Palomar Holdings
An insurer best known for paying to rebuild Californian houses after an earthquake, now growing mostly through crop, casualty and surety cover.
The scores
Growth3/4 →
Sales have compounded for a decade, but surety and much of crop were bought in the last eighteen months, and crop is federally subsidised.
Business quality3/4 ↓
The regulator's data shows private carriers taking earthquake share from the state pool, but that franchise is under a third of sales and losing price.
Financial position3/4 →
Debt is real but small and covered many times by underwriting profit, and it matures in 2031, outside the horizon. Not net cash.
Execution record4/4 →
Fifteen consecutive quarters of beating guidance, three raises this year, five product lines built without a stumble, and capital returned in cash.
Per share capture3/4 →
Deals are cash and buybacks have outrun stock pay since mid-2025, but the 2024 share sale left the count about 6% above its 2023 level.
Potential2/4 ↓
The founding market is shrinking in policies, and the crop ambition needs share taken from eleven rivals in a programme already 89% enrolled.
Risk profile3/4 ↑
California is down to 30.9% of sales and the earthquake tail is reinsured past the 1-in-250-year loss, but two agencies control a third of it.
Valuation2/4 →
The price needs about 27% on net worth after goodwill; reported profit delivers 23.5%, and 18.8% without reserve releases. The plan reaching it is credible.
Evidence quality3/4 →
Central figures come from filings, a state regulator and a federal agency, but the two sharpest reserve facts are management statements no filing carries.
If it works 1.2× · if it fails you keep about 60%.
The Verdict
We rate Palomar unattractive. It made its name selling earthquake cover to Californian homeowners, and it is a solid insurer. But earthquake has stopped growing, and growth has moved to crop, casualty and surety cover, with thinner margins and less-tested reserves. The price is the highest against net worth of the insurers it names and needs returns above today's for good, so it is full. We think today's returns lean on money released from earlier years' reserves. As a bet it is priced for success: if crop insurance earns the federal programme's average the shares are worth little more than today, and if it never earns shareholders keep about three-fifths.
The long casePalomar's earthquake market is shrinking, so growth must come from crop and casualty; federal crop insurance pays approved insurers a steady gain over time, which Palomar must earn.
The evidence under this - 7 claims
These growth rates are stated on the call and not tabulated in the 10-Q, which reports gross written premium only in total. The surety and crop figures are also not like-for-like organic growth: Gray Surety was acquired on 31 January 2026 and Advanced AgProtection in April 2025, so a large part of both is bought rather than grown. Total gross written premium growth of 27.0% for the quarter is the one figure in this set that is filed.
Three things run the other way. On adjusted profit - the basis management steers on and the basis consensus is struck on - the same calculation gives 31.2%, comfortably above the implied 27%. The denominator is depressed by a January acquisition, so tangible equity understates the capital actually earning: measured on total stockholders' equity the return is 19.9% against a 3.7 times multiple, a much less demanding pairing. And the implied figure is only as good as assumption A01: at a 9.0% cost of equity the requirement falls to about 24.4%, which the reported return already meets.
The figure appears in no filing - the Q2 2026 10-Q does not use the term IBNR at all - so it rests on the call. It is corroborated in direction and roughly in level by the 10-K's own filed split at 31 December 2025, which puts net IBNR at 82.2% of net reserves. Management presents the high share as conservatism, and the six consecutive quarters of favourable development are evidence for that reading rather than against it: reserves that were too high are exactly what releases come from.
Measured against adjusted underwriting income, which is the figure management steers on, the share is lower - 19.0% in the first half of 2026 against 10.8% - because adjusted underwriting income adds back stock compensation, intangible amortisation and transaction costs. The direction and roughly the doubling hold on either denominator. Management also argues the releases are the designed output of reserving conservatively rather than a prop: the same pattern has run for six consecutive quarters and began in the property book before reaching casualty.
Beating guidance 15 times running is as much a statement about how guidance is set as about delivery, and the chief financial officer described the reserving philosophy in the same terms - take the bad news early, let good news arrive later. The independently checkable half is the analyst estimate, and the pack records beats there too, on the adjusted basis consensus is struck on. The one place the record is thinner is that no large catastrophe has tested it: the earthquake retention has not been breached in this period.
The company
A Californian homeowner's ordinary policy does not cover earthquakes. Palomar sells the separate policy that does, pricing each address by its soil, the risk of the ground turning liquid and the distance to a fault, where it says rivals use broad zones. It reaches buyers through agents, brokers and more than 35 insurers that offer its cover beside their own.
