# How Title Insurers Get Paid: A Nickel in Claims, Seventy Cents to the Agent, and Nobody Shopping on Price Only about a nickel of every title-premium dollar ever pays a claim, and roughly 70 cents goes to the agent. The buyer who pays the premium almost never picks the insurer. Author: J.A. Watte Published: July 20, 2026 Source: https://jwatte.com/blog/how-title-insurers-are-paid/ --- *Thirty-first in a series on jobs whose pay system is stranger than the salary. This batch is about the toll collectors nobody names on the receipt: the [card networks](/blog/how-card-networks-are-paid/) that skim every swipe, the [pharmacy benefit managers](/blog/how-pharmacy-benefit-managers-are-paid/) that sit between your drug and your plan, the [credit rating agencies](/blog/how-credit-rating-agencies-are-paid/) that stamp a bond and get paid by the issuer. Title insurers belong right in that row. They collect a fee on nearly every home sale and refinance in America, the fee is paid by the buyer at closing, and the buyer is almost never the person who chose them. Every figure below is cited to a federal audit, a company filing with the SEC, a consumer-protection agency, or an industry statistics body, and where a number could not be verified at a primary source I say so rather than estimating.* Here is the strangest fact about title insurance, and it is the whole article in one sentence. It is the rare insurance line that almost never pays a claim. On property and casualty insurance overall, roughly 73 cents of every premium dollar went back out as losses in 2005. On title insurance that same year, the figure was about 5 cents. A nickel. Ninety-five cents of every premium dollar did something other than cover a loss, and the largest single thing it did was pay the agent who sold it. That is not a scandal by itself. It is a design. But it is a design that produces a toll booth, and like every entry in this series, the person who pays the toll is not the person who chose it. ## A nickel of every dollar Start with the loss ratio, because it is the number that makes this line unlike any other insurance you buy. When the Government Accountability Office studied the industry in 2007, it found that losses and loss-adjustment expenses for title insurers as a whole came to only about 5 percent of total premiums written in 2005. For property and casualty insurers the same year, the comparable figure was about 73 percent. The GAO treated that gap as the defining feature of the line, and it is easy to see why. On most insurance, the premium is a pool. You and thousands of other drivers pay in, some of you crash, and the money flows back out as claims. The insurer keeps the spread. Title insurance does not work that way, and the reason is structural. The National Association of Insurance Commissioners describes title as an indemnity policy that protects against defects which already exist at the moment the policy is issued, a forgotten lien, a botched deed, a missing heir. That is the opposite of auto or life insurance, which cover events that happen after the policy is written. Because the risk already exists, it can be hunted down and eliminated up front by the title search, rather than pooled and paid out over time. The search is the product. The insurance is the backstop on the search. So the loss ratio is structurally tiny, not because insurers are stingy, but because the whole business is built to make sure almost nothing is ever paid out. Which raises the obvious question. If only a nickel goes to claims, where do the other 95 cents go? ## Seventy cents to the agent The GAO answered that with a pie chart, and the pie chart is the pay system. Of every premium dollar in 2005, about 5 cents went to losses. About 25 cents went to what the GAO called other expenses, meaning salaries, rent, and equipment. And about 70 cents went to title agents, as payment for the title search and examination and as commission. Seventy cents of the premium the consumer pays goes to the party that produced the referral and did the search, not to a pool that protects the consumer against loss. At the level of the individual agent, the share is even larger. The GAO found that title agents typically retain 80 to 90 percent of the premium the consumer pays. The industry-wide figure comes out lower, around 70 percent, only because the insurers also run direct operations where there is no agent to split with. Look at the GAO's state-by-state sample and the agent's cut is enormous almost everywhere: about 90 percent to the agent in California, about 86 percent in Colorado, about 85 percent in New York and in Texas, about 80 percent in Illinois. Iowa is the lone exception, because it has no split at all, running title through a state-owned entity instead. The GAO noted that it did