Costs & the Loan Estimate

Three pages of fees just landed in Brandon and Katie Sullivan's inbox. They found the house — a three-bedroom in Parma, outside Cleveland — got past the application and the underwriting (Lesson 15), and now a document titled "Loan Estimate" is sitting there with forty-odd line items, some in the thousands, most with names they've never heard: "origination charge," "tax service fee," "prepaid interest," "initial escrow payment." And the three questions running through their heads are the same three anyone has at this moment, so let's name them out loud. Is this a rip-off — are these fees real, or is someone padding the bill? Which of these can I actually push back on? And how on earth do I compare this against another lender's version without a math degree? This lesson answers all three, and the reassuring news comes first: that intimidating three-page form is not the lender's sales sheet. It is the opposite. It is a form the U.S. government designed, standardized, and requires — built specifically so that a first-time buyer like the Sullivans can read the true cost of a mortgage and shop it against any other lender on equal terms. Learning to read it is the single most valuable money skill in the entire home-buying process.

The document is called the Loan Estimate, and it exists because of a hard-won consumer-protection rule usually shortened to TRID (the "TILA-RESPA Integrated Disclosure" — two older laws stitched into one form). Before it existed, every lender sent its own confusing pile of paperwork and comparing two mortgages was nearly impossible; fees hid in the fog. Since 2015, every lender offering the same type of loan has to disclose it on the exact same three-page form, in the exact same order, with the exact same words. That standardization is the whole point: it turns "trust me" into "compare us." The Sullivans are going to use it to do exactly that — they'll get a Loan Estimate from Lakefront Lending (we'll call them Lender A) and a second one from Cuyahoga Home Loans (Lender B), and by the end they'll know which loan is genuinely cheaper, which is not the one with the lower rate on the sign.

Here's the ground we'll cover, and it follows the form itself. First, what the Loan Estimate is and the rules that force it into the Sullivans' hands — the three-day delivery clock and what triggers it. Then a page-by-page walkthrough of the whole form: page 1 (the loan and the monthly payment), page 2 (every closing cost, itemized), and page 3 (the comparisons that let you shop). Along the way we'll take apart the anatomy of closing costs — which are truly yours to shop and which aren't — and the three numbers people confuse constantly (the note rate, the APR, and the total-interest percentage). Then the payoff: comparing the two lenders' estimates on true cost, the tolerance rules that stop a lender from quietly raising fees on you, the predators who work this exact moment, and where to turn if something's wrong. The Sullivans' loan is the one they've carried since Lesson 13 — a $285,000 home, 5% down, a $270,750 loan, 30-year fixed — and every number on their Loan Estimate reconciles to it. A companion document, the Closing Disclosure, arrives three days before closing to confirm all of this; that's the LE's twin, and it's the whole of Lesson 17. This lesson is about the estimate that comes first — the one you shop with. It starts with what it actually is.

A Loan Estimate is a three-page, federally standardized form that a mortgage lender must give you after you apply, laying out the loan's terms, your projected monthly payment, and every cost to close — in a fixed format designed so you can compare it line-for-line against any other lender's. Four features define it, and each one changes how the Sullivans should treat the paper in their hands.

First, it is standardized — that is the source of its power. Because Lakefront and Cuyahoga must both use the identical form, with Section A in the same place meaning the same thing, the Sullivans can lay the two estimates side by side and compare them directly. No lender can bury a fee somewhere unexpected or rename it to look smaller; the form's order is fixed by regulation. This is what makes real shopping possible, and it's why the rest of this lesson is essentially a guided tour of one form.

Second — and this is the fear-disarming part — it is an estimate, not an approval and not a contract. Receiving a Loan Estimate does not mean the Sullivans have been approved for the loan, and it absolutely does not obligate them to take it. The form even says so on page 3: signing it "only confirms that you have received this form. You do not have to accept this loan." So the natural instinct — "I got the Loan Estimate, I guess I'm locked in with this lender now" — is exactly wrong. The Sullivans owe Lakefront nothing for sending it (beyond, at most, a small credit-report fee), and they can walk to Cuyahoga or anyone else with no penalty. The Loan Estimate is not a cage; it's a quote.

Third, it is triggered by a specific, minimal application. A lender must issue a Loan Estimate once you've given six pieces of information — your name, your income, your Social Security number (so they can pull credit), the property address, an estimate of the property's value, and the loan amount you're seeking. That's it. Those six items are the legal definition of a mortgage "application," and here's the protective teeth in it: the lender cannot demand more paperwork before handing over the estimate, and cannot charge you any fee beyond the cost of a credit report until you've received the Loan Estimate and told them you want to proceed. (There used to be a vague seventh "anything else the lender wants" catch-all — it was deliberately removed in 2015 so lenders couldn't stall the estimate by demanding endless documents.) For the Sullivans, this means they can collect estimates from several lenders cheaply and fast, just by handing each the same six facts.

Fourth, it is mostly a snapshot that can shift — but only in specific ways, and this is the single most useful thing to understand before you shop. Until the Sullivans "lock" their interest rate, only three things on the Loan Estimate can actually change: the interest rate, the points, and the lender credits (money the lender puts toward your closing costs in exchange for a higher rate — the mirror image of points, taken apart in §10). Everything else — the estimated closing costs — is held to strict limits we'll cover in the tolerance section, and the whole estimate carries an expiration date and time after which the lender can revise it. So those three items are the real negotiating levers, and the rest is a promise with legal weight behind it. Knowing that reframes the whole document: it's not "here are some numbers that'll probably change," it's "here is a quote the lender can be held to, with three moving parts you can negotiate." How that quote actually reaches the Sullivans — and how little time the lender has to produce it — is the delivery clock, and it's §2.

The Loan Estimate isn't something the Sullivans have to beg for or wait weeks on — the law puts it on a tight clock, and knowing the clock is what lets them shop several lenders at once without losing time. There are really two deadlines, one for getting the estimate and one for getting to closing, and it's worth keeping them straight because they use different day-counts.