It also writes four more lines across all 50 states: casualty, other property, crop and, by lending its licence, cover for other insurers.
The evidence under this - 3 claims
How it makes money
Customers pay a premium, the price of the policy, mostly for a year, and Palomar earns it as the months pass, keeping what is left after claims and costs. It passes 48.4% of premium to reinsurers, who in return pay for the largest losses. Profit also depends on old claims: when claims it set money aside for in earlier years cost less than expected, the difference is released as profit.
The evidence under this - 3 claims
Where the money comes from, stream by stream
| Stream | Share of sales (fiscal year 2025) | How it earns it |
|---|---|---|
| Earthquake | 28.2% | Residential and commercial earthquake cover, mostly in California, priced address by address and sold through agents and partner insurers. |
| Casualty | 26.8% | Liability, professional and surety lines written mostly excess-and-surplus; premiums paid now, claims settled years later. |
| Inland Marine and Other Property | 22.0% | Builders' risk, specialty homeowners, commercial property and Hawaii hurricane cover, much of it through program administrators. |
| Crop | 12.2% | Federally reinsured multi-peril crop insurance; the government subsidises premium and shares losses and gains under its standard agreement. |
| Fronting | 10.8% | Rents Palomar's licensed paper to other carriers for a fee, ceding most of the premium and risk back to them. |
Who runs it
Mac Armstrong founded Palomar in 2014 and still runs it; Chris Uchida has run its finances since 2017. Under them it grew from one earthquake product to five lines and has beaten its own forecasts for years. Insiders own under 4%, most of it his. Bonuses ride on adjusted profit, long-term shares mostly on return on equity against a goal set at half what the company earns, and nothing rewards careful reserving.
The evidence under this - 5 claims
The premium goal was raised 28% for 2026, the relative-return modifier can cut payouts by a fifth, half the long-term award is plain restricted stock that tracks the share price, and the 2026 board added enterprise risk management to the individual goals, so not every lever points at volume and adjusted profit.
Beating guidance 15 times running is as much a statement about how guidance is set as about delivery, and the chief financial officer described the reserving philosophy in the same terms - take the bad news early, let good news arrive later. The independently checkable half is the analyst estimate, and the pack records beats there too, on the adjusted basis consensus is struck on. The one place the record is thinner is that no large catastrophe has tested it: the earthquake retention has not been breached in this period.
Why it wins, and who could beat it
Why customers come back
Palomar sits inside somebody else's sale. More than 35 insurers offer its earthquake cover with their homeowners policy, and one agency, Arrowhead, brings in a fifth of all its sales. A homeowner who bought the cover with the house insurance renews it with the house insurance, and in the June 2026 quarter more than 96% of earthquake premium renewed.
The state regulator's data shows it working: private insurers lifted their share of earthquake policies from 32.63% in 2020 to 39.15% in 2025.
The evidence under this - 4 claims
The figure appears nowhere in the 10-Q or the results release, so it cannot be checked against a filing; retention is stated for residential earthquake alone, which is about 64% of an earthquake book that is 28.2% of gross written premium, so it describes roughly a fifth of the company. Commercial earthquake, the other 36% of the same franchise, saw renewal rates fall more than 20% in the same quarter.
The threat
In earthquake the rival is a state-run pool, the California Earthquake Authority, which holds 60.85% of policies. In casualty it is larger specialists such as Kinsale, which Palomar's own report names first. In crop, every approved provider sells a policy set by federal rule, so only the agent relationship is competed for.
The evidence under this - 5 claims
The rivals: KNSL, CB, AFG, BOW, SKWD, WRB, RLI, and 1 more that is private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| Not listed | The California Earthquake AuthorityPriced against it: Not listed | Palomar's specialty lines: earthquake, inland marine and property, casualty, surety and fronting | On sale | Not listed |
| KNSL | Kinsale Capital Group, Inc.Priced against it: Yes | Palomar's specialty lines: earthquake, inland marine and property, casualty, surety and fronting | On sale | Yes |
| CB · AFG | The eleven other Approved Insurance Providers in the federal crop programme, including Chubb and American Financial GroupPriced against it: One segment only | Palomar Crop: federal crop insurance | On sale | One segment only |
| BOW | Bowhead Specialty HoldingsPriced against it: Yes | - | On sale | Yes |
| SKWD | Skyward Specialty Insurance GroupPriced against it: Yes | - | On sale | Yes |
| WRB | W. R. BerkleyPriced against it: Yes | - | On sale | Yes |
| RLI | RLIPriced against it: Yes | - | On sale | Yes |
How big it could get
How big this could get
California has 12.75 million home insurance policies, and only 12.48% carry earthquake cover, at $956.43 a year on average: $1.52B of cover sold in 2025, against Palomar's $571.4M. The market is not converting. Take-up has fallen from 13.77% in 2020 and so has the number of policies, so the market grew 16.2% on price alone.