not independently verify these split figures, which came from interviews with insurers, agents, and regulators, so treat them as the industry's own description of itself. Now hold two facts next to each other. The agent keeps 80 to 90 cents of the premium dollar. The claims pay out a nickel. The premium, in other words, is priced mostly to fund the distribution of the premium. That is a toll, and the next question is the one that always matters in this series: who chose the toll collector, and who pays? ## Reverse competition, or why the shopper never shops In a normal market, the person paying picks the seller, and the sellers fight over that person by cutting price. Title insurance breaks that loop on purpose, and the industry has a name for the break. It is called reverse competition. The GAO described the mechanism precisely. Consumers, it wrote, generally do not select their title agent or insurer, and title agents do not market to consumers but rather compete among themselves for referrals from those who do, that is, real estate and mortgage professionals. Read that again. The company you pay is chosen for you by your realtor, your lender, or your closing attorney. So the title agent has no reason to advertise a lower price to you, the buyer. The agent's customer is not you. The agent's customer is the professional who steers you. The competition runs backward, away from the person holding the checkbook and toward the person holding the referral. I want to be careful with the label. The GAO fully documents the dynamic, that the referrers choose and the buyer pays. The two-word term reverse competition is standard terminology in the NAIC and academic literature for exactly this pattern, and I am attributing the concept to the GAO and the label to the broader field, rather than claiming the GAO prints that specific phrase on the page. The dynamic has a predictable consequence, and the GAO named it too. Because the referrer is choosing, and because the referrer sometimes has a financial stake in the agent being chosen, the arrangement creates conflicts of interest. Whenever the party who picks the vendor also profits from the pick, the buyer's price is nobody's priority. ## The captive at the closing table The buyer's position in all this is the weakest in the series, because the buyer cannot walk away either. The GAO put it plainly. The consumer, it concluded, is in a potentially vulnerable situation where, to a great extent, they have little or no influence over the price of title insurance but, at the same time, they have little choice but to purchase that insurance. That is the captive payer in two clauses. You cannot really move the price, because you did not choose the vendor and the vendor is not competing for you. And you cannot really decline, because the lender requires the policy before it will fund the loan. The premium is paid once, at closing, to the title agent, at a moment when the deal is done and the last thing any buyer wants is to reopen it. The consumer agency confirms the shape of the trap while pointing at the one exit. The Consumer Financial Protection Bureau tells buyers that most lenders require you to purchase a lender's title insurance policy, which protects the amount they lend. Note whose loss that policy covers. It protects the lender, not you. The separate owner's policy, the one that protects your equity, is something the CFPB says you may want to buy, an option, not a requirement. And the exit the CFPB flags is the one reverse competition hides: you can usually shop for your title insurance provider separately from your mortgage, and if you shop, you could save money. The saving is real. It is just structurally invisible, because the buyer has been trained to accept whoever the closing table hands them. ## Five companies, ninety-two percent of the country A toll booth pays best when there are few operators, and this market is concentrated at the top. In 2005 the GAO found that five insurers accounted for 92 percent of the national title market. The two largest, Fidelity and First American, split roughly 29 percent and 27 percent between them, almost evenly. LandAmerica held about 18 percent, Stewart about 11 percent, Old Republic about 8 percent, and everyone else combined took about 6 percent. Five underwriters, nearly the entire country. Those exact shares are a 2005 snapshot, and the roster has since changed. LandAmerica was absorbed by Fidelity during the financial crisis, leaving the surviving underwriting families widely described today as Fidelity National Financial, First American, Old Republic, and Stewart. I could not pull precise current market-share percentages from a primary source, because the detailed tables from the American Land Title Association are paywalled, so I am not