The practical upshot for the Sullivans is a simple shopping rhythm. On a Monday, they give the same six facts — names, income, Social Security numbers, the Parma address, the ~$285,000 value, and the $270,750 loan amount — to Lakefront, to Cuyahoga, and to their own credit union. By roughly Thursday of the same week, three Loan Estimates are in hand, each on the identical form, ready to compare. Because they got them close together and for the same loan, the comparison is apples-to-apples. And because the estimate has to reach them at least seven business days before any closing, there's structurally no way a lender can rush them to the table before they've had time to read it. The clock is on the borrower's side.

One honest note on why "business day" appears twice with a subtle difference: the three-day delivery deadline counts days the lender's office is open, while the seven-days-before-closing and the mailing presumption count all calendar days except Sundays and federal holidays. You don't need to memorize the distinction — the takeaway is just that these windows are measured in business days, not calendar days, so they stretch a little further than they first appear, which works in your favor. With the estimate in hand and time to read it, the walkthrough begins where the form does: page 1, the loan and the payment. That's §3.

Page 1 answers the two questions a borrower cares about most: what am I borrowing, and what will I pay every month? It has three parts — a header of loan facts, a "Loan Terms" box, and a "Projected Payments" box — and it ends on two summary numbers ("Costs at Closing") that the whole of page 2 exists to explain. Here is the Sullivans' page 1 from Lakefront Lending, in full:

Read the header first, because it fixes what loan this is. It's issued to Brandon & Katie Sullivan for the Parma property, a $285,000 sale price, a 30-year term, "Purchase" (not a refinance), a "Fixed Rate" "Conventional" product, and — importantly — "Rate Lock: NO." That last one matters: because the rate isn't locked yet, the interest rate, points, and lender credits can still move (the §1 rule), and the estimate itself expires at the date and time shown. The header is how the Sullivans confirm both estimates they're comparing describe the same loan; if one said "30 years" and the other "Product: 5/1 Adjustable Rate," comparing their costs would be meaningless.

The "Loan Terms" box is next, and the genuinely important part isn't the dollar figures — it's the right-hand column, "Can this amount increase after closing?" The Loan Amount is $270,750 (what they're borrowing), the Interest Rate is 6.75% (the yearly price of the money — an illustrative scenario rate; in mid-2026 the 30-year-fixed survey average sits a bit below this, around 6.4%), and the Monthly Principal & Interest is $1,756 (the part of the payment that repays the loan itself). For a fixed-rate loan like the Sullivans', every answer in that increase column is a reassuring "NO" — the loan can't grow, the rate can't rise, the principal-and-interest payment is locked for 30 years — and the "Prepayment Penalty: NO / Balloon Payment: NO" lines below confirm there's no charge for paying early and no sudden lump sum due. On an adjustable-rate loan those answers would flip to "YES," pointing to two extra tables a fixed loan like the Sullivans' simply omits — the Adjustable Payment (AP) and Adjustable Interest Rate (AIR) tables, which spell out how much and how often the payment and rate can move. So that increase column is the fastest read on the whole page: it tells you, in a glance, whether the loan has teeth. Every "NO" is a promise the loan will behave.

The "Projected Payments" box turns principal-and-interest into the real monthly number, and it does something clever: it shows the payment in phases. For Years 1–11, the Sullivans pay $1,756 in principal and interest, plus $113 in mortgage insurance (PMI — the premium they pay for putting only 5% down, covered in Lesson 14), plus $595 in estimated escrow (the lender collecting their property taxes of ~$475 and homeowner's insurance of ~$120 each month) — a total of $2,464 a month. Then, around year 12, something good happens: the PMI drops off (the "+ 113" becomes "+ —"), because by then they've paid the loan down enough to cross the equity threshold where PMI ends, and the payment falls to $2,351. That drop-off is a real feature the form is showing them — it's why the box has two columns — and the mechanics of removing PMI are Lesson 18. The escrow line carries a quiet warning too: it "can increase over time," because taxes and insurance premiums rise even though the principal-and-interest never will.

One easy-to-miss line under that box matters more than its small type suggests: the "Estimated Taxes, Insurance & Assessments" note, which shows $595 a month and answers "In escrow? YES" for property taxes and homeowner's insurance. That "YES" means the lender bundles those costs into the monthly payment and pays the bills for the Sullivans out of an escrow account — so their $2,464 really does cover the whole housing cost. If it had said "In escrow? NO" for some item, the estimated total monthly payment would not include it, and they'd owe it separately — a common way a payment looks smaller than it really is. The footnote spells this out: "You must pay for other property costs separately." Always check what's in escrow and what isn't.

Page 1 ends on the two numbers everything else builds toward: "Estimated Closing Costs" of $10,400 (helpfully broken down right there as $4,400 in Loan Costs plus $6,000 in Other Costs, minus $0 in Lender Credits) and "Estimated Cash to Close" of $18,800. Those are the punchline of the whole form — the total one-time cost of getting this loan, and the cash the Sullivans need to bring. But page 1 only summarizes them; it's page 2 that itemizes every dollar inside that $10,400 and shows exactly how the $18,800 is calculated. Before we open that itemized page, it helps to know the four families those costs fall into — because only some of them are the Sullivans' to shop. That's §4.

"Closing costs" sounds like one lump, but it's really four different families of cost with four different owners — and the reason that matters is intensely practical: you can only shop some of them. Spend your energy trying to negotiate the wrong ones and you'll get nowhere; spend it on the right handful and you can genuinely save. The Loan Estimate sorts every fee into lettered sections (A through H) precisely so you can tell which is which. Here's the whole $10,400 sorted by who sets the price:

The first family is lender fees (Section A) — the charges the lender keeps for making the loan: the Sullivans' $1,700 covers a $500 application fee and a $1,200 underwriting fee, and this is also where discount points would sit if they were buying the rate down (they're not — it reads "0.000% of Loan Amount"). You don't shop these fee-by-fee; you shop them by comparing whole Loan Estimates across lenders, because a lender with a fat origination charge often has a lower rate, and vice versa. The whole estimate is the unit of comparison, not the single line.

