no government backing — most loans, 5–20% down, PMI under 20% down. Government-backed loans (FHA, VA, USDA) trade easier qualification for different fees. Ask which one your lender is quoting; the numbers compare differently.
• Conforming loan vs jumbo. Conforming loans fit Fannie Mae / Freddie Mac limits ($832,750 baseline for 2026) and get the best rates; jumbo loans exceed those limits and price slightly higher.

Offer and contract

• Earnest money. The 1–3% deposit you put up when an offer is accepted — credited toward closing if the deal closes, forfeited if you back out without a covered contingency.
• Contingency. Escape clauses — financing, inspection, appraisal, home-sale. Waiving them strengthens your offer and raises your risk; hot markets expect some waived, so know what you’re giving up.
• Appraisal. The lender orders a licensed valuation against three recent Comparable from the neighborhood. Come in below contract and the deal usually gets renegotiated or falls apart.
• Seller concessions. A negotiated amount the seller pays toward your closing costs, capped by loan type (typically 3–6%) — ask your lender for the cap that applies to your loan. A hidden price reduction structured as a closing-cost credit.

Application and closing

• Underwriting. The lender’s risk audit, and the slowest stretch of closing; most “conditional approvals” are underwriting saying “yes if you produce documents X, Y, and Z.” Don’t change jobs or open new credit cards during this window.
• Rate lock. Locks your quoted rate for 30–60 days while underwriting completes; without it, market drift can change your monthly payment between application and closing. Always lock.
• Loan estimate (LE) and Closing disclosure (CD). The two federally required forms, standardized so you can compare lenders line-by-line — the comparison the opening of this guide turns on. The LE arrives within three days of application; the CD arrives at least three days before closing. If a number changed unexpectedly, ask why before signing.
• Closing costs vs Cash to close. Closing costs are the fees themselves (typically 2–5% of the price); cash to close is the full total you bring — down payment + closing costs + prepaids, minus credits. The bigger number, and the one your bank account has to deliver.
• Title insurance. Two policies: the lender’s is required and covers the lender against ownership-claim disputes; the owner’s is optional, one-time at closing (a few hundred to a few thousand dollars), and protects you for as long as you own the home. Most buyers take it.

Sources

• IRS Publication 936 — Home Mortgage Interest Deduction (deduction limits, qualified residence rules, the $750,000 cap)
• IRS Publication 530 — Tax Information for Homeowners (broader homeowner-tax picture)
• Consumer Financial Protection Bureau — Owning a Home (mortgage shopping, PMI rules, closing disclosures, loan estimates)
• HUD — Homeowners Protection Act of 1998 (the 80% / 78% LTV thresholds for PMI cancellation)
• Freddie Mac — Primary Mortgage Market Survey (PMMS) (historic fixed and ARM rate spreads, weekly back to 1971)
• Tax Foundation — Property Taxes by State (median effective property-tax rates, used in the escrow section)
• The Money Guy Show — Housing-cost framework (the 25% PITI-of-gross guardrail and the order-of-operations placement of mortgage payoff)