asserting today's exact split. The industry body itself notes that a small number of underwriters, on the order of 27, represent about 98 percent of the U.S. market, which tells you the concentration at the underwriting tier remains extreme even if the precise 2005 figures are stale. ## The kickback that puts a number on the toll The clearest proof that the premium is mostly a distribution fee, not a risk pool, is what happened when regulators pulled apart one common scheme and did the arithmetic. The scheme was captive reinsurance. A builder, lender, or broker would set up its own reinsurance company, and a slice of every title premium on its referred deals would be routed to that captive reinsurer, ostensibly to share the risk. The GAO walked through a specific example from a multistate settlement administered by the Colorado Division of Insurance in 2005. Take a $250,000 transaction in Colorado carrying a $1,614 premium. Of that, about $632 flowed to the captive reinsurer. And what did the reinsurer take on in exchange? About $36 of expected losses, using Colorado's roughly 4.5 percent combined loss ratio. Collect $632, expect to pay out $36, net profit about $596. That $596 is roughly 37 percent of the entire premium, handed to a reinsurer that bore almost no real risk. As the settlement noted, virtually no claims were filed. Sit with that. More than a third of the premium, on that example, was not insurance at all. It was disguised referral compensation, a way to pay the party that steered the customer while calling it risk-sharing. The reinsurance was a costume. Underneath it was a kickback, and the kickback was quantified at about 37 cents on the premium dollar. ## RESPA, and why the fence is low There is a federal law that is supposed to stop exactly this, and it is worth understanding why it does not stop much. Section 8 of the Real Estate Settlement Procedures Act generally prohibits the giving or accepting of kickbacks and referral fees in a real-estate settlement. That is the fence. But look at the penalty behind it. The criminal sanction is a fine of up to $10,000 or up to one year in prison, and the GAO noted that it is rarely used, in part because it requires a Department of Justice prosecution. A $10,000 maximum fine is not a deterrent to an industry collecting billions in premiums. It is a cost of doing business, if it is ever imposed at all. And RESPA leaves a wide, legal door open beside the fence. Affiliated business arrangements, where a realtor or lender owns a piece of the title agency it refers you to, are permitted, provided the relationship is disclosed, the customer is not required to use the affiliate, and the return to the owner is limited to their ownership interest. Disclosed, in practice, means a form in a stack of forms at closing. So the law bans the crude kickback, permits the structured version with a disclosure, and backs the ban with a penalty most operators will never face. The wall against reverse competition is real, but it is low, and everyone in the business knows exactly where the gate is. ## What the buyer actually pays, and how the price is set For a sense of scale on the individual bill, the GAO cited a 2006 survey figure it did not itself validate: the average all-in cost of simultaneously issued lender's and owner's policies on a $200,000 loan was about $859, roughly 28 percent of total loan origination and closing fees. That figure comes from the Bankrate 2006 Closing Costs Survey, so treat it as a dated, secondary snapshot rather than a current price. What matters more than the exact dollar is how the number is produced. The price is not a quote you negotiate. It is the underwriter's or the state's published rate applied mechanically to the size of the transaction, to the loan value for the lender's policy and to the home price for the owner's policy. And the GAO found that in most states, regulators do not examine whether that rate bears any relationship to the agent's actual costs. Rate regulation is a state function, and it is aimed at loss projections, even though losses are only about 5 percent of title premiums. Among the GAO's six sample states, one did not regulate title rates at all, and only one of the four all-inclusive-rate states regularly reviewed agents' costs. Some states, including New York and Texas, simply promulgate the rate outright. So the premium is set by applying a formula to a number, and the formula is rarely tested against what the work costs. The rate is the toll schedule, and the toll schedule is mostly unexamined. The insurers do not pretend otherwise about the split. The GAO reported that the insurer-agent division was negotiated on the basis of the agent's volume of business, the quality of the agent's past work, and the insurer's desire to increase