The second family is third-party services the lender chooses (Section B) — outside services the loan requires but you can't shop, because the lender picks the vendor: the Sullivans' $750 covers the appraisal ($600, to confirm the house is worth the price), a credit report ($50), a flood determination ($25, checking whether flood insurance is required), and a tax service fee ($75). These are real costs for real work, and while you can't shop them, they shouldn't vary wildly between honest lenders. The third family is the one that matters most for shopping: third-party services you can choose (Section C). The Sullivans' $1,950 here — lender's title insurance $1,050, a settlement/closing fee $450, a survey $250, and a pest inspection $200 — are services the loan needs but where the law gives you the right to pick the provider. The lender must hand you a written list of suggested providers, and you're free to use someone cheaper. This is the one bucket where calling around actually moves the number.

The remaining families aren't really "shoppable" in the same way, because they're set by outside forces, not the lender. Government fees (Section E) — the Sullivans' $150 in recording fees, plus transfer taxes where they apply — are set by the county. Prepaids (Section F, $2,291) and the initial escrow deposit (Section G, $2,735) are set by the calendar and by escrow math, which §6 takes apart. Trying to negotiate those is wasted breath. So the honest map of where shopping pays off is narrow and worth memorizing: Section C services, the optional owner's title policy (more on that in §6), and your choice of homeowner's-insurance company. Everything else is either compared lender-to-lender (A and B) or simply fixed (E, F, G). With that map in hand, page 2's itemized list stops being a wall of fees and becomes a sorted, readable thing. That's §5.

Page 2 is where the $10,400 comes apart into its pieces, and it's laid out in two columns — "Loan Costs" on the left (Sections A, B, C, totaling D) and "Other Costs" on the right (Sections E, F, G, H, totaling I) — with the grand total (J = D + I) below, and then the cash-to-close math. Here is the Sullivans' whole page 2; we'll read the Loan Costs column now and the Other Costs column in §6:

Section A, "Origination Charges" ($1,700), is what Lakefront charges to create the loan, and the very first line is the one to check: "0.000% of Loan Amount (Points)." That $0 means the Sullivans are not paying any discount points to buy down their rate — they're taking Lakefront's rate as-is. If they wanted a lower rate, they could pay points here (each point costs 1% of the loan, or $2,708 on their $270,750), which is exactly the lever Lender B pulls in §11. Below it sit the lender's own fees: a $500 application fee and a $1,200 underwriting fee — real charges for processing and evaluating the loan, and the numbers to compare hardest against other lenders, because this is money the lender simply keeps. A borrower comparing two estimates should look at Section A first: it's the most lender-controlled, most negotiable pile on the page.

Section B, "Services You Cannot Shop For" ($750), is the honest label for third-party costs the lender assigns: the $600 appraisal (an independent valuation confirming the Parma house is worth what the Sullivans agreed to pay — protection for them and the lender), the $50 credit report, the $25 flood determination, and the $75 tax service fee. "Cannot shop" is not a red flag here — it just means the lender picks these vendors, and the amounts should be ordinary. What matters is that the Sullivans know these are pass-through costs for genuine services, not padding. Section C, "Services You Can Shop For" ($1,950), is the opposite and the important one: the loan requires these, but the Sullivans get to choose the provider. The $1,050 lender's title insurance (which protects the lender's stake if someone later challenges who really owns the house), the $450 settlement/closing fee (paid to whoever runs the closing table), the $250 survey, and the $200 pest inspection are all shoppable. Lakefront had to give the Sullivans a written list of suggested providers, but they can call around — a cheaper title company or pest inspector directly lowers this number, and it's the one section where their legwork pays.

Add the three and you get D, "Total Loan Costs" of $4,400 ($1,700 + $750 + $1,950) — the whole cost of the loan itself, separate from the costs of the property and the calendar. That $4,400 is the number to compare most carefully against Lender B, because it's the part most under a lender's control. But it's only part of the story: the other $6,000 of the Sullivans' closing costs isn't about the loan at all — it's about prepaying property expenses and seeding an escrow account. That's the "Other Costs" column, and its two trickiest members, prepaid interest and the initial escrow deposit, deserve their own section. That's §6.

The right column of page 2, "Other Costs," is where new homeowners get blindsided, because these aren't fees for the loan — they're money for the property and the calendar, and they're the reason "closing costs" are bigger than most people expect. The Sullivans' Other Costs total $6,000, and they break into four sections (E, F, G, H). Two are simple; two — prepaid interest and the escrow deposit — trip up almost everyone, so we'll slow down on them.

Start with the simple ones. Section E, "Taxes and Other Government Fees" ($150), is the county's recording fee — the cost of officially entering the sale and mortgage in the public record (in some places a "transfer tax" would swell this line, but Ohio's is modest and often the seller's to pay). Section H, "Other" ($824), holds the optional owner's title insurance policy. This one's worth a beat: the $1,050 lender's title insurance back in Section C protects the lender; this owner's policy protects the Sullivans' own equity if someone later surfaces claiming a right to the property. It's optional, which is why it sits in "Other" — but for a modest one-time premium it protects the largest purchase of their lives, and it's usually far cheaper bought at the same time as the lender's policy. That "(optional)" tag is the form telling them it's a real choice, not a required fee.

