ACIC — American Coastal Insurance Corporation · 2026-09-09 · Verdict: WATCH · Conviction 2
Price $9.31 (screen row, universe_v2.csv, screen built 2026-09-09; no live price used) · Mkt cap ~$440M (47.3M shares implied by equity / BVPS) · P/TBV 1.58x · P/E 4.6x trailing GAAP, 6.1x normalised, 8.0x on the current run rate · Senior notes $150M due Dec 2027 · ADV $2.4M Sources read: 10-K 2026-03-09 (Items 1, 1A, 7), 10-Q 2026-08-06 (Item 2 MD&A). No DEF 14A, 8-K, transcript or Form 4 data is in the bundle for this ticker, so ownership, compensation and insider activity are not covered below.
Desk stats - Revenue trend: FY2025 total revenue $335.4M, up 13.1% from $296.7M, but that headline is a reinsurance artefact: gross premiums written fell 5.4% to $612.5M, and revenue rose because the external quota share cession was cut from 40% to 20% to 15%, dropping ceded premiums earned 6.4% (10-K, Item 7). Q2 2026 total revenue $82.6M, down 4.5%; gross premiums earned down 16.2% to $138.7M, "attributed to a 24% decrease in our net pricing year-over-year as the market continued to soften", with policies in force up 1.2% to 4,453 (10-Q, Item 2). - Normalised after-tax operating profit: TTM (FY2025 less H1 2025 plus H1 2026) pre-tax income $133.1M. Adjustments: strip $12.8M of net realized investment gains and unrealized gains on equity securities (10-K and 10-Q income statements); add a normal catastrophe load of $25.8M, being the five-year average current-year cat impact of 8.7% of net earned premium (5.4%, 23.5%, 4.9%, 9.3%, 0.5% for 2021-2025, 10-K Item 7 development table) against the $4.7M actually incurred; strip $3.3M of net favorable prior-year reserve development. Discontinued operations net to zero in the TTM window, and the $4.5M non-recurring employee retention tax credit refund (10-Q) falls in both FY2025 and H1 2025 so it cancels. Normalised pre-tax $96.1M, taxed at 25% = $72.1M normalised after-tax against $100.2M TTM GAAP net income. - EV / normalised after-tax profit: EV is not a meaningful measure for an insurer, and the desk prompt bars EV/EBIT here. Price / normalised after-tax profit is 6.1x on the trailing base and 8.0x on the current earned-premium run rate (below). - Financial-lane metrics: P/E 4.6x on TTM diluted EPS of $2.02, 5.0x on TTM core income of $92.0M, 6.1x normalised. P/TBV 1.58x (tangible book $279.1M, $5.91 a share). ROTE 39.6% trailing on average tangible equity, about 19.7% on the normalised run rate. Book value per share $7.21 at 30 June 2026 against $6.00 a year earlier, up 20.2% (10-Q). Equity is 27.4% of assets and AmCoastal's RBC ratio is 1,757% (10-K, Item 1). - Is the growth sustainable? No. It is organic and cash-backed (H1 2026 operating cash flow $52.8M), but the reported growth came from ceding less, not selling more, and net pricing is down 24% year over year while total insured value rose from $69B to an estimated $78B (10-K and 10-Q reinsurance sections). Price per unit of exposure is falling on both axes. - What the screen got wrong: (1) The 4.2x P/E is a look-back on a year in which "no named storms made landfall in our geographic footprint" (10-K, Item 7), a 0.5% cat load against a five-year average of 8.7%. (2) "NI growth 41.1%" is measured against a 2024 base carrying $25.4M of retained Hurricane Milton losses. (3) "Revenue growth" is the quota share cut, not sales. (4) The screen's 48.66M share count is about 3% above the ~47.3M implied by equity of $340.8M over the stated $7.21 BVPS, consistent with the treasury repurchases in the 10-Q, so market cap is nearer $440M than $453M. (5) Dividend yield shows 0.0%, but H1 2026 financing outflows of $56.0M were "driven by the payment of dividends and treasury stock repurchases" (10-Q), while the 10-K says there is no ordinary quarterly dividend, so this was a special distribution the bundle does not size.
1. What the business actually does
ACIC, renamed from United Insurance Holdings in July 2023, writes commercial multi-peril property insurance for Florida condominium associations, apartments and assisted living facilities through one carrier, AmCoastal (10-K, Item 1). One state, one product line, 4,453 policies, 68 employees. Condominium business comes entirely through an exclusive managing general agency contract with AmRisc, which "represents 100% of our condominium revenue"; apartment and assisted living business runs through affiliated MGA Skyway Underwriters. The former personal lines carrier UPC went into receivership in 2023 and Interboro was sold to Forza on 1 April 2025 for $25.7M cash.
2. Why it is mispriced — the edge case
There is no clean edge case, so the verdict is capped at WATCH. This is a Nasdaq-listed $440M insurer with an ordinary float, no spin-off, no index event and no identified forced seller; it is 24% below its 52-week high and down 19% over six months for the reason on the tape, which is that Florida commercial property pricing has fallen 24% and the premium base is shrinking. The one company-specific overhang is the FLOIR proceeding under s.624.4073 Fla. Stat., requiring AmCoastal's officers and directors to show they were not "a substantial contributing cause of the insolvency" of UPC, unresolved at the 10-K date (10-K, Item 1A). That is a management-continuity risk, not a mispricing.
