James River Group Holdings, Inc. (JRVR) stock intrinsic value fair value estimate & owner earnings analysis

James River Group writes casualty insurance for hard-to-place US risks. Roughly 80-90% of its business is now Excess and Surplus Lines (contractors, excess casualty, general casualty and similar lines, sold through wholesale brokers; three brokers supply about 72% of that segment's premiums). The rest is Specialty Admitted, a fronting business that passes most of its risk to reinsurers and is being deliberately shrunk (its premiums fell 76% in the second quarter of 2026). The company sold its casualty reinsurance arm, JRG Re, in April 2024, so today's company is narrower than the one that existed before. How we measure earnings. For an insurer we start from operating cash flow and strip out every swing in loss reserves, unearned premiums, receivables and reinsurance balances, then subtract stock compensation and the capital spending needed to stay in business. Measured that way the company produced about $66 million (2016), $32 million (2017), $90 million (2018), $30 million (2019), $19 million (2020), a loss of $214 million (2021, when the reinsurance arm took a large reserve hit and commercial auto reserves were strengthened), $68 million (2022), $26 million (2023), a loss of $92 million (2024, the year of the JRG Re sale and the purchase of reserve protection) and $33 million (2025). We anchor on the ten-year median, about $30 million a year. Ten years is used deliberately: the good years of 2016-2019 were later shown to have been under-reserved, and the losses of 2021 and 2024 are the bill for them, so a shorter window catches only one side of that cycle. Owner earnings rose from a loss to $33 million in 2025; we start year one at about $30 million, slightly below that, and credit no growth. What changed since our previous analysis. The adverse development cover bought from Enstar's Cavello Bay, which absorbed further reserve increases on Excess and Surplus Lines accident years 2023 and prior, had $23.6 million of limit left at the end of 2025. In the first half of 2026 those old years developed adversely by about another $21 million, and by the second quarter the cover was fully used. From here, any further strengthening on those years falls directly on shareholders. On the other side, reserve movements on business not covered by that contract were close to zero in both 2026 quarters, the Excess and Surplus Lines combined ratio was 92.8% in the second quarter, and tangible equity per share edged up. We also now treat the Series A preferred (held by GPC Partners, $133.1 million liquidation preference, convertible at $8.32) as a claim ranking ahead of common shareholders whenever converting would not pay its holder, and we credit shareholders with the insurer's tangible capital when the business story ends, so the previous figures are not directly comparable. What is in the bag. Nothing. Excess and Surplus Lines premiums fell 5.3% in 2025 and 17% in the second quarter of 2026, management describes the market as increasingly competitive, and fronting is being cut. No case credits premium growth, rate increases, margin expansion or a stronger insurance market. Base case, about $6.75 per share. Owner earnings of about $30 million a year for 20 years, supported by a 20-plus-year record in this niche and an A- (Excellent) rating from A.M. Best on each insurance subsidiary. Because the old-year development is a demonstrated, continuing pattern and the cover is gone, we charge roughly $16 million after tax in each of 2027 and 2028 for further strengthening on 2023-and-prior years. At the end of 20 years we credit the release of today's tangible capital, about $442 million (shareholders' equity plus the preferred, less $182 million of goodwill and $32 million of intangibles), since an insurer that stops writing can return its capital as claims are paid. That values the whole business at about $455 million. The company also owes $225.8 million of senior debt (mostly a revolving credit facility maturing June 12, 2028) and $104.1 million of junior subordinated debt (maturing from 2034); interest on that debt is already deducted in the earnings above, so it is not subtracted again. The preferred ranks ahead of common and its dividend resets in October 2029 to the five-year Treasury yield plus 5.2% (capped at 8%), so we count it as about a $140 million claim. That leaves about $315 million, or roughly $6.75 per each of 46.3 million common shares. Bear case, about $1.75 per share. One full casualty reserve cycle of the kind this company has already lived through: since 2020 it has recorded about $354 million of cumulative net adverse development. We assume a further $150 million pre-tax (about $120 million after tax) of strengthening spread over 2027 and 2028 with no reinsurance protection left, followed by an A.M. Best downgrade that pushes brokers to place business elsewhere, cutting owner earnings to about $15 million a year for five years and $25 million thereafter, with the story ending after 15 years and only about $300 million of capital left to return. The business is then worth about $222 million against the same $140 million preferred claim, leaving about $82 million, or about $1.75 per common share, and in practice less if new equity has to be raised at low prices to restore capital. The revolving credit facility must be renewed by June 2028, right in the middle of this case. Bull case, about $10.90 per share. The current underwriting discipline holds (Excess and Surplus Lines earned a $59.5 million underwriting profit in 2025 at an 89.4% combined ratio), old-year development stops, and owner earnings settle near $45 million a year for 20 years, still below the 2018 and 2022 peaks, with the same capital release at the end. In that case the preferred holders would convert into 13.5 million shares, and the business is worth about $654 million, or roughly $10.90 per share on 59.8 million shares. Accounting judgements. Reserve and premium movements are removed from cash flow because a shrinking or growing book distorts them; 2024 cash flow was hit by the transfer of $552 million of reinsurance balances in the JRG Re sale and 2019 by a large funds-held movement. Earnings before 2024 include the sold reinsurance arm; we keep those years because their later losses are exactly the record of reserve shortfalls the bear case draws on. For 2016-2018 the cash-flow statement has no separate equipment purchase line, so the small 'other' investing outflow is used as the capital spending figure; it is immaterial. Retroactive reinsurance accounting creates a $107.8 million deferred gain on the balance sheet; we give it no separate value. How we set the margin-of-safety price. We are not sure of the earnings level: the ten-year record runs from a $90 million year to a $214 million loss, and the company has strengthened reserves repeatedly. The worst plausible outcome is a large new reserve charge with no reinsurance cover left, followed by a downgrade; with three brokers supplying most of the core business, $330 million of debt, a revolver due in 2028 and a preferred ranking ahead of common, the structure would survive but common shareholders could be heavily diluted, which is why the bear case is so low. If we are wrong about the stability of the $30 million earnings level or about old-year reserves being nearly fully recognized, the shares are worth something between about $1.75 and tangible common book of about $6.70. During the holding period the business pays only a $0.01 quarterly dividend, so the whole return must come from value recognition, and we estimate that could take three to four years of clean reserve results. At $3.25, the price is about half of tangible common book for a business that, if the base case holds, would be worth about $6.75 within roughly four years, which the base case implies is around a 20% annual return, while the price already discounts much of the bear outcome. The margin-of-safety price of $3.25 applies only once a year of quarterly results shows the old accident years have stopped developing adversely.