Now the two that confuse everyone. Section F, "Prepaids" ($2,291), is money paid in advance for things you'd owe anyway — here, a full year of homeowner's insurance ($1,440) and "prepaid interest" ($851). That prepaid-interest line is the classic head-scratcher, so here's exactly what it is: interest doesn't start neatly on the first of the month, so the lender charges the interest for the days between closing and the start of the first full payment period — a "per-diem" (per-day) charge. The math is grounded and worth seeing:

So the Sullivans' $851 in prepaid interest is simply 17 days of interest at $50.07 a day — real interest they genuinely owe, just collected up front. (A small but useful lever: closing near the end of the month shrinks this line, since there are fewer days left to prepay.) Section G, "Initial Escrow Payment at Closing" ($2,735), is the other confusing one, and it's not a fee at all — it's the Sullivans' own money, seeding their own escrow account. Remember from page 1 that the lender collects taxes and insurance monthly and pays those bills for them; to make sure the account has enough when the first big tax bill comes due, the lender collects a few months up front — here, 5 months of property taxes ($2,375) and 3 months of homeowner's insurance ($360). This money isn't lost; it's their cushion, sitting in their account to pay their bills. There's a legal ceiling on how much the lender can pad this cushion, and it's worth knowing:

Add E, F, G, and H and you get I, "Total Other Costs" of $6,000. One quick note the Sullivans' insurance line illustrates: they're paying 12 months of homeowner's insurance up front, but that number varies — many lenders collect 6 to 12 months, not a fixed year, so it's not a sign of anything wrong. Now the two columns combine: Total Loan Costs D ($4,400) plus Total Other Costs I ($6,000) equals J, "Total Closing Costs" of $10,400 — the exact figure that appeared on page 1. The form has now shown its work. What's left on page 2 is turning that $10,400 into the actual cash the Sullivans must bring to the table — which involves their down payment, the earnest money they already paid, and the credit they negotiated from the seller. That cash-to-close math is where people miscount, so it gets its own section. That's §7.

Total closing costs ($10,400) and cash to close ($18,800) are different numbers, and confusing them is how buyers get an unpleasant surprise a week before closing. Closing costs are the fees. Cash to close is the total money the Sullivans actually have to wire to the closing table — which includes their down payment, adjusts for money they've already put down, and subtracts credits. The Loan Estimate lays this out in a "Calculating Cash to Close" table, and the arithmetic is worth walking slowly, because there's a trap in it:

Here's the build, top to bottom. Start with the down payment: 5% of the $285,000 home is $14,250 — the Sullivans' own equity going in. Add the $10,400 in total closing costs. That's $24,650 so far. Now subtract two things. First, the $2,850 "deposit" — this is the earnest money they already paid weeks ago when their offer was accepted (1% of the price, a good-faith deposit held in escrow), so it's credited back here because they've already handed it over. Second, the $3,000 seller credit — a concession they negotiated in the purchase offer, where the seller agreed to cover $3,000 of the buyer's closing costs to close the deal. Subtract both and the Loan Estimate's "Estimated Cash to Close" lands at $18,800. (One note for later: this "Calculating Cash to Close" table is the purchase version. A refinance or any deal without a seller uses a slightly different "alternative" version of the same table — it starts from the loan amount and nets out the old loan's payoff instead of a down payment — but its job is identical, turning the costs into the cash that actually changes hands.)

And here is the trap, the reason the widget separates two numbers. That $18,800 is the cash owed at the closing table — but it is not the total cash the Sullivans had to come up with, because it already subtracts the $2,850 they paid earlier. The real question their savings had to answer is the total: down payment $14,250 plus closing costs $10,400 minus the $3,000 seller credit equals $21,650 of cash the deal requires. They've already paid $2,850 of it (the earnest money), leaving $18,800 due at closing — which is exactly the LE figure. Both numbers are true; they answer different questions. "$18,800" is what to wire on closing day; "$21,650" is what the whole purchase drained from their accounts. A buyer who budgets only for the $18,800 line and forgets they already spent $2,850 has actually under-counted what the house cost them.

Where does the $21,650 come from? The Sullivans had $22,000 in savings and received a documented $5,000 gift from Katie's parents (the kind of gift that needs a paper trail, which they set up back in Lesson 15) — $27,000 available. After the $21,650 the purchase consumes, they're left with about $5,350 in reserves. That reserve number matters as much as any fee on the form: it's the difference between owning a home with a small cushion and owning one with nothing left if the furnace dies in month two. Reading cash to close correctly — total needed, not just the closing-table line — is what lets a buyer know whether they can actually afford to close, and with how much breathing room. That completes page 2. Page 3 turns from "what does this loan cost" to "how does it compare" — and it's where the three most-confused numbers on the whole form live. That's §8.

Page 3 exists for one purpose the first two pages can't serve: comparing this loan against others. It has a "Comparisons" box with three carefully chosen numbers, a list of "Other Considerations," and the signature line. Here is the Sullivans' page 3:

The "Comparisons" box holds three measures, and they're designed to be used against another lender's page 3, not admired in isolation. The first, "In 5 Years," shows two numbers: $116,533 (everything the Sullivans will have paid — principal, interest, mortgage insurance, and loan costs — through month 60) and $16,582 (how much of their loan balance they'll have actually paid off in that time). This is the most honest shopping number on the form, because it reflects a realistic holding period rather than the full 30 years, and we'll see in §11 why that matters enormously. The second is the Annual Percentage Rate (APR) at 7.24%, which the form is careful to label "your costs over the loan term expressed as a rate. This is not your interest rate." The third is the Total Interest Percentage (TIP) at 133%. Those last two — APR and TIP — are the numbers borrowers misread most often, so §9 is devoted to them; for now, just note where they live and that they're comparison tools.

The "Other Considerations" list is the fine print that turns out to matter, and it's worth actually reading rather than skimming. The Sullivans' says the lender may order the appraisal and will give them a copy; that assumption is not allowed (a future buyer can't take over this loan on its original terms); that homeowner's insurance is required but they may choose the company; that a late payment costs "5% of the monthly principal & interest" if more than 15 days late (about $88 on their $1,756 payment — a real number worth knowing); a sober note that "you may not be able to refinance this loan," a reminder that refinancing is never guaranteed; and that the lender intends to service the loan itself (meaning the Sullivans will send payments to Lakefront, not a stranger — though servicing can be sold later, which is Lesson 18's territory). None of these are cost lines, but each is a term of the deal, and the late-payment percentage in particular is one people wish they'd read.