3. Underwriting quality, combined ratio and reserve development
The underwriting record is genuinely good. Combined ratio 60.9% in 2023, 67.5% in 2024, 60.1% in 2025; underlying 60.9%, 59.6%, 61.5% (10-K, Item 7). Prior-year development was favorable every year from 2021 to 2025: $4.2M, $8.0M, $12.7M, $3.7M, $5.8M. The deterioration is now visible: Q2 2026 combined ratio 74.3% against 60.6%, underlying 68.7% against 62.2%, expense ratio 47.3% against 40.8% because fixed costs and reinsurance sit on a net earned premium base of $69.7M, down 11.1%. The underlying gross loss ratio is stable at 10.8% of gross earned premium, so this is a price and expense-leverage story, not a loss-cost story. Q2 2026 also carried $767K of unfavorable prior-year development, the first break in a five-year run of releases (10-Q).
The June 2026 reinsurance renewal is the item to underline. Core programme limit rose to $1.68B first occurrence, but GAAP first-event retention rose from $29.75M to $49.0M and second-event retention from $18.5M to $25.0M (10-K and 10-Q). ACIC is retaining roughly 65% more per event on 13% more insured value at 24% lower pricing.
4. Balance sheet and capital allocation
$650.0M of cash, restricted cash and investments at 30 June 2026, 81.7% of fixed maturities Treasuries or A-rated or better. Debt is $150M of senior notes maturing 15 December 2027, whose coupon fell from 7.25% back to 6.25% effective 16 December 2025 after Kroll upgraded the issuer to BBB- in July 2025 (10-K, Item 7). Holdco liquidity looks fine: AmCoastal upstreamed $87M in H1 2026 against $23M in all of 2025. Capital return has turned real, with $56.0M of dividends and buybacks in H1 2026 and the share count down about 3%. Against that, the 2023 at-the-market equity programme has 3.6M shares still available, so dilution optionality remains. $15M was contributed to a new excess and surplus entity that had not begun operations at the 10-K date.
5. Management: what they said versus what they did
The 10-K states the vision is "to be a top-quartile underwriter of catastrophe-exposed property insurance" with a process balancing "risk appetite and underwriting profit opportunities with our available capital." The 2026 renewal tests that claim: they took a materially larger net retention into a softening market rather than shrinking the book, and gross written premium still fell 14.2% in H1. There is no transcript or 8-K in the bundle, so guidance versus delivery cannot be checked directly here.
6. Valuation
Forward run rate, from H1 2026 annualised: net earned premium $270.6M, operating expenses $132.1M, underlying losses at 19.8% of NPE ($53.6M), normalised cat load 8.7% ($23.5M), giving underwriting profit $61.3M; plus $21.0M net investment income, less $9.4M interest, equals $73.2M pre-tax and $54.9M after tax, or 8.0x at $440M, a 19.7% return on tangible equity.
Base (50%): pricing declines moderate after the 2027 renewal, normalised earnings near $55M at 9x, equals $10.47 a share. Bear (30%): another 15% net price decline earns through, NPE falls to about $200M, expense ratio passes 52%, earnings $35M at 7x, equals $5.18. Bull (20%): a market-turning storm or capacity withdrawal re-hardens Florida commercial property and the E&S entity adds a second engine, earnings $75M at 10x, equals $15.86. Probability-weighted $9.96, about 7% above price.
Reverse DCF: at $9.31 against tangible book of $5.91, the price implies a sustained 18.9% return on tangible equity in perpetuity with zero book growth at a 12% cost of equity, almost exactly the forward run-rate return. The market has the arithmetic about right.
7. Catalysts and timeline
Q3 2026 results in November, the first read on this hurricane season at the new $49M retention. The 1 June 2027 reinsurance renewal and the pricing disclosed alongside it. Refinancing or repayment of the $150M notes before December 2027. First written premium from the new E&S entity. Continued buybacks at 1.6x tangible book.
8. Risks and pre-registered kill criteria
- Gross premiums written down more than 15% year over year in either Q3 or Q4 2026, or the June 2027 renewal disclosing another double-digit net pricing decline.
- Underlying combined ratio above 75% for two consecutive quarters (68.7% in Q2 2026, 62.2% a year earlier).
- Net unfavorable prior-year reserve development in two consecutive quarters, following the $767K unfavorable in Q2 2026.
- The $150M senior notes not addressed by 30 June 2027, or refinanced above 7.25%. Other risks: total dependence on AmRisc, which has "substantial leverage in negotiations"; a Demotech A or Kroll A- downgrade; the unresolved FLOIR s.624.4073 matter; and a single event above the $49M retention plus reinstatement cost, which would roughly halve a year's earnings.
9. Verdict and one-paragraph summary
WATCH, conviction 2. American Coastal is a genuinely well-run Florida commercial property underwriter, with a 60.1% combined ratio in 2025, five straight years of favorable reserve development, RBC of 1,757%, $650M of investments and a Kroll upgrade that cut its coupon; but the screen's 4.2x P/E is a picture of a year in which no named storm made landfall, and putting back a normal 8.7% catastrophe load while stripping $12.8M of investment gains and $3.3M of reserve releases takes normalised after-tax profit to $72M and the multiple to 6.1x, with the current run rate of $55M at 8.0x. It is not an IDEA because everything is moving the wrong way at once: management says net pricing is down 24% as the market softens, gross written premium fell 14.2% in the first half, the underlying combined ratio and the expense ratio each rose about 6.5 points on a shrinking net premium base, the first unfavorable reserve development in five years appeared in Q2, and the June 2026 renewal raised first-event GAAP retention from $29.75M to $49.0M on insured value up from $69B to $78B, so the company keeps more risk per event on more exposure for less price. At $9.31 it trades at 1.58x tangible book, implying a sustained 19% return on tangible equity with no growth, which is essentially what the forward run rate produces, so there is no discount to pay for that risk and no forced seller or misunderstood segment to explain one. Revisit after the Q3 print and the 2027 renewal pricing; a re-hardening market with the book still intact would make this cheap quickly.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.