Comparing Two Mortgage Offers Properly

Set the case up so the two offers are actually comparable

Two offers can only be compared when the loan amount and the term are identical, because both of those change the payment far more than any plausible rate difference does. Fix them: $400,000 over thirty years. Offer A quotes 6.25 percent and $2,000 of lender fees. Offer B quotes 5.875 percent and $12,000, the extra typically taking the form of a discount point plus a fatter origination charge. The annuity formula — derived in the article on how a loan payment is built, and not repeated here — gives $2,462.87 for A and $2,366.15 for B. B saves $96.72 a month and costs $10,000 more on day one.

The first thing worth noticing is how small the rate difference is next to the fee difference. Three-eighths of a percentage point buys $96.72 a month. Written over a whole year that is $1,160.64, and the fee gap is $10,000 — nearly nine years of the monthly saving. Rate shopping and fee shopping are not the same activity, and a broker who has moved the rate down while moving the fees up has not necessarily done you a favour. Whether they have is a question you cannot answer without deciding how long you intend to keep the loan.

The right metric: total cost of credit over a stated holding period

Almost nobody holds a thirty-year mortgage for thirty years. Houses are sold, jobs move, families change size, and rates fall far enough to justify refinancing. So the honest comparison is not over the term but over the period you actually expect to keep the loan, and the measure has three parts. Take the fees you paid at the start. Add every payment you will have made by the end of the period. Add the balance you still owe at that moment, because you will have to hand it over when you sell or refinance. Subtract the amount you borrowed. What is left is what the credit cost you over that window.

Run it at five years. Offer A: $2,000 of fees, sixty payments of $2,462.87, and a balance of $373,348.97, less the $400,000 borrowed, is $123,121.09. Offer B: $12,000 of fees, sixty payments of $2,366.15, and a balance of $371,637.76, less $400,000, is $125,606.83. A is cheaper by $2,485.73. Run it at twenty years and B is cheaper by $18,225.42. Run it to term and B is cheaper by $24,818.39. The comparison did not change; only the window did.

The simple payback is wrong, and it is wrong in the same direction every time

Divide the extra fees by the monthly saving and you get $10,000 divided by $96.72, which is 103.4 months: eight years and seven months. That figure appears in a great deal of mortgage advice and it is systematically too long. It counts only the cash flowing out and ignores what the two loans do to the balance. Because B carries a lower rate, more of every payment is principal, so the debt shrinks faster. After sixty payments B owes $1,711.21 less than A, and that gap is real money that will be handed to the lender at closing if you sell.

Put the balance back into the comparison and the crossover moves from month 103 to month 80 — six years and eight months, twenty-three months earlier. In other words the high-fee offer starts winning nearly two years sooner than the rule of thumb says. That matters because the decision usually sits in exactly that band: buyers who expect to move in three or four years should take the low-fee offer without hesitation, buyers who expect to stay ten should take the low-rate one, and it is the buyers in the six-to-eight-year zone who need the arithmetic done rather than guessed.

What the disclosed rate exists to do, and the one assumption it cannot escape

Regulators saw this problem long before you did, which is why every developed mortgage market makes lenders publish a single comparison figure alongside the nominal rate: the annual percentage rate in the United States, the taux annuel effectif global in France, the effektiver Jahreszins in Germany, the tasa anual equivalente in Spain, the taxa anual efetiva global in Portugal, the tasso annuo effettivo globale in Italy. The construction is the same everywhere. Take every cash flow the borrower makes, including the up-front costs, and find the rate at which their present value equals the amount advanced. It compresses rate and fees into one number precisely so they cannot be traded off invisibly.

The limitation is structural and no amount of regulation removes it: the disclosed rate is computed on the assumption that the loan runs to term. It is the answer to the bottom row of the table above and to no other row. On our two offers it will say that B is cheaper, which is true after year seven and false before it. That is not a defect in the disclosure; it is the disclosure answering the question it was designed to answer. Read it as what it is — a fee-inclusive ranking for a borrower who never moves — and supply the holding period yourself.

The same acronym does not contain the same costs on both sides of the Atlantic

This is where cross-border comparison quietly breaks. Under the European mortgage credit directive, the total cost of credit includes the cost of ancillary services — insurance above all — where taking the service is compulsory to obtain the credit on the terms marketed. In France that pulls two large items inside the disclosed rate that an American borrower will not find in theirs: the borrower's death and disability insurance, and the cost of the guarantee, whether that is a mortgage registration or a surety company. Notary fees on the purchase itself stay outside, as do early repayment charges.