Finally, the "Confirm Receipt" signature line, which closes the loop on the §1 reassurance in the clearest possible language: "By signing, you are only confirming that you have received this form. You do not have to accept this loan because you have signed or received this form." If the Sullivans sign page 3, they've acknowledged delivery — nothing more. It is not acceptance, not a commitment, not a step they can't walk back. The whole three-page form, read end to end, is a quote with legal weight and an explicit invitation to shop it. So let's use it for that — starting with the two numbers on page 3 that mislead the most, the APR and the TIP. That's §9.

Three percentages appear on the Loan Estimate — the note rate (6.75%), the APR (7.24%), and the TIP (133%) — and they look similar enough that borrowers treat them as versions of the same thing. They are not. Each answers a different question, and reading them wrong leads straight to picking the wrong loan. Here's what each one actually measures:

The note rate (or interest rate), 6.75%, is the simplest: it's the yearly price of borrowing the money, interest only, no fees. It's what builds the Sullivans' $1,756 principal-and-interest payment, and it's the number lenders advertise because it's the smallest. But it deliberately leaves out the fees, which is why it can't be the whole story. The APR, 7.24%, fixes that: it folds the finance-charge fees — the origination charge, any points, and the mortgage insurance — back into the rate and re-expresses the whole cost as a single yearly percentage. Because it includes those fees, the APR is always at least as high as the note rate, and usually a bit higher; the gap between 6.75% and 7.24% is essentially the fees, spread across the loan and re-quoted as a rate. This is genuinely useful, and it's why the rule of thumb is "compare APR to APR" — two lenders' APRs put their rates and their fees on one comparable number, so a low rate with fat fees and a higher rate with no fees can finally be measured against each other.

The TIP, 133%, is the one that shocks people, because it looks enormous next to 6.75% — and the key is that it is not a rate at all. The Total Interest Percentage is the total amount of interest paid over the entire 30-year loan, expressed as a percentage of the amount borrowed. The Sullivans borrow $270,750 and, over 30 years at 6.75%, pay about $361,000 in interest — and $361,000 is roughly 133% of $270,750. That's all the TIP is: a lifetime total, not a yearly rate. So the single most important thing about the TIP is a warning — never read it next to the note rate or the APR as though they share a unit. "6.75% versus 133%" is meaningless; they measure different things (a yearly price versus a lifetime sum). The TIP's real value is the gut-punch of scale: it shows that financing a house for 30 years means paying more in interest than the house's loan itself. That's a reason to consider extra principal payments someday, not a number to compare against a rate.

Now the catch that makes APR genuinely misleading, and it's the hinge of the next two sections. Both the APR and the TIP are calculated assuming the borrower keeps the loan for its full 30-year term. That assumption quietly flatters a loan that charges big upfront fees to buy down the rate: spread a $3,000 fee across 30 years and it barely nudges the APR, so the low-rate/high-fee loan looks great by APR. But if the Sullivans sell or refinance in year 5 — which first-time buyers very often do — they paid that fee up front and never lived with the loan long enough to earn it back through the lower rate. So the loan with the lower APR is not automatically the cheaper loan for them. This is exactly why the form also shows "In 5 Years," and exactly why the right comparison is over your own expected holding period. One place this shows up directly is the choice to pay points or take lender credits — the dial that trades upfront cash against the rate. That's §10.

The Sullivans' Loan Estimate shows "0.000% of Loan Amount (Points)" and "$0" in lender credits — meaning they took Lakefront's rate straight, paying nothing extra to lower it and getting nothing extra to offset costs. But those two lines are a dial they could turn in either direction, and understanding it is what lets a borrower see through a too-good rate or a too-good "no cost" offer. The dial trades upfront cash against the interest rate, and it has two ends:

On one end are discount points: the Sullivans could pay an extra fee up front to buy their rate down. Each "point" costs 1% of the loan amount — $2,708 on their $270,750 — and lowers the rate by some amount. (How much? Here's an honest caveat the industry rarely volunteers: the old "one point lowers the rate about 0.25%" is only a rough rule of thumb with no rule behind it. The real cut varies by lender and by day, so the move is to ask each lender for the actual rate at 0, half a point, and a full point, and compare.) A sobering finding worth knowing before you reach for points: lender research (Freddie Mac) has found they often don't pay off for the average buyer, who keeps a mortgage only about five years — frequently not long enough to clear the break-even. Points appear in Section A of the form. They're a bet on staying put: you pay more today for a lower payment every month after, so they pay off only if you keep the loan long enough for the monthly savings to exceed the upfront cost. On the other end are lender credits: the mirror image. The lender gives you money toward your closing costs — reducing your cash to close — in exchange for a higher rate. It's "negative points," and it appears on the Lender Credits line in Section J. Credits are a bet on leaving: you pay less today and more each month, which wins if you sell or refinance soon.

The far end of the lender-credit side has a marketing name worth decoding: the "no-closing-cost" loan. It sounds like the fees vanish. They don't — they can't, because someone did real work to make the loan. What actually happens is the lender covers the closing costs with a credit and charges a higher interest rate to pay for it, or rolls the costs into the loan balance so you pay interest on them for 30 years. So "no-closing-cost" is precisely, mathematically, the lender-credit dial turned all the way up. That doesn't make it a scam — it's genuinely the right choice for someone who'll move or refinance within a few years, since they escape before the higher rate adds up. But for someone who'll stay 20 years, it's expensive, because that higher rate compounds the whole time. The rule the whole section points to is the same one the APR warning gave: whether points, par, or credits is right depends entirely on how long you'll keep the loan — which is exactly the calculation the Sullivans have to make when they compare their two real Loan Estimates. That's §11.

