💰 Finance & Access
A $50,000 Ring Costs $750 a Year to Insure. The Insurer Expects to Pay Out $428 of It.
Scheduled jewelry floaters run 1 to 2 percent of appraised value a year. We ran an illustrative, dollar-by-dollar decomposition of the premium, back-solved the implied 0.86% loss rate, and found when self-insuring wins.
A $50,000 engagement ring costs about $750 a year to insure through a specialty carrier. Over twenty years, that is $15,000 in premiums on a policy whose maximum payout is $50,000. Nobody believes they will lose the ring every 67 years, which is what a fair bet at that rate would require. And yet the price sticks, year after year, across an entire industry, which means either millions of buyers are bad at arithmetic or the arithmetic is subtler than it looks, and the answer is a little of both: the buyers are insuring something real, and the price is layered three deep.
We ran the numbers. Here is the receipt.
Decomposing the Dollar
Start with the rate itself. Industry coverage guidance puts annual specialty coverage at 1 to 2 percent of appraised value: Jewelers Mutual's own example is a $5,000 ring for $50 a year, or 1 percent flat. Bogleheads policyholders report about 1.2 percent a year for the company's Perfect Circle policy; the worked examples below use the 1.5 percent midpoint. Deductibles run from zero to $500, varying by zip code, safe, and carrier.
To see where the money goes, decompose the dollar the way state filings do: regulators sort scheduled jewelry floaters into the inland marine line, and the National Association of Insurance Commissioners' industry snapshot puts that line's combined ratio at 83.9 percent, the share of each premium dollar consumed by claims, handling, and expenses. Apply that to a 1.5 percent rate on our $50,000 ring, and the $750 premium splits, illustratively, into about $428 of expected claim payments, about $203 of commissions and administration, and about $120 of underwriting profit.
| Component | Share of premium | Dollars | Basis |
|---|---|---|---|
| Expected claim payments | ~57% | ~$428 | Illustrative loss ratio; NAIC inland marine anchor |
| Commissions and administration | ~27% | ~$203 | Illustrative expense ratio; NAIC inland marine anchor |
| Underwriting profit | ~16% | ~$120 | 100% minus NAIC combined ratio of 83.9% |
At a more typical $10,000 appraisal, the same split gives $150 a year: about $86 of expected claims, $41 of expenses, and $24 of profit. Scales linearly.
That $428 is the number that matters: an expected loss cost of roughly 0.86 percent of insured value per year (1.5 percent x 57 percent), and that single multiplication is the hinge of this article, because every argument about whether the price is fair reduces to whether 0.86 percent resembles the real world. The math is not hiding. If you imagine insurance as a bet against total loss, 0.86 percent a year means a total loss every 116 years, which sounds absurd. But total loss is not what jewelers' insurers mostly pay for. In a single year, per the company's figures, damage and loss can account for nearly 90 percent of all claims filed, with theft and burglary the other 10 percent. Accidental loss or mysterious disappearance, the industry's term for "I have no idea where it went," made up nearly 40 percent of the company's 2009 claims; both pages are company-sponsored posts, which is worth knowing. These are shares of claims filed, not dollars paid: a theft claim pays far more than a prong repair. Disappearance coverage is the moral-hazard frontier: once "I lost it" is a covered peril, every claim is the insured's word against nothing, and the price carries that suspicion.
Partial claims change the frequency math completely: if the average claim pays 25 percent of insured value, an expected loss rate of 0.86 percent implies a claim frequency around 3.4 percent a year, and a loss rate that looks impossibly low as a total-loss probability becomes an entirely ordinary claim frequency the moment you stop assuming every claim is a total loss. A BriteCo survey of buyers who purchased in the past ten years found 59 percent of the insured ones had filed a claim, and Jewelers Mutual's own 2026 travel study, a company release, found jewelry losses nearly 2.5 times more likely while pieces were being worn than while packed. That 59 percent sits well above what the 3.4 percent annual frequency implies over a decade, about 29 percent, which is itself evidence of how skewed the insured pool is: the survey reached a jewelry insurer's own audience, not a random sample. Forget the average owner: the insured population is the population that wears the thing.
| Reading of the rate | Implied frequency |
|---|---|
| 1.5% as a fair total-loss probability | One total loss every ~67 years |
| 0.86% expected loss cost as a total-loss probability | One total loss every ~116 years |
| Reality: partial claims | ~3.4% claim frequency a year at 25% average severity |
Three Forces Pushing the Rate Above Naive Probability
First, adverse selection: buyers are disproportionately people who wear expensive pieces daily and travel with them. Second, the insured value is soft: retail appraisals can exceed what you paid, and since the premium is a percentage of the appraisal, you pay 1.5 percent of an inflated number, while the payout is capped at actual replacement cost. Third, the loading is heavy: only about 57 cents of your premium dollar funds claims. That is the markup stack.