The evidence under this - 3 claims
The results
The numbers
Sales grew 60% over the twelve months to Jun 2026, about the same pace as the 58% in the year to Dec 2025.
It keeps 19% of sales as profit, down from 21% in the year to Dec 2023.
Yearly sales growth has ranged from +15% to +58% over the last 5 years.
Shareholders' money earned 22% a year, up from 12% in the year to Dec 2021.
It turned 38% of sales into free cash, $415M over the last twelve months.
It owes $233M more than the cash it holds, about 0.6 years of the cash its operations bring in.
The share count rose 3.6% in 4.5 years, and pay in shares costs 2.5% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-10-04. Fiscal years end in December.
| Dec 2021 | Dec 2022 | Dec 2023 | Dec 2024 | Dec 2025 | 12 mo to Jun 2026 | |
|---|---|---|---|---|---|---|
| Size and growth | ||||||
| Sales | $248M | $327M | $376M | $554M | $876M | $1.09B |
| Sales growth | +47% | +32% | +15% | +47% | +58% | +60% |
| Profit | ||||||
| Net margin | 19% | 16% | 21% | 21% | 22% | 19% |
| What growth costs, per sales dollar | ||||||
| Pay in shares | 2.3% | 3.6% | 4.0% | 3.0% | 2.4% | 2.5% |
| Capital spending | 0.0% | 0.1% | 0.0% | 0.0% | 0.0% | 0.0% |
| Cash | ||||||
| Cash from operations | $87.8M | $170M | $116M | $261M | $409M | $415M |
| Free cash flow | $87.8M | $169M | $116M | $261M | $409M | $415M |
| Balance sheet | ||||||
| Cash and short-term investments | $50.3M | $68.1M | $51.5M | $80.4M | $107M | $62.7M |
| Debt | - | - | - | - | $0 | $296M |
| Net cash | - | - | - | - | $107M | −$233M |
| Returns and shares | ||||||
| Shares (diluted) | 26.1M | 25.8M | 25.3M | 26.2M | 27.5M | 27.1M |
What growth costs
For an insurer the cost of growing is the commissions and overheads paid on each dollar of premium. Those have fallen as Palomar grew, but the newer crop and liability business pays out more in claims, so the saving went to claims: the profit kept from each premium dollar has stayed about level.
The evidence under this - 2 claims
The 2023 ratios are computed here from the filed lines (losses $72.6M, expenses net of commission income $192.6M, over $345.9M) rather than read from a reported table, and 2025 carried almost no catastrophe losses against 5.5% in 2024, so the like-for-like expense gain is flattered less than the loss ratio is. Against the reading, the Q2 2026 adjusted combined ratio widened to 76.7% from 73.1% (C21), so the flat run may already be ending.
The last report (4 Aug)
Quarter ended June 2026, filed figures.
The second quarter of 2026 was, on the company's telling, its 15th straight beat, and it raised its full-year forecast for the third time, to $270M to $280M of adjusted profit.
The detail pointed the other way. Claims took 34.5% of premium earned, up from 25.7%. Strip out releases from earlier years' reserves and this year's claims rose from 29.3% to 39.5%. Those releases supplied 27.8% of first-half underwriting profit, against 13.2% a year before.
What it did to the thesisIt is the core evidence for our negative view. Behind the raised forecast, this year's claims rose and more profit came from releasing old reserves. The case stays open on whether the crop claims are seasonal, as management says, or a lasting change.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $2.36 against $2.19 expected; $1.94 as filed, the gap being what the company adds back.
The evidence under this - 4 claims
Beating guidance 15 times running is as much a statement about how guidance is set as about delivery, and the chief financial officer described the reserving philosophy in the same terms - take the bad news early, let good news arrive later. The independently checkable half is the analyst estimate, and the pack records beats there too, on the adjusted basis consensus is struck on. The one place the record is thinner is that no large catastrophe has tested it: the earthquake retention has not been breached in this period.
Management gave the same figure on the call - 39.6% for the accident-year loss ratio excluding catastrophes - and framed it as expected rather than as deterioration, attributing it to crop growing faster than planned and to crop losses arriving earlier in the year than crop premium is earned. On that account the seasonal shape reverses in the second half, which the 2025 pattern supports. The current-year figure also contains $6.2M of losses on derivative instruments, against an insignificant amount a year earlier; excluding those the accident-year ratio is about 37.3%, still eight points higher.