Now the payoff. Everything so far — the standardized form, the itemized costs, the APR, the points dial — exists to let a borrower do one thing: put two lenders' Loan Estimates side by side and find the genuinely cheaper loan. The Sullivans got a second estimate, from Cuyahoga Home Loans (Lender B), for the identical $270,750 loan. Cuyahoga's rate is lower — 6.50% versus Lakefront's 6.75% — which makes it look like the obvious winner. But a lower rate always comes from somewhere, and the somewhere is upfront cost. Here are the two estimates on the numbers that decide it:

Watch where the green — the winner of each row — sits, because it tells the whole story. Lender B wins the headline numbers: a lower rate (6.50% vs 6.75%), a lower monthly payment ($1,711 vs $1,756, saving about $45 a month), a lower APR (7.10% vs 7.24%), and a lower TIP (128% vs 133%). By every number a borrower usually looks at, Cuyahoga is cheaper. But look at the cash-upfront row: Lender B costs $13,400 to close versus Lakefront's $10,400 — $3,000 more, because Cuyahoga baked a full discount point ($2,708) plus heavier fees into that lower rate. And now the row that flips everything: "Total paid — first 5 years." Over five years, Lakefront (A) costs the Sullivans $115,765 and Cuyahoga (B) costs $116,079 — meaning Lakefront, the loan with the higher rate, is actually the cheaper loan for the first five years. (These totals count all the upfront closing costs plus five years of principal and interest, so the $3,000 cash gap between the two lenders is fully captured — a slightly different basis than page 3's "In 5 Years" line, which by the form's definition counts principal, interest, mortgage insurance, and loan costs only. Either way, A wins the five-year race.) Only in the "full 30 years" row does B finally pull ahead, by about $13,100.

The reconciling number is the break-even, and it's the single most useful figure in the comparison: 67 months, about 5.6 years. That's how long the Sullivans must keep the loan before Cuyahoga's $45 monthly savings adds up to the $3,000 extra they paid upfront. Keep it longer than 5.6 years and B wins; sell or refinance sooner and A wins. This is the §9 warning made concrete: B has the lower APR and lower TIP precisely because those numbers assume all 30 years, but the Sullivans are a young family in their first home — the odds they move, upsize, or refinance before 5.6 years are real. For them, the "worse" loan on every headline may well be the cheaper one in practice. The lesson isn't "always pick A"; it's "the headline rate is not the price — compare the total cost over the years you'll actually keep it." A quarter-point lower rate is worthless if it costs $3,000 to buy and you leave before you earn it back.

Two practical points make this shopping real rather than theoretical. First, the CFPB's own advice is to get Loan Estimates from at least three lenders and compare them apples-to-apples — same loan amount, same day, same assumptions — and it costs almost nothing to do: the only fee a lender may charge before you commit is for a credit report, and all your mortgage credit inquiries within a 45-day window count as a single inquiry, so shopping barely dents your score. Requesting an estimate commits you to nothing. Second, competing estimates are leverage. A better Section A from one lender is a concrete thing to take back to your preferred lender and ask them to match — lenders expect it, and shopping this way saves buyers hundreds to over a thousand dollars a year on the loan. The Sullivans' two estimates aren't just a comparison; they're a negotiating position. But there's a fear lurking under all this shopping: what stops a lender from quoting a nice low estimate and then quietly raising the fees at closing? The answer is a set of rules with real teeth, and it's §12.

Here's the fear that makes people distrust the whole document: if it's just an "estimate," what stops Lakefront from quoting the Sullivans $10,400 to win their business and then billing $13,000 at closing? The answer is one of the strongest consumer protections in the mortgage rules, and it's the reason the Loan Estimate is worth trusting. The law sorts every fee into "tolerance" buckets that cap how much it can rise between the Loan Estimate and the final Closing Disclosure. Some fees can't rise a penny. Here are the three buckets, and which fees fall where:

The zero-tolerance bucket is the strongest promise: these fees cannot increase at all. It holds the lender's own charges — origination and underwriting (Section A) — the services you couldn't shop for (Section B), and transfer taxes. In other words, the money the lender controls is the money the lender is held to exactly. If Lakefront's Loan Estimate says $1,700 in origination, the Closing Disclosure must say $1,700 (or less); a single dollar more is a violation. The second bucket, "10% cumulative," allows some movement but caps it: recording fees, plus the shoppable services (Section C) if the Sullivans picked a provider from the lender's written list, can rise — but the whole group together can't exceed its estimate by more than 10%. The crucial subtlety is that the 10% is measured on the sum of the group, not each fee: one fee inside it can jump much more than 10% and still be legal, as long as the group's total stays within 10%. The third bucket, "no set limit," covers the fees genuinely outside the lender's control — prepaid interest, homeowner's insurance, the escrow deposit, property taxes, and any shoppable service where the Sullivans went off the lender's list. These can change without a fixed cap.

Three nuances turn this from trivia into protection, and they're exactly the things a sharp borrower uses. First, "no set limit" is not "anything goes." A no-limit fee can only rise if the lender's original estimate was made in good faith with the best information available — so if Lakefront lowballed the Sullivans' property taxes to make the estimate look cheap and then "corrected" it upward at closing, that fee snaps back to zero tolerance and the increase is a violation. Second, a fee's bucket can depend on the Sullivans' own choice: a title or pest fee is in the 10% group if they pick from the lender's list, but has no limit if they shop off-list — so shopping around trades a cap for freedom, a real trade-off to know about. Third, a decrease in the lender credit shown on the estimate is itself a zero-tolerance violation — the lender can't quietly shave the credit it promised. And there's an escape hatch that's legitimate: a genuine "changed circumstance" — a surprise the appraisal turns up, a change in what the Sullivans qualify for, a request they make — lets the lender issue a revised Loan Estimate that resets the tolerances. That's the honest flexibility in the system; it's not a loophole to raise fees for no reason.

The enforcement is what makes it real: if a lender exceeds a zero- or 10%-tolerance limit, it must refund the excess to the borrower — and send a corrected Closing Disclosure — generally within 60 days of closing. The refund can come as cash, a reduction of the loan principal, or a lender credit, but it must come. So the Sullivans' defense is concrete and simple: keep the Loan Estimate, and when the Closing Disclosure arrives (three business days before closing, per §2), lay them side by side and check every line. A zero-tolerance fee that rose a dollar, or a 10% group that blew its cap, is money the lender owes back. This is why the "estimate" deserves trust: it's an estimate with a legal spine. Which brings us to the people who try to work around all of this — the closing-cost predators. That's §13.