Ruin vs. Replacement
Liability insurance sells protection against the right tail: a small chance of a ruinous judgment, plus defense costs that arrive whether you win or lose, which is why the product feels expensive even when the underwriter is losing money on it. That same NAIC snapshot showed private passenger auto liability running a combined ratio of 109.5 percent, meaning underwriters lost money on the line before investment income. Jewelry floaters have no tail; the worst case is the appraised value, printed on page one. Ruin protection barely breaks even on underwriting, while the line selling a capped, predictable risk keeps 16 cents on the dollar. Nobody is cheating; everybody is charging. That asymmetry is the point of this piece: the scarier-sounding product is the worse business, and the boring one is the annuity.
The answer is about the shape of your utility curve rather than the expected value, which is negative for almost every buyer, and keeping those two ideas separate is the whole argument. A loss is worth insuring when replacing the item would force real tradeoffs. You do not need to be buying jewelry on margin for that to bite: a $30,000 ring financed over twelve months that vanishes in month three means paying installments on a ghost, and a $50,000 piece that is a quarter of your liquid net worth, gone a month before the wedding, means a forced replacement at retail prices on a deadline. Homeowners policies typically cap jewelry theft at $1,000 to $5,000 and exclude mysterious disappearance entirely, so the floater also buys a coverage gap, not just a repricing. What ties these cases together is not recklessness; it is concentration, and the same logic covers business inventory, where a jeweler's traveling case or a dealer's show stock puts livelihood and merchandise in the same bag.
The Strongest Case Against This Story
Steel-man the objection at full strength: maybe the price is fair and the buyers are the rational ones, because a market where most buyers eventually file a claim is not obviously a market where the house is winning too much. Engagement rings are worn daily by distracted people in their twenties, taken off at gyms and beaches, and cleaned over open drains. Taken at face value, a 59 percent claim incidence suggests the 1.5 percent rate could be underpriced for the actual risk pool, not overpriced. Insurers know appraisals are inflated and price accordingly; the game is priced in. And "expected value is negative" indicts all insurance, including the liability coverage nobody calls a scam. That rebuttal has teeth because jewelry risk is behavioral, and behavior is exactly what the buyers are insuring.
What This Analysis Did Not Prove
Start with the biggest caveat: the NAIC's 83.9 percent combined ratio covers the entire inland marine line, including commercial cargo, not just personal jewelry, and no public filing breaks out the personal-jewelry book, so the $428 figure is illustrative rather than audited. Those claim-mix statistics come from Jewelers Mutual analyses of 2009 and 2013 claims. BriteCo's 59 percent figure is a vendor survey with an unspecified claim window. And 1.5 percent is a midpoint rate, not a quote. None of this changes the structure, but your exact split will differ.
What You Can Do
If the piece is worth more than about 5 percent of your liquid net worth, price a standalone floater against a homeowners rider and compare the mysterious-disappearance language line by line. Get the appraisal from an independent appraiser who charges by the hour, not a percentage of value. Take the highest deductible you can fund from cash, because the zero-deductible option is where the loading is fattest. Photograph everything, keep records in the cloud, and re-appraise every three to five years. If you travel with jewelry, the hotel safe beats the jewelry roll. And if the number is small relative to your wealth, skip the policy and self-insure. That is not bravado; it is the same math, run in your favor. This is the math, not personal advice; your risk tolerance and state rules may differ. Know someone shopping for a ring? Forward this before they sign.
The Bottom Line
A 2 percent premium on jewelry is not a 2 percent probability of loss, and nearly every confused conversation about this product, including the one that prompted this article, comes from reading the rate as a probability. It is roughly a 0.9 percent expected loss rate, marked up through expenses, profit, inflated appraisals, and a risk pool full of people who actually wear the merchandise, and every one of those markups is individually defensible, which is what makes the total so hard to argue with. Liability insurance is priced against ruin; jewelry insurance is a marked-up replacement plan with excellent marketing. Buy it when the loss would genuinely hurt, when the item is financed, or when you need the mysterious-disappearance coverage your homeowners policy excludes, because those are the three cases where the premium math stops being the whole story. Otherwise, keep the $750 a year. That expected $428 payout is already in the price, and the other $322 is the cost of someone else holding your risk.