Measured against adjusted underwriting income, which is the figure management steers on, the share is lower - 19.0% in the first half of 2026 against 10.8% - because adjusted underwriting income adds back stock compensation, intangible amortisation and transaction costs. The direction and roughly the doubling hold on either denominator. Management also argues the releases are the designed output of reserving conservatively rather than a prop: the same pattern has run for six consecutive quarters and began in the property book before reaching casualty.
What management did with the money
It bought The Gray Casualty and Surety Company for about $314.0M in January, borrowing $300M to do it, and started a $0.45 quarterly dividend while buying back $64.1M of stock. Much of the price was goodwill, the amount paid above the assets bought, so net worth with goodwill stripped out fell from $881.6M to $744.2M in six months.
The evidence under this - 4 claims
The price, and where we disagree
In short. Full. The price pays for Palomar earning more on its net worth, for good, than it earns today, and today's return leans on money released from earlier years' claim reserves.
Management expects its claims ratio to keep rising for two more years as the mix shifts to crop and casualty, and is booking claims on this year's crop near 100% of its premium, against about 80% last year.
The next test is the third-quarter results, due 5 November: whether releases from old reserves still supply a large share of underwriting profit, and whether this year's claims keep climbing.
The evidence under this - 4 claims
Management pairs the statement with an expectation that the adjusted combined ratio stays attractive because the expense ratio falls and the share of premium retained rises - and the record supports the offset, since the expense ratio fell from 53.1% to 48.8% year over year in Q2 2026 while the loss ratio rose. So a rising loss ratio is not by itself a falling margin, and the company has so far delivered the offset it describes.
The chief financial officer attributes the figure to the timing of crop accounting rather than to expected results: most crop premium is not recognised until acreage reports arrive in July, so losses run ahead of premium in the first half and the ratio 'trues up' in the third and fourth quarters. The 2025 crop year, booked the same way, ended near 80%, which is close to the federal programme's own long-run average of 0.85 - so the pattern has completed once already at a normal outcome.
Measured against adjusted underwriting income, which is the figure management steers on, the share is lower - 19.0% in the first half of 2026 against 10.8% - because adjusted underwriting income adds back stock compensation, intangible amortisation and transaction costs. The direction and roughly the doubling hold on either denominator. Management also argues the releases are the designed output of reserving conservatively rather than a prop: the same pattern has run for six consecutive quarters and began in the property book before reaching casualty.
Management gave the same figure on the call - 39.6% for the accident-year loss ratio excluding catastrophes - and framed it as expected rather than as deterioration, attributing it to crop growing faster than planned and to crop losses arriving earlier in the year than crop premium is earned. On that account the seasonal shape reverses in the second half, which the 2025 pattern supports. The current-year figure also contains $6.2M of losses on derivative instruments, against an insignificant amount a year earlier; excluding those the accident-year ratio is about 37.3%, still eight points higher.
How we read the multiple
Buyers pay about $4.90 for each $1 of Palomar's net worth: what it owns minus what it owes, leaving out what it paid above value for businesses it bought. Its filed rivals' middle is about $2.90, and none costs more.
That only makes sense if Palomar earns about 27% a year on that net worth, for good. It earned 23.5% in the first half, and 18.8% without money released from reserves set aside for earlier years' claims.
Against its own history. No dearer than its own normal, and cheaper on profit. But that normal was set while earthquake drove the growth, so the comparison flatters today's mix.
The evidence under this - 2 claims
Three things run the other way. On adjusted profit - the basis management steers on and the basis consensus is struck on - the same calculation gives 31.2%, comfortably above the implied 27%. The denominator is depressed by a January acquisition, so tangible equity understates the capital actually earning: measured on total stockholders' equity the return is 19.9% against a 3.7 times multiple, a much less demanding pairing. And the implied figure is only as good as assumption A01: at a 9.0% cost of equity the requirement falls to about 24.4%, which the reported return already meets.
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on price for each $1 of net worth excluding goodwill, cheapest first:
| Ticker | Company | price for each $1 of net worth excluding goodwill |
|---|---|---|
| BOW | Bowhead Specialty Holdings | 2.3× |
| SKWD | Skyward Specialty Insurance Group | 2.9× |
| WRB | W. R. Berkley | 2.9× |
| RLI | RLI | 3.4× |
| KNSL | Kinsale Capital Group | 4.2× |
| PLMR | Palomar Holdings | 4.9× |
The middle of the other 5 is 2.9. Palomar Holdings, Inc. is dearest of 6.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
The dearest of its filed rivals on net worth, with middling returns among them, so the high price is this company's own rather than the group's.