Regulation Z goes the other way on a whole class of costs. For a transaction secured by real property, fees that are bona fide and reasonable in amount are excluded from the finance charge, and therefore from the American disclosed rate, in five listed categories: title examination, abstract, title insurance and survey; preparation of loan documents such as deeds and mortgages; notary and credit-report fees; appraisals and pre-closing condition inspections; and amounts paid into escrow. Those are precisely the third-party charges a buyer meets at an American closing table, which means the American figure ranks lender pricing well and settlement costs not at all.

One consequence is worth spelling out because it dwarfs everything else in this article. Where borrower's insurance sits inside the disclosed rate, its price is a comparison variable — and it is a big one. On our $400,000 loan, insurance at 0.34 percent of the initial capital each year costs $1,360 a year, $40,800 over thirty years. At 0.10 percent it costs $12,000. The difference, $28,800, is nearly three times the entire fee gap between the two offers we have spent this article comparing. In markets where you may take that cover from someone other than the lender, that single substitution is the largest single lever available to a borrower.

A comparison you can actually run in twenty minutes

Write down a holding period first, before you look at either offer, so the number is not chosen to justify a decision already taken. Three to five years for a first flat you expect to outgrow, seven to ten for a family home, longer only if you are genuinely certain. Then take from each offer four inputs and no others: the amount advanced, the term, the periodic rate, and every charge you must pay to get the loan. Charges you would pay in any case — a survey you would commission regardless, a tax on the purchase — belong in the cost of buying the house, not in the comparison of the two loans.

Then compute the three-part total for both offers at your chosen horizon and, since it costs nothing, at half and double that horizon as well. If the ranking is the same at all three, the decision is robust and you can stop. If it flips, you have learned the thing that actually matters: the answer depends on a fact about your life that no lender knows, and you should choose according to how confident you are about that fact rather than according to which rate is smaller.

Frequently asked questions

Why would a lower rate ever be the worse deal?

Because the rate is paid gradually and the fees are paid all at once. A rate advantage accrues month by month over the life of the loan; a fee disadvantage is complete on day one. Hold the loan long enough and the slow accumulation overtakes the immediate cost, which is exactly what happens at month 80 in our case. Hold it for less than that and it never catches up. The mistake is not in preferring a lower rate, it is in treating the two prices as if they were the same kind of thing.

Is the disclosed comparison rate useless then?

No, and it would be a bad reading of this article to conclude that. It does one job extremely well: it stops a lender advertising a low nominal rate and recovering the difference in charges you would not otherwise notice. Use it as the screening filter — an offer whose disclosed rate is far above its nominal rate has costs in it you should ask about — and then do the holding-period calculation on the two or three offers that survive the screen. What it cannot do is tell you which of two honestly quoted offers suits a borrower who will move in four years, because it was never computing that.

How do I choose a holding period honestly?

Choose it before you see the offers, and choose it from your life rather than from the loan. The relevant questions are whether the property fits the household you expect to have in five years, whether your work is portable, and whether you have moved more or less often than every seven years in the past. Then, because you will be wrong, test the ranking at half and double your estimate. If the winner is the same across that whole range the estimate did not need to be right. If it is not, prefer the offer that wins in the shorter window, because the cost of being wrong is smaller in that direction: an unexpected move with a low-fee loan costs you a little, an unexpected move with a high-fee loan costs you the fees.

Does the possibility of refinancing change the answer?

It shortens the effective holding period, which pushes the comparison toward the low-fee offer. A refinance terminates the loan just as decisively as a sale, and it does so at a moment you do not choose: when rates fall enough. So a borrower who paid heavily up front for a low rate is exposed twice, first to moving and then to the possibility that the rate they bought becomes unremarkable. That is a real argument for treating a high-fee, low-rate offer as the less flexible of the two, and it applies with particular force in markets where early repayment charges are small or absent, because there the refinance is easy and therefore likely.

The two fee lists do not contain the same items. How do I line them up?

Sort every line into one of two buckets and ignore the labels entirely. Bucket one: money you would not pay if you took the other offer. Bucket two: money you would pay either way. Only bucket one belongs in the comparison. A tax on the purchase, a survey the seller requires, the removal van, the cost of the property itself — all of these are identical under both offers and adding them to both totals changes nothing except your confidence in a wrong figure. Where a charge is unavoidable but differently priced, such as a valuation that one lender does in-house and the other outsources, the difference goes in bucket one and the common part in bucket two. When the two lists have been sorted this way they are comparable even if no two line items share a name.