its share of business in a certain geographic area. In other words, the split is negotiated on referral leverage and volume, not on the agent's costs. The insurers told the GAO they did not analyze whether the amount the agent retained actually reflected the agent's expenses. They set the rate knowing the split, and they do not check whether the split is earned. ## The pattern is not history Someone reading the 2005 numbers might reasonably ask whether any of this still holds twenty years later. One company's own current filing says it does. First American Financial is a title pure-play, one of the surviving big underwriters, and its 10-K for fiscal 2025 discloses its benefits, claims, and losses as a share of premium revenue in the supplementary insurance schedule. Computed from those figures, the ratio runs about 8.9 percent in 2019, about 10.1 percent in 2020, about 8.0 percent in 2021, about 7.3 percent in 2022, about 7.2 percent in 2023, about 6.4 percent in 2024, and about 5.7 percent in 2025, against title premium revenue of $5,722.1 million that last year. This is a single company's figure, not an industry universal, and I am presenting it as First American's own. But it corroborates the shape the GAO found two decades earlier. Even at the low single digits, the nickel-on-the-dollar loss pattern is not a dusty 2005 artifact. It is visibly still in the filings. I want to be honest about the one thing I could not pin. A current, industry-wide NAIC composite loss ratio for 2025 was not fetched from a primary source, and the current exact 70 percent agent split is likewise a 2005 GAO figure that I could not re-verify against a fresh primary document, because the SEC archive host blocked automated fetches and the newest filings are not yet archived elsewhere. The structure is widely described as unchanged, and First American's own numbers point the same way, but I am flagging that the precise current split is inferred from the older primary figure rather than freshly sourced. ## A toll that scales with the housing market The last thing to grasp is how large this flow is, because a small percentage of an enormous river is still an enormous amount of money. According to the American Land Title Association, the U.S. title industry booked about $18.5 billion in premiums in 2025, up 13.8 percent from the year before, which implies roughly $16.3 billion in 2024. The first quarter of 2026 alone ran about $4.5 billion, up from about $3.9 billion in the same quarter a year earlier. These ALTA figures came through the association's newsroom, and I could not pin the specific dated press-release URL or the exact base-year dollar, so I am marking them secondary even though ALTA is the authoritative body for the number. The scale is the point. This is a toll that grows with the housing market, collected on nearly every home sale and refinance, funding a distribution channel through a premium that pays a nickel in claims. When rates fall and refinancing surges, the river swells and the toll swells with it. Nobody at the closing table is shopping. The buyer signs, pays once, and moves on, and the seventy cents goes where the referral pointed. ## What a salaried reader should take from this **When the person who chooses the vendor is not the person who pays, price stops working.** Reverse competition is the cleanest example in this whole series of a market where the ordinary discipline of shopping is simply absent by design. Your realtor or lender picks the title company, the title company competes for their referral rather than your dollar, and you pay the bill. Any time you find yourself paying for something a professional selected on your behalf, ask who that professional's real customer is. If it is not you, the price is not being held down for you, and the CFPB is right that you can often shop it yourself and save. **A low loss ratio is not automatically a rip-off, but it tells you what you are really buying.** Title insurance pays a nickel on the dollar because the risk is eliminated up front by the search, not pooled over time. That is legitimate. But it also means the premium is overwhelmingly a payment for distribution and search, not for a claims pool, which is exactly why 80 to 90 cents can go to the agent. When an insurance product almost never pays out, you are not buying protection against a common event. You are buying a certificate, and most of the price is the cost of the certificate reaching you through a chain of referrals. **The size of the penalty tells you how serious the rule is.