Private or foreign rivals that cannot be priced: Lloyd's of London, California Earthquake Authority, National Flood Insurance Program, The other ten federal crop Approved Insurance Providers.
The bet: what it pays if it works
The bet is on crop insurance, Palomar's fastest-growing line. Its premium is rising fast, but this year's crop is booked at close to no profit, so the rest of Palomar is counted separately, at the median price to net worth of the insurers it names.
If crop reaches management's target and earns the federal programme's long-run average, the shares are worth only a little more than today. If it never earns, shareholders keep about three-fifths.
| Palomar Crop: federal crop insurance earnings in two to three years | Hundreds of millions of dollars of crop premium needed | Valued like its peers | Like the cheapest |
|---|---|---|---|
| $100.00M | about 7 at 100% kept | 1.0× | 1.0× |
| $150.00M (the big win) | about 10 at 100% kept | 1.2× | 1.1× |
To pay 2× (a Fair bet) it would need about $360.40M of Palomar Crop: federal crop insurance earnings (about 24 hundreds of millions of dollars of crop premium); to pay 5× (a Cheap bet), about $1.13B of Palomar Crop: federal crop insurance earnings (about 75 hundreds of millions of dollars of crop premium), at the same multiple.
Each row is Palomar Crop: federal crop insurance alone, plus the rest of the company (Palomar's specialty lines: earthquake, inland marine and property, casualty, surety and fronting) counted at about $2.16B.
The steps it has to take
- Done Crop built to an eighth of gross written premium
- In progress Crop premium passes $400M a year due 2026-12
- In progress The 2026 crop year trued up below a 100% net loss ratio due 2027-03
- Not started A crop year booked at or below the programme's long-run loss ratio due 2028-03
- Not started Crop premium reaches $1B a year due 2029-12
The detail behind the label
The evidence under this - 6 claims
The chief financial officer attributes the figure to the timing of crop accounting rather than to expected results: most crop premium is not recognised until acreage reports arrive in July, so losses run ahead of premium in the first half and the ratio 'trues up' in the third and fourth quarters. The 2025 crop year, booked the same way, ended near 80%, which is close to the federal programme's own long-run average of 0.85 - so the pattern has completed once already at a normal outcome.
These growth rates are stated on the call and not tabulated in the 10-Q, which reports gross written premium only in total. The surety and crop figures are also not like-for-like organic growth: Gray Surety was acquired on 31 January 2026 and Advanced AgProtection in April 2025, so a large part of both is bought rather than grown. Total gross written premium growth of 27.0% for the quarter is the one figure in this set that is filed.
No date is attached to either target, and the crop ambition has to be read against the federal programme it sells into: 89% of the acreage of eight major field crops is already enrolled, and the whole pool of underwriting gains shared among every approved insurance provider was $2.31B in 2024. Growth of that size is therefore share taken from the other eleven providers in a programme whose pricing and terms are set by a federal agreement, not a market expanding underneath the company. Asked directly whether reserve releases must grow to fund Palomar 2X, the chief financial officer said he hoped they would continue but called them 'one smaller component'.
Where we disagree with the market
A proven compounder: fifteen straight beats, three raises this year and a 26% return on equity.
The returns are real but their make-up changed: earthquake stalled, growth moved to weaker lines, and reserve releases supply 27.8% of underwriting profit.
The market sees the record and the raised forecasts. What nobody sets beside the price is the state regulator's own earthquake data, showing take-up falling since 2020 in the franchise that built the reputation.
We would be wrong if casualty policies from 2024 to 2026 cost less than reserved, crop claims settle near the federal programme's long-run 85% of premium, and net worth after goodwill starts compounding again.
What analysts say, and the evidence
The evidence under this - 7 claims
Beating guidance 15 times running is as much a statement about how guidance is set as about delivery, and the chief financial officer described the reserving philosophy in the same terms - take the bad news early, let good news arrive later. The independently checkable half is the analyst estimate, and the pack records beats there too, on the adjusted basis consensus is struck on. The one place the record is thinner is that no large catastrophe has tested it: the earthquake retention has not been breached in this period.
These growth rates are stated on the call and not tabulated in the 10-Q, which reports gross written premium only in total. The surety and crop figures are also not like-for-like organic growth: Gray Surety was acquired on 31 January 2026 and Advanced AgProtection in April 2025, so a large part of both is bought rather than grown. Total gross written premium growth of 27.0% for the quarter is the one figure in this set that is filed.
