You have agreed the price. You are in a small office off the showroom floor with someone you have not met before, and there is a screen turned slightly toward you. This part takes about five minutes, and one of the things offered in it is GAP.
The pitch is simple and it is true. A new car loses value faster than the loan pays down. Write it off in year two and your insurer pays what the car is worth, which can be less than what you still owe. GAP covers the difference.
That is a real risk. Some people genuinely have it. The question this chapter answers is narrower and harder: what does it cost, and what do you get?
The first half has an answer. The second half does not, and the reason it does not is the most useful thing in here.
The Consumer Financial Protection Bureau looked at loan-level data on roughly 34 million auto loan originations. The average cost of a GAP product financed into a US auto loan was $952.1
Hold that against the numbers that circulate. Consumer articles usually quote $400 to $700. The agency's measured average is above the top of that range. If you have been using the popular figures, you have been going easy on the dealership.
Almost nobody writes a check for GAP. It goes into the amount financed, which means you borrow it, and you pay interest on it for the length of the loan.
Federal Reserve data for early 2026 puts the average new car loan at a finance company at $42,504 over 66 months at 6.1%.2 Industry origination data for the same quarter puts new loans at 6.39% over about 69 months, and used loans at 11.43% over about 68.3
Run $952 through each of those:
| The loan | Interest on the GAP alone | What GAP really cost |
|---|---|---|
| 6.1% over 66 months | $171 | $1,123 |
| 6.39% over 69 months | $189 | $1,141 |
| 11.43% over 68 months | $344 | $1,296 |
So the number on the form is not the price. On a typical used car loan the true cost is about 36% higher than the figure you agreed to.4
And notice who pays the most. The used car buyer, at the higher rate, on the older vehicle, with the smaller deposit — the person the product is aimed at hardest — pays the largest financing surcharge on it.
Now the other half. Of the money American drivers pay for GAP, how much comes back to them as claims?
Nobody publishes it. Not the states, not the federal agencies, not the industry.
That is not a gap in this research. It was looked for, hard. Here is the whole of what exists: one underwriter's own filing showing a 20.9% GAP loss ratio for the first half of 2010, against 90.4% the year before.5 One company, one half-year, sixteen years ago, and a swing so wide between adjacent years that it cannot stand for anything at all.
There is a structural reason for the silence. In much of the country GAP is not sold as insurance. It is sold as a debt waiver — an agreement by the lender to forgive the shortfall — and a debt waiver is not an insurance product, so it never enters the data that insurance departments collect and publish. The money is real and the reporting category does not exist.
Two states have at least set a standard. Oregon requires filed GAP insurance rates to anticipate that at least 50% of premium will come back as claims.6 Washington's regulator weighs whether an insurer can reasonably be expected to reach 60%.25 Both are floors on what a regulator will approve. Neither is a measurement of what happens, and no state publishes the second thing.
There is a further reason the number cannot exist, and it is the sharpest part of this. Where GAP is sold as a waiver rather than insurance, the price a buyer pays splits three ways: the reserve set aside to pay claims, an administration fee, and the dealer's markup. Only the reserve reaches an insurer. The other two are never counted as premium at all — not for premium tax, not for capital rules, not for any experience return.26 So even a regulator who collected every figure the insurance system holds would be measuring claims against the smallest slice of what the buyer actually handed over. The markup is not missing from the loss ratio by oversight. It is outside it by construction.
One state shows the whole thing at once, in three provisions of a single code.
Virginia requires insurers to file credit life and credit sickness experience every year, and its Commission compares actual loss ratios against the approved standard every three years, then republishes rates.30 It measures credit insurance carefully, on a statutory schedule.
Virginia also requires a retail seller of a GAP waiver to insure that obligation under a contractual liability policy, to report the sales, to forward the money, and to hold what it collects in a fiduciary capacity.31 So an insurance policy does stand behind a Virginia GAP waiver. It covers the seller's obligation, and it sits behind the reserve — not behind the price the buyer paid.
And Virginia says a GAP waiver is not insurance, and is exempt from the insurance laws of the Commonwealth.32
One code, three provisions, and the payout still cannot be measured. The state measures credit insurance, requires the GAP obligation to be funded and held in trust, and puts the waiver itself outside the regime that does the measuring. The markup sits outside the loss ratio by law, not merely by industry habit.
And the silence is not for want of attention. In 2023 Colorado looked directly at GAP, decided the price needed a ceiling, and set one: no more than 4% of the amount financed or $600, whichever is greater, with sales barred above 150% loan-to-value.33 A legislature examined this product closely enough to cap what it may cost, and there was still no published figure anywhere for what it pays back. The cost got regulated. The return was never measured.
So when someone at the desk tells you GAP is good value, ask how they know. The national credit insurance return that states file with each other covers credit life and credit disability and never mentions GAP.27 Where GAP is written as insurance, at least one state folds it into a combined credit line with no separate figure.29 Nobody at that desk has a payout number, because the category that would hold one does not exist.
There is one thing the complaint record does show, and it is worth sitting with.
Every complaint Americans file about consumer finance goes into a public federal database. Roughly one in five carries a written account, and each account carries a ZIP code. ZIP codes can be matched to income from tax returns. Two facts come out of doing that.
The typical GAP complaint comes from a ZIP code where the median income is about $69,000. For every other add-on product in this book, about $83,000.