** RESPA bans kickbacks and caps the criminal fine at $10,000 against an industry that collects billions, and the GAO found the sanction is rarely even used. A rule with a trivial penalty and a wide, disclosure-only exception is a rule the industry has already priced in. The captive-reinsurance example, where about 37 percent of a premium flowed to a reinsurer bearing almost no risk, shows what happens next: the banned behavior reappears in a costume the fence was too low to catch. When you evaluate whether a protection actually protects you, look past whether the conduct is prohibited and look at what the prohibition costs the party doing it. **Concentration plus captive demand equals a durable toll.** Five underwriters covered 92 percent of the country, buyers cannot decline the lender's policy, and buyers do not choose the vendor. Each of those alone would be tolerable. Stacked together they produce a fee that behaves like a toll booth on the housing market, the same engine that runs under the [card networks](/blog/how-card-networks-are-paid/) skimming every swipe and the [pharmacy benefit managers](/blog/how-pharmacy-benefit-managers-are-paid/) standing between your prescription and your plan. The tell is always the same. A margin survives not because the service is uniquely good, but because the customer has been structurally removed from the choice. ## Related reading - [How card networks are paid](/blog/how-card-networks-are-paid/): a fraction of a cent on every swipe, collected from a merchant who cannot route around the network. - [How pharmacy benefit managers are paid](/blog/how-pharmacy-benefit-managers-are-paid/): the middleman between your drug and your plan, paid on a flow the patient never sees. - [How credit rating agencies are paid](/blog/how-credit-rating-agencies-are-paid/): paid by the issuer to stamp the bond, another toll where the payer is not the party the rating is supposed to protect. - [How mortgage bond sellers are paid](/blog/how-mortgage-bond-sellers-are-paid/): the same mortgage tolled again on its way up to the banks, one layer above the closing table. - [How insurance agents are paid](/blog/how-insurance-agents-are-paid/): a cut of the premium, front-loaded and baked invisibly into the price. ## Fact-check notes and sources The structural facts, the loss ratio, the agent split, reverse competition, the captive-reinsurance arithmetic, RESPA, and the concentration figures all come from a single federal audit, the GAO's 2007 title-insurance study, which analyzed NAIC data. The consumer guidance comes from the CFPB. The indemnity structure comes from the NAIC. One company's current loss figures come from its SEC filing. The industry premium total comes from ALTA. Where a figure is secondary, dated, or could not be verified at a primary source, it is flagged in the text and here. - **The loss ratio of about 5 percent of title premiums versus about 73 percent for property-casualty insurers in 2005**, **the pie-chart breakdown of the premium dollar** (about 5 percent to losses, about 25 percent to other expenses, about 70 percent to agents), **the agent-level retention of 80 to 90 percent and why the industry figure is lower**, **the state-by-state agent splits** (California about 90 percent, Colorado about 86 percent, New York and Texas about 85 percent, Illinois about 80 percent, Iowa with no split via a state-owned entity), **the reverse-competition description** (consumers generally do not select their agent or insurer; agents compete for referrals from real estate and mortgage professionals; the resulting conflicts of interest), **the captive-payer conclusion** (little influence over price, little choice but to buy; premium paid once at closing), **the market concentration** (five insurers at 92 percent of the national market in 2005; Fidelity about 29 percent and First American about 27 percent, LandAmerica about 18 percent, Stewart about 11 percent, Old Republic about 8 percent, all others about 6 percent), **the state-based rate regulation that rarely examines agents' costs**, **the captive-reinsurance example** (a $250,000 Colorado transaction, $1,614 premium, about $632 to the reinsurer for about $36 of expected losses at a 4.5 percent combined loss ratio, net profit about $596, roughly 37 percent of the premium, with virtually no claims filed, from a 2005 multistate settlement administered by the Colorado Division of Insurance), **the RESPA Section 8 prohibition and its penalties** (up to a $10,000 fine or up to one year in prison, rarely used, and the disclosed affiliated-business-arrangement exception), **the average $859 all-in cost on a $200,000 loan at about 28 percent of closing costs**, and **the statement that insurers negotiate the split on volume and referral leverage rather than agents' costs** are all from the [U.S. Government Accountability Office, GAO-07-401, Title Insurance: Actions Needed to Improve Oversight of the Title Industry and Better Protect Consumers (April 2007)](https://www.gao.gov/products/gao-07-401), specifically pp. 3, 6, 7, 8, 9, 29 to 31, 37 to 42, and 53 and Figures 1, 10, and 11 (GAO analysis of NAIC and industry data). **The state-split percentages are from GAO interviews and were not independently verified by the GAO. The $859 average cost is from the Bankrate.com 2006 Closing Costs Survey as cited by the GAO, which did not validate the survey, and is treated as dated and secondary. The two-word term "reverse competition" is standard NAIC and academic terminology for the dynamic the GAO fully describes; the concept is attributed to the GAO and the label to the broader literature.