Measured against adjusted underwriting income, which is the figure management steers on, the share is lower - 19.0% in the first half of 2026 against 10.8% - because adjusted underwriting income adds back stock compensation, intangible amortisation and transaction costs. The direction and roughly the doubling hold on either denominator. Management also argues the releases are the designed output of reserving conservatively rather than a prop: the same pattern has run for six consecutive quarters and began in the property book before reaching casualty.
The figure appears in no filing - the Q2 2026 10-Q does not use the term IBNR at all - so it rests on the call. It is corroborated in direction and roughly in level by the 10-K's own filed split at 31 December 2025, which puts net IBNR at 82.2% of net reserves. Management presents the high share as conservatism, and the six consecutive quarters of favourable development are evidence for that reading rather than against it: reserves that were too high are exactly what releases come from.
What could go wrong, and how we would know
The risks
- Reserves nobody has tested. More than 84% of the money set aside for casualty claims is for claims nobody has filed yet, in a business under three years old.
- Growth in thinner lines. Casualty sales rose from $90.4M to $542.9M in two years, and this year's crop is booked with claims near 100% of premium.
- Two agencies, a third of sales. Arrowhead brings in 20.3% of sales and a second agency 11.7%, under agreements none of which is filed.
- Prices softening. Commercial earthquake renewal rates fell more than 20% in the June quarter, and some large new business is priced below what Palomar considers adequate.
- Losses that are not claims. About a tenth of first-half losses, $18.5M, came from derivatives rather than insurance claims.
The evidence under this - 6 claims
The figure appears in no filing - the Q2 2026 10-Q does not use the term IBNR at all - so it rests on the call. It is corroborated in direction and roughly in level by the 10-K's own filed split at 31 December 2025, which puts net IBNR at 82.2% of net reserves. Management presents the high share as conservatism, and the six consecutive quarters of favourable development are evidence for that reading rather than against it: reserves that were too high are exactly what releases come from.
The chief financial officer attributes the figure to the timing of crop accounting rather than to expected results: most crop premium is not recognised until acreage reports arrive in July, so losses run ahead of premium in the first half and the ratio 'trues up' in the third and fourth quarters. The 2025 crop year, booked the same way, ended near 80%, which is close to the federal programme's own long-run average of 0.85 - so the pattern has completed once already at a normal outcome.
Management frames this as evidence of discipline - that it is walking away - rather than as share loss, and commercial earthquake is 36% of the earthquake book. It is also self-serving in one direction: a company describing rivals as pricing below technical levels is describing business it did not win. RLI's filing independently reports catastrophe reinsurance rate decreases of 15% to 20%, which is the cost side of the same soft market and cuts the other way, because cheaper reinsurance is a margin benefit.
What we still do not know
- Whether casualty policies from 2024 to 2026 cost more or less than reserved. Only time on the books will tell.
- How each product line really earns. Palomar reports a single segment, so no line carries an audited profit figure.
- What the Arrowhead agreement says. Like every programme agreement, it is not filed.
The evidence under this - 3 claims
The figure appears in no filing - the Q2 2026 10-Q does not use the term IBNR at all - so it rests on the call. It is corroborated in direction and roughly in level by the 10-K's own filed split at 31 December 2025, which puts net IBNR at 82.2% of net reserves. Management presents the high share as conservatism, and the six consecutive quarters of favourable development are evidence for that reading rather than against it: reserves that were too high are exactly what releases come from.
What would change our view
- Unattractive→AttractiveThe return is earned on current business, at a cohort-like multiple
It needs all of:
- ○Casualty policies from 2024 to 2026 cost less than reserved, and adjusted claims plus costs stay below 78% of premium for four quarters.checked at the next results, due by 5 Nov
- ○The shares fall to about 2.90 times net worth after goodwill, the middle of its filed rivals.about $78.03 today, 43% below the last close
- Reopens the researchThe earthquake book starts growing again
It needs:
- ○Earthquake sales grow more than 10% for two quarters running, and net worth after goodwill per share beats its December 2025 level for two quarters.checked at the next results, due by 5 Nov
We would also be wrong if
- Its return on equity rises above the middle of the five rivals its report names.
What is coming
- 5 November 2026 Q3 2026 results: The crop quarter: whether prior-year reserve releases still carry about a third of underwriting income and the accident-year loss ratio climbs past 39.5%.
Every dated event ahead
- 5 November 2026 Q3 2026 results: The crop quarter: whether prior-year reserve releases still carry about a third of underwriting income and the accident-year loss ratio climbs past 39.5%.