And GAP is the only one of them where complaints from the poorest fifth of ZIP codes roughly match complaints from the richest — 25.6% against 24.8%. For mortgage insurance the richest outnumber the poorest four to one. For title insurance, more than three to one.
The second fact matters more than the first, because of which way the bias runs. Across the whole database, richer areas complain more. The richest fifth files 34.5% of all complaints and the poorest fifth 19.8%. Everything in this data leans toward higher earners, for reasons that have nothing to do with any product. GAP is the one that comes out level anyway.
So it appears that GAP is sold hardest to the people with the least room to absorb it — and that the complaint record understates it rather than overstating it. If the poorest fifth complains less across the board and still shows up here in equal numbers, the people buying it are very likely poorer still.23
Here is what the CFPB says GAP does at the end of its life. A product like GAP "will not offer any possible benefits after either early payoff or repossession."7
Read that against how it is sold. One premium, paid at signing, financed across the whole term of the loan.
Put the two together and you get this: you can finish paying for GAP long after it stopped being able to help you. Pay the loan off early, refinance, trade the car in, or have it repossessed, and the cover is finished — while the borrowed premium and its interest are still sitting inside a balance somewhere.
That is not a dealership problem. That is the shape of the product.
There is machinery meant to handle exactly this. When the loan ends early, the unused portion of the premium is supposed to come back to you.
The enforcement record on that machinery is not close to acceptable.
The CFPB found that Wells Fargo "did not ensure that unearned GAP fees were refunded to all borrowers who paid off their loans early" — while the same firm "obtained such GAP fee refunds when it would benefit Respondent."8 Refunds flowed reliably in one direction. The court-filed settlement in the related class action was $45 million.9
At another servicer, at least $4 million was never refunded to an estimated 5,600 customers who paid off early, and more than $1 million was left sitting inside the deficiency balances of 2,870 customers whose cars had been repossessed.10
That last detail is the one to sit with. An unrefunded amount inside a deficiency balance does not just go unpaid. It gets collected — and those balances get sold on to debt buyers, who pursue people for money that should have been returned to them years earlier.11
Examiners found the same failure across supervised servicers generally, not at one firm.12 Colorado's attorney general has recovered more than $23.5 million for nearly 132,000 people on GAP refunds alone.13 One firm's remediation under a single state's rules covered about 90,000 accounts and more than $25 million.14
Servicers also kept taking monthly payments after they already knew a total loss was covered, then got the reimbursement calculation wrong.15
Several states cap what can be charged. Read them against a real loan.
Colorado caps GAP at the greater of 4% of the amount financed or $600.16 California caps it at 4%.17 Texas at 5%.18
On the Federal Reserve's average amount financed of $42,504, those caps permit:
Against a measured average price of $952.
The caps in the largest capped states sit at roughly double what the average buyer already pays. They do not bind an ordinary transaction. And a cap set as a percentage of the loan has a strange shape: the bigger and longer your loan, the more they may charge you — which is the reverse of how the product's cost behaves.
Indiana caps only the non-refundable version at $400, and leaves refundable waivers uncapped.19 South Carolina and Utah set no price cap at all.2021
This is not an argument that nobody should ever buy GAP. The risk it covers is real, and for a specific buyer it is worth having.
You are that buyer if all of these are true. You put little or nothing down. Your loan runs long — sixty months or more. You rolled negative equity from an old loan into this one. And you could not absorb the shortfall out of savings if the car were written off next year.
That combination is not unusual. It is exactly what a long, low-deposit loan on a fast- depreciating car produces, and it is the situation GAP was designed for.
But being the right buyer for the risk does not make the dealership the right place to buy it. Those are two separate questions, and the five minutes in the small office asks you only the first one.
Ask what it costs as a number, not as a monthly figure. Then add the interest — you are borrowing it.
Ask your own auto insurer first. Many carriers will add GAP-style cover to an existing policy. What that costs could not be established from any regulator filing for this chapter, and quoted ranges disagree with each other too much to print. That is precisely why it is worth a phone call before you sign — you will get a real number for your car, in about ten minutes, and it is the only comparison that matters.
Know that you can decline it and still have the loan. GAP is optional. Where sellers have told buyers otherwise, that has appeared in the federal record as a complaint.22
That is not only anecdote. Across the federal complaint database, complaints about GAP mention being told the product was required about half again as often as complaints about ordinary financial products — 1.52 times the rate, comparing accounts of matching length.24 For private mortgage insurance the same language runs higher, at 2.5 times, but there it is mostly people describing a real rule: mortgage insurance genuinely is required above a certain loan size. GAP never is. So it appears that a meaningful share of GAP buyers are being told something that is not true.
Keep the paperwork, and diary the loan. If you pay off early, refinance, trade in, or have the car repossessed, a refund is likely owed. On the record above, it may well not arrive unless you ask.
You will pay around a thousand dollars, and closer to thirteen hundred once it is financed on a used car loan. Nobody in the United States can tell you what share of that comes back to buyers, because no regulator collects it. The cover can end while you are still paying for it. And the refunds meant to handle that have failed at a scale measured in tens of millions of dollars and hundreds of thousands of people.
Whether you need the risk covered is a real question with a real answer for some people. Whether this is the place to buy it is a different question, and the five minutes is not designed to let you ask it.
Plenee Academy provides financial information and education, not personalized financial advice. Plenee Co. is not a registered investment adviser, broker-dealer, or financial planner. Legal Disclosures & Notices →