** - **The consumer guidance** that the lender's policy protects the amount the lender lends, that the owner's policy is optional, and that buyers can usually shop for title insurance separately from the mortgage and could save money, is from the [CFPB, Ask CFPB: What is the difference between a loan policy and an owner's policy of title insurance?](https://www.consumerfinance.gov/ask-cfpb/whats-the-difference-between-a-loan-policy-and-an-owners-policy-of-title-insurance-en-164/), a primary source. - **The indemnity structure** (title insurance protects against defects that already exist at the time the policy is issued, unlike auto or life insurance which cover later events, which is why the loss ratio is structurally low) is from the [NAIC Center for Insurance Policy and Research, Title Insurance topic page](https://content.naic.org/cipr-topics/title-insurance), a primary source. - **First American Financial's benefits, claims, and losses as a share of premium revenue** (about 8.9 percent in 2019, about 10.1 percent in 2020, about 8.0 percent in 2021, about 7.3 percent in 2022, about 7.2 percent in 2023, about 6.4 percent in 2024, and about 5.7 percent in 2025, against 2025 title premium revenue of $5,722.1 million) are computed from XBRL Supplementary Insurance Information (Schedule VI) tags in the [First American Financial Corporation FY2025 Form 10-K (CIK 1472787, filed 2026-02-18), via data.sec.gov companyfacts](https://data.sec.gov/api/xbrl/companyfacts/CIK0001472787.json), a primary source. **These are one company's figures, not an industry universal, and are presented as First American's own.** - **The U.S. title industry premium total** (about $18.5 billion in 2025, up 13.8 percent year over year, implying about $16.3 billion in 2024; about $4.5 billion in the first quarter of 2026, up from about $3.9 billion a year earlier; and that roughly 27 underwriters represent about 98 percent of the U.S. market) is from the [American Land Title Association Market Share Analysis / industry research](https://www.alta.org/industry-research/). **ALTA is the authoritative industry body, but this figure came through its newsroom via web fetch; the specific dated press-release URL and the precise 2024 base-year dollar were not pinned, and ALTA's detailed market-share tables are paywalled, so these figures are treated as secondary.** - **Not asserted, and why:** a current 2025 industry-wide NAIC composite loss ratio (the verified about 5 percent figure is the GAO's 2005 NAIC-sourced number; First American's own about 5.7 percent for 2025 corroborates the pattern, but no 2025 primary composite was fetched); the current exact agent-retention split (the 70 percent industry and 80 to 90 percent agent-level figures are the GAO's 2005 numbers, and the SEC archive host and the corvus proxy both returned 403 to automated fetches of the 2026 title-underwriter 10-Ks, with no Wayback snapshot yet available, so a fresh exact split is inferred from the older figure rather than freshly sourced); and precise current underwriter market shares (the 92 percent and five-insurer concentration is 2005; the surviving families today are Fidelity National Financial, First American, Old Republic, and Stewart after Fidelity absorbed LandAmerica in 2008, but exact 2025 share percentages are behind ALTA's paywall and are not asserted). *This post is informational and journalistic, not legal, financial, or real-estate advice, and nothing here is a recommendation about any title insurer or policy. It describes a federal audit, a consumer-protection agency's guidance, an insurance-regulator topic page, one company's SEC filing, and an industry statistics body. Rates, market shares, filings, and rules change year to year, and several figures here are dated to 2005 through 2006 or flagged as secondary as noted, so verify current data before relying on any of them. Mentions of specific companies, agencies, and trade bodies are nominative fair use, and no affiliation is implied.* --- Canonical HTML: https://jwatte.com/blog/how-title-insurers-are-paid/ RSS: https://jwatte.com/feed.xml JSON Feed: https://jwatte.com/feed.json Hero image: https://jwatte.com/images/how-title-insurers-are-paid.webp