Most mortgage lenders are legitimate, and the Loan Estimate's standardization and tolerance rules make the honest ones easy to compare. But the same complexity that the form tames is exactly what a dishonest operator exploits — and the moment a stressed buyer is juggling forty line items and a moving date is when padding and switches slip through. Four patterns work this moment specifically, and every one of them is beaten by the same defense: comparing documents. Here they are, with a blame-free guide to reporting:

The first pattern is junk-fee padding — vague, official-sounding charges with no real service behind them, stacked into Sections A, B, and C to fatten the total. A second "processing fee" on top of underwriting, a "document preparation fee," an "administrative fee," a duplicate "application fee": each sounds plausible, and each may be pure margin. The tell is that they don't survive a direct question ("what does this fee actually pay for?") or a competing Loan Estimate that doesn't have them. The second pattern is the lowball-rate bait-and-switch — an advertised or phone-quoted rate that sits suspiciously below everyone else's, then the Loan Estimate arrives carrying points and heavy origination charges that quietly pay for that rate, or the rate simply isn't honored when it's time to lock. The tell is the §9 lesson: a rate means nothing without its Section A, so you compare the whole estimate, never the number in the ad.

The third pattern is the "no-closing-cost" trap, and it's the §10 dial dressed up as a gift. "We'll cover your closing costs!" almost never means the costs vanish — they're baked into a higher rate the borrower pays for 30 years, or rolled into the balance to accrue interest. It's not always a bad deal (it's genuinely good for a short stay), but sold as "free" to someone who'll keep the loan for decades, it's a quiet, expensive switch. The tell is to ask for the same loan with and without the "no-cost" structure and compare the rates, then do the break-even. The fourth pattern is tolerance creep — the estimate looks clean, but fees drift upward on the final Closing Disclosure, betting the buyer is too rushed to compare the two documents line by line. This one is beaten cold by §12: keep the Loan Estimate, check it against the Closing Disclosure in the three-day window, and any zero-tolerance rise or over-cap 10% group is a refund the lender owes. Across all four, the pattern is the same — they survive on the borrower not comparing, and a second Loan Estimate plus the final Closing Disclosure is the entire defense.

And if one of these got past the Sullivans, reporting it is a civic act, not a confession — the how-to-report block in the panel lays out where to go (start with the lender in writing, then the state attorney general or mortgage regulator, the CFPB, HUD for settlement issues, and the FTC for deceptive ads), what to have ready (the Loan Estimates, any advertised rate, the provider list, the Closing Disclosure, and every email), and why it's worth doing (a tolerance violation caught before closing must be cured, and a complaint builds the record regulators use to stop the operation for the next family). Being targeted by one of these is evidence of the scheme's design, not of any failing by the buyer — which is exactly the frame the next section holds for anyone the warning reached too late. That's §14.

Maybe this lesson arrived after the Sullivans — or you — already accepted a Loan Estimate that, in hindsight, is thick with fees, or committed to a lender before shopping anyone else, and now it feels too late and too tangled to change. If that's where you are, the first thing to hear is that this is an ordinary story, not a personal failure — and the second is that a Loan Estimate is the one point in the whole process where you have the most power and the most time to fix things.

Set the self-blame down first, because it's the thing that keeps people stuck. These are three dense pages, the numbers blur together, the words are unfamiliar, and the entire process runs on the urgency of not losing the house — lenders and the calendar both count on that rush. Being caught by fees isn't a sign you're careless or bad with money; it's a sign the form is genuinely hard and the moment is genuinely pressured. "I should have read it more carefully" points at the wrong culprit. The useful move is not shame; it's the concrete list of things you can still do, and there are more of them than most people realize.

Concretely: a Loan Estimate is not a commitment, so you can still shop right now — ask two or three other lenders for an estimate on the same loan, get each within three business days, and (because inquiries in a 45-day window count as one) barely touch your credit doing it. A better estimate is leverage to make your current lender match it, or a reason to switch. You can walk before closing — right up until you sign the final papers you can change lenders; you may forfeit an application or appraisal fee already paid, but that's usually far less than overpaying on rate and fees for 30 years. And the Closing Disclosure gives you a last, legally-protected checkpoint: you must receive it at least three business days before closing, so lay it beside your Loan Estimate and any zero-tolerance fee that rose, or a 10% group over its cap, is money the lender must refund. Even after closing, a tolerance violation can be curable for up to 60 days, and a free HUD-approved housing counselor can help you read what happened. One rushed or fee-heavy start is a setback, not a verdict — and the tools to fix it are the same ones this whole lesson taught. Several of those steps rest on rights and channels worth naming plainly, which is §15.

Several moves in the last two sections leaned on rights — the tolerance refund, the three-day window, the ability to complain — so this section names the recourse ladder for a Loan Estimate or closing-cost problem, along with an honest read of which rungs actually have force behind them in 2026. As with the high-cost-credit protections back in Lesson 10, the most dependable help is no longer only the federal agency you'd expect.

Start with the lender, in writing, because most problems genuinely end there. If a fee changed or looks wrong, ask exactly why and cite your Loan Estimate; a charge that broke a zero- or 10%-tolerance limit must be cured and refunded, generally within 60 days of closing, and a lender that knows you've spotted it usually fixes it rather than fight. If that fails, the next rung is your state — the state mortgage or banking regulator and the attorney general's consumer-protection office. This is worth emphasizing: with federal enforcement stretched thin (below), the state channel is often the fastest real answer, so treat it as a first-class option, not a fallback. The CFPB comes next — you can file at consumerfinance.gov/complaint or call 855-411-2372, and companies generally respond within about 15 days — but honesty requires the caveat: the CFPB's enforcement and supervision capacity has been sharply cut and is legally contested through 2025 and 2026. File the complaint, because it still routes to the company and builds the record, but don't treat it as a guaranteed or fast backstop.