- February 2027 Q4 2026 results and the 2026 10-K: The year's reserve tables: whether casualty claims still sit mostly unreported, and how the 2026 crop year trued up against the near-100% booking.
- April 2027 2027 proxy statement: Whether the return-on-equity goals in the pay plan are raised toward what the company earns, or left where they pay the maximum.
- 1 June 2027 June 1 reinsurance renewal: What the earthquake tail costs to lay off, and whether commercial earthquake rate cuts are still running at the renewal.
Since we wrote this
- 11 OctThesis reviewPalomar Crop declared an early business, with a betThe owner's rule of 2026-10-11: every thesis names its lines, and an early line owes a bet. Crop is Palomar's fastest-growing line and is booked near no profit this year.
- 5 OctThesis reviewThe long case now sits under the verdictAfter the five-name pilot, the owner asked for the long case on every thesis.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the proxy and filings.
- 26 SepThesis reviewJournal completed from the written record; view heldThe opening entry lacked the view and conviction the thesis was written with; they are filled from the original file. Nothing in the argument moved.
- 7 SepThesis writtenUnattractive, medium convictionWritten a month after a record quarter in which earthquake premium grew 1% and releases from old reserves supplied a growing share of profit.
Sources 14
Sources - 14
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-02-24Item 1 in full: the five product lines, the gross written premium tables by state and by product for 2023 to 2025, the Competition section naming five listed rivals and the California Earthquake Authority, the program administrator concentration table, the distribution model, and the case-versus-IBNR reserve split at 31 December 2025 and 2024.backsC01C02C03C07C13C15C18C19C27C38C47C52C53C54C55C59C60C61C62
- S02primarycompany filing2026-08-05The balance sheet, the reserve for losses roll-forward separating current-year from prior-year incurred losses, the goodwill and intangibles roll-forward with future amortisation guidance, the Gray Surety purchase price allocation and its revenue and net loss contribution, the single-operating-segment statement, the term loan terms and the share repurchase detail.backsC14C22C25C26C29C31C32C33C35C37C39C42C45C48C50C60C61
- S03primarycompany release2026-08-04The quarterly and half-year operating summary, the adjusted-net-income reconciliation naming every add-back, the key ratios, the tangible stockholders' equity reconciliation for 30 June 2026 and 31 December 2025, the dividend announcement and the raised full-year adjusted net income guidance.backsC21C26C30C32C34C35C36C37C43C44C55
- S04primarytranscript2026-08-05The figures with no filed home: residential earthquake premium retention, the 84% casualty IBNR share, growth by product line for the quarter, commercial earthquake and layered property rate changes, the crop premium outlook and the loss ratio at which the 2026 crop year is being booked, the ISO general liability loss cost benchmark, and management's stated expectation that the loss ratio keeps rising. The question section carried the reserve-release exchange the case against rests on.
New business production was very strong in the quarter with both new business premium and policy count increasing year-over-year from the second quarter of 2025. Premium retention exceeded 96% and renewal policies continue to include a 10% inflation guard. ... In Commercial Earthquake, which now constitutes 36% of the Earthquake premium, market conditions remain highly competitive. Average rate decrease in the book was more than 20%. ... Lastly, we maintain a conservative reserving philosophy with more than 84% of casualty reserves held with IBNR. ... For the Earthquake franchise, year-to-date written premium is up 1% year-over-year. ... Notably, auto exposure is intentionally limited. We do not write auto within our E&S casualty business and only 9% of our primary general liability policies have auto coverage. ... Importantly, Laulima, the reciprocal we manage and use to write Hawaii hurricane business, purchased its own reinsurance and maintains $1.5 million event retention.
backsC04C10C20C23C24C28C29C43C49C51C62 - S05primarycompany release2026-02-11FY2025 net income, adjusted net income, combined ratio and return on equity - the base every 2026 growth rate in this thesis is struck against, and the half-year split needed for the guidance arithmetic.backsC35
- S06primarycompany release2026-05-29The reinsurance tower: $3.92B of earthquake limit, $135M of continental United States hurricane limit, the $20M and $11M per-occurrence retentions, the seventh Torrey Pines Re catastrophe bond, and the statement that coverage exceeds the 1-in-250-year peak zone probable maximum loss.backsC40
- S07primaryregulator2026-07-03California Department of Insurance, Earthquake Premium and Policy Count Data Call - Summary of 2025 Residential Totals California Department of InsuranceThe regulator's own count of the whole California residential earthquake market for 2025: 12,754,225 residential policies in force, 1,591,648 of them carrying earthquake cover (12.48%), an average earthquake premium of $956.43 and $1,522,292,004 of written premium, split between the California Earthquake Authority at 968,479 policies and 60.85% and non-CEA carriers at 623,169 and 39.15%. This is the prize measured from outside the company.