Below those sit three more channels, best used in parallel rather than in sequence. HUD is the front door for RESPA and settlement issues and for FHA loans, and also the gateway to housing counseling — though the RESPA enforcement authority itself actually sits with the CFPB now. The FTC (reportfraud.ftc.gov) is where deceptive mortgage advertising goes — the lowball rate, the false "no-cost" promise — under its truth-in-advertising rules. And a HUD-approved housing counselor is free or low-cost, unbiased help reading your Loan Estimate and figuring out next steps, reachable through consumerfinance.gov/find-a-housing-counselor or HUD at 1-800-569-4287. The through-line for 2026 is the one Lesson 10 drew: the rights are written into law and hold regardless of who's enforcing them this year, but the durable, responsive help is increasingly the channel closest to you — your state, and the documents in your own file. Which is the real lesson under all of this: keep every Loan Estimate, the provider list, the Closing Disclosure, and every email, because a borrower who can lay two documents side by side is a borrower who can prove what they were promised. With the products, the pages, the comparison, the traps, and the recourse all covered, only the wrap-up remains — the questions buyers actually ask, and a chance to compare two estimates yourself. That's §16.

"Is the Loan Estimate a bill, or am I locked in once I get one?" Neither. It's a standardized quote, and receiving or even signing it commits you to nothing — the form itself says signing "only confirms that you have received this form." You owe the lender nothing beyond, at most, a credit-report fee, and you can walk to another lender freely (§1).

"Why is my 'cash to close' different from my 'closing costs'?" Because they answer different questions. Closing costs ($10,400 for the Sullivans) are the fees; cash to close ($18,800) is the money you actually bring, which adds your down payment and subtracts what you've already paid (the earnest-money deposit) and any seller credit. And note the total cash the purchase drains ($21,650) is larger still — it counts the earnest money you paid earlier that the closing-table figure already nets out (§7).

"The APR is higher than the interest rate on my form — is that a mistake?" No — it's supposed to be. The interest rate (6.75%) is the price of the money alone; the APR (7.24%) folds the fees back in and re-states the whole cost as a yearly rate, so it's always at least the note rate. Compare APR to APR across lenders; just don't compare the APR to the TIP, which isn't a rate at all (§9).

"My TIP says 133% — am I paying 133% interest?!" No. The Total Interest Percentage is the total interest over all 30 years as a share of the amount borrowed, not a yearly rate — $270,750 borrowed, about $361,000 in interest, which is ~133% of the loan. It's a lifetime sum meant to show the scale of long-term interest, and it should never be read next to your 6.75% rate as if they matched (§9).

"One lender has a lower rate — isn't that just the better loan?" Not necessarily. A lower rate almost always comes with higher upfront cost. The Sullivans' Lender B is 0.25% lower but costs $3,000 more to close, so it doesn't break even for about 5.6 years — meaning the higher-rate loan is actually cheaper if they move or refinance before then. Compare the total cost over the years you'll really keep the loan, not the headline rate (§11).

"Which closing costs can I actually shop or negotiate?" A specific handful: Section C services (title, settlement, survey, pest), the optional owner's title policy, and your homeowner's-insurance company. The lender's own fees (Section A) you shop by comparing whole estimates across lenders; the government fees, prepaids, and escrow deposit (E, F, G) are set by the county and the calendar and won't move (§4).

"What stops a lender from quoting low fees and charging more at closing?" The tolerance rules. The lender's own fees can't rise a penny (zero tolerance); recording and on-list shoppable services can rise only 10% as a group; and anything the lender exceeds it must refund, generally within 60 days. Keep your Loan Estimate and check it line-by-line against the Closing Disclosure, which you get three business days before closing (§12).

"A lender is advertising a 'no-closing-cost' mortgage — is that real?" The costs are real; they're just moved. A "no-closing-cost" loan covers the fees with a higher interest rate (or by adding them to your balance), so you pay over time instead of upfront. It's a good deal if you'll move or refinance soon and a costly one if you'll stay for decades — do the break-even before you take it (§10, §13).

"How many lenders should I get Loan Estimates from — and won't all those credit checks wreck my score?" Get at least three, so you have real leverage and a true read on the market. The credit hit is tiny by design: all your mortgage-shopping inquiries within a 45-day window count as a single inquiry, and requesting an estimate commits you to nothing. Shopping this way saves buyers hundreds to over a thousand dollars a year on the loan (§11).

"Can a lender charge me fees just to hand me a Loan Estimate?" No. Beyond the actual cost of a credit report, a lender may not charge you anything until you've received the Loan Estimate and told them you want to proceed. So if someone demands an application fee or a "processing deposit" before they'll give you the estimate, that's both a red flag and against the rules (§1).

"The form lists both 'lender's title insurance' and an optional 'owner's title policy' — do I need both?" They protect different people. The lender's policy (required, in Section C) protects the bank's stake in the property; the owner's policy (optional, in Section H) protects your own equity if someone later surfaces claiming a right to the home. You're not required to buy the owner's policy, but for a modest one-time premium it guards the largest purchase of your life — and it's cheapest bought at the same time as the lender's (§6).

Now, a chance to do the core skill yourself — put two Loan Estimates side by side and find the one that's truly cheaper over the years you'd actually keep it:

That closes the lesson's content. Step back to where it began: three intimidating pages of fees that turned out to be the most powerful consumer-protection tool in the whole home-buying process. The Sullivans — and you — can now read all three pages of a Loan Estimate and know it's a quote with a legal spine, not a sales sheet; sort every closing cost into the four families and find the handful that are yours to shop; tell the note rate from the APR from the TIP, and know why the lowest rate isn't always the cheapest loan; compare two lenders on true cost over a real holding period; name exactly which fees a lender can't raise on you and the refund you're owed if it does; and spot the padding, the bait-and-switch, and the "no-cost" trap, with a place to turn if one lands. The estimate is where the money is decided — and reading it well is the difference between a mortgage you chose and one that was sold to you. What confirms all of it, three days before closing, is the Loan Estimate's final twin — the Closing Disclosure — and that's Lesson 17.

The fresh terms from this lesson, in one place. Each was defined where it first appeared; this is the quick-reference version.