Insurers with EQ coverage provided by California Earthquake Authority (CEA) ... 934,116,138$ 968,479 595,329,683,756$ 964.52$ 1.57$ 60.85%11.08% Insurers with EQ coverage provided by Non-CEA ... 588,175,866 623,169 297,809,149,294 943.85 1.98 39.15%15.53% Total Residential Market 20,205,816,926$ 12,754,225 5,766,074,487,463$ 1,584.24$ 3.50$ 100.00%1,522,292,004$ 1,591,648 893,138,833,050$ 956.43$ 1.70$ 100.00%12.48%
backsC05C06C11C12 - S08primaryregulator2021-01-01California Department of Insurance, Earthquake Premium and Policy Count Data Call - Summary of 2020 Residential Market Totals California Department of InsuranceThe same series five years earlier - 1,687,712 earthquake policies at a 13.77% take-up rate and a $776.11 average premium - which is what makes the 2025 reading a trend rather than a level, and shows the market grew on price while shrinking in policies.
Insurers with EQ coverage provided by Non-CEA ... 241,618,734,925 854.03 1.95 32.63%18.31% Total Residential Market 10,776,213,180$ 12,255,597 4,218,359,359,690$ 879.29$ 2.55$ 100.00%1,309,849,432$ 1,687,712 808,688,181,853$ 776.11$ 1.62$ 100.00%13.77%
backsC05C12 - S09primarygovernment data2026-09-07USDA Economic Research Service, Risk Management - Crop Insurance at a Glance United States Department of Agriculture, Economic Research ServiceThe federal crop programme measured by the agency that runs it: a long-run loss ratio of 0.85 since 1997, 89% of the acreage of eight major field crops already enrolled in 2024, and the size of the pool approved insurance providers actually earn from - $2.31B of underwriting gains and $2.34B of administrative reimbursement in 2024.
In 2024, roughly 89 percent of acreage of eight major U.S. field crops (barley, corn, cotton, oats, rice, sorghum, soybeans, and wheat) were enrolled in FCIP. ... Premium subsidies totaled $10.4 billion for 2024. In addition, FCIP incurs other costs due to its public-private partnership structure, which is defined via the Standard Reinsurance Agreement (SRA). ... In 2024, the cost of the program delivery (administrative costs) was $2.34 billion and $2.31 billion was paid as underwriting gains to AIPs. ... Since 1997, FCIP's annual loss ratios has averaged 0.85, indicating that average annual indemnities total about 85 percent of total premiums.
backsC16C17 - S10primarycompany filing2026-02-20A named competitor's own account of the same market: that it underwrites property, casualty and surety on admitted and E&S paper, its California earthquake reinsurance retention of $25M against $850M of limit, that it secured 15 to 20 percent rate decreases on its catastrophe programs, and that capacity from managing general agents is increasing competition across property and casualty markets.backsC09C41
- S11primarycompany filing2026-02-20Kinsale Capital Group, Inc., Form 10-K for the fiscal year ended December 31, 2025 Kinsale Capital Group, Inc.The cohort's strongest performer describing itself: one reportable segment, exclusive focus on the United States excess and surplus lines market, and a 2025 gross written premium mix of 70.7% casualty and 29.3% property.backsC08
- S12primaryvendor data2026-09-07Research pack for PLMR, frozen at content/PLMR/research-pack.jsoncontent/PLMR/research-pack.jsonPrice and market value, the 20-day range and the return series, the analyst split and consensus estimates, short interest from FINRA, the insider filing window, and the filed competitor cohort's multiples - including price to tangible book and return on equity for all six names. Also the cross-check flagging the vendor's trailing sales multiples as unusable for this filer.backsC36C37C44C46C50
- S13primarycompany filing2026-08-24The transaction codes behind the pack's aggregate insider line: three sales at code S totalling 3,500 shares, against 322,388 shares still held afterwards - which is what turns 'the chief executive sold' into a much weaker signal than the count suggests.backsC46
- S14primarycompany filing2026-04-10executive officers and their tenure, beneficial ownership as of March 31, 2026, the 2025 annual incentive and PSU metrics, goals and payouts, the 2023 PSU payout, the 2026 PSU goals, the adjusted return on equity reconciliation for 2021 to 2025, and related-person employmentbacksC56C57C58C59