Conduit Holdings: cheap book value, unproven cycle

Conduit Holdings is the parent of Conduit Reinsurance, a Bermuda-based multi-line reinsurer writing property, casualty and specialty business. It earns premiums for taking on tail risk, invests the float in a conservative fixed-income portfolio, and pays claims when hurricanes, earthquakes or other large losses hit. After a rough 2025, H1 2026 looked much better: combined ratio down sharply, comprehensive income back in the black, and buybacks running ahead of dividends. The question is how much of that improvement is skill and how much is simply a quiet catastrophe season.

At 462p on September 25, 2026, the shares trade 15.2% below my triangulated Base value of 545p, an upside of 18.0%. The Bear value of 390p is 15.6% below the price. For a reinsurer this exposed to single-event tail risk, the framework's required margin of safety is 30%, which puts the maximum buy price at 382p. Conduit is about 21% above that level, so this stays a watchlist name rather than a buy.

Route: SF-03 Insurance, P&C reinsurance subtype (Bermuda-based multi-line reinsurer). Economic classification: Quality growth / standard insurer blend (8/12 on the sector scorecard). Primary method: justified price to fully diluted tangible net asset value. Secondary method: normalized earnings and residual-income cross-check. Values are present fair-value estimates, not 12-month price targets.

What Conduit is

Conduit Reinsurance Limited, a Bermuda Class 4 insurer rated A- by A.M. Best, underwrites property, casualty and specialty reinsurance for primary insurers around the world. It is younger and smaller than the global reinsurance incumbents, and its edge is underwriting discipline and capital strength rather than brand or scale. As the reinsurance market has softened, management has been actively repositioning the book: growing Casualty, trimming lower-margin Property quota-share exposure, shifting Property toward excess-of-loss, and buying more retrocession and aggregate cover to lower the modeled net probable maximum loss from a major North Atlantic windstorm.

H1 2026 premiums by segment were Property $454.8 million (down 9.3%), Casualty $217.0 million (up 21.2%) and Specialty $117.2 million (down 4.7%). The mix shift toward excess-of-loss and away from quota share is a reasonable risk-selection choice, but it comes with lower premium volume and depends on continued underwriting discipline as competition for business increases. Overall risk-adjusted rate change was minus 6% for the half, with Property down 10%, Casualty down 1% and Specialty down 7%, net of claims inflation.

The latest half-year

The turnaround is real, but the H1 2026 combined ratio was primarily a function of a more benign natural-catastrophe environment than a demonstrated structural improvement in underwriting margin. FY2025's 101.5% and H1 2025's 122.1% are the more sobering reference points for what a normal or bad year can look like.

Which numbers matter for a reinsurer

Free cash flow isn't a meaningful valuation anchor here. The primary metric is fully diluted tangible book value per share, read alongside sustainable ROE and the undiscounted combined ratio, since claims liabilities and invested assets are the real economic balance sheet. Company-reported basic tangible net asset value per share was $7.56 at June 30, 2026; the fully diluted figure, which accounts for equity awards, was $7.39, and this analysis uses the fully diluted number as the valuation anchor.

A through-cycle planning range of 96% to 101% undiscounted combined ratio is used, with sustainable ROE of 9% to 11%, both below the H1 2026 annualized-equivalent result and near or under the FY2025 return. This deliberately does not extrapolate the benign H1 2026 loss experience forward.

Shares, buybacks and capital

At June 30, 2026, Conduit had 165.24 million issued shares, 16.94 million held as own shares, and 3.43 million dilutive shares from equity awards, giving a fully diluted book-value denominator of 151.74 million. Own shares held rose from 10.93 million at December 2025, reflecting $38.9 million of H1 buybacks covering 6.8 million shares, on top of $28.7 million of dividends. Total shareholders' equity was $1,121.1 million. Buybacks below intrinsic value can lift per-share value, but they also consume capital that would otherwise cushion a bad catastrophe year, which is a genuine tradeoff for a company at this stage of its capital build.

Primary valuation: justified price to tangible book

Converting the $7.39 fully diluted book value using the company's own quoted $7.56 / £5.70 equivalence gives approximately £5.57 (557p) fully diluted tangible book per share.

For a reinsurer with a sustainable through-cycle ROE of 9% to 11% against an 11% to 12% cost of equity and 2% to 3% long-term growth, the justified multiple sits below or around 1.0x. The Base case uses 0.98x, a modest discount for catastrophe volatility and softening prices despite improved portfolio protection. The primary Base value is 545p. A sustained combined ratio below 95% with a high-teens ROE could support a richer multiple, but one benign half-year doesn't establish that record.

Secondary valuation: normalized earnings cross-check

H1 diluted EPS was $0.52; simply doubling it would overstate confidence given catastrophe, investment-mark and seasonality noise. Using normalized annual EPS of $0.85 to $1.00 (midpoint $0.92) at a 9x to 10x multiple gives $8.28 to $9.20/share, or roughly £6.25 to £6.95 at the company's published FX equivalence. A residual-income check using 9% to 11% sustainable ROE against an 11% to 12% cost of equity points to a value near or somewhat below tangible book unless underwriting returns persist above the cost of capital, which argues against paying a sustained premium to book until Conduit has a track record through a full catastrophe cycle. This earnings-based approach carries more uncertainty than the book-value method and is treated as an upside cross-check.

Historical cross-check

Fully diluted tangible book value per share was $6.37 in June 2025, $7.03 at December 2025 and $7.39 at June 2026, a genuine growth trend including dividends paid. The current price is about 0.83x that fully diluted figure and 0.81x reported basic TNAV, a discount to current balance-sheet value. Reinsurers can reasonably trade below book when expected future ROE is below cost of equity or catastrophe risk is elevated, so the discount alone isn't proof of mispricing. A reliable long-period price-to-book median wasn't available in the reviewed materials, so book growth and the present discount serve as the cross-check instead of a precise historical multiple.

Valuation triangulation

The weighted calculation comes to about 549p; the working figure is rounded down to 545p to stay consistent with the primary method and avoid false precision. The methods aren't fully independent, since both book value and earnings respond to the same catastrophe and investment outcomes.

What is the market already pricing in?

At 462p, CRE trades at about 0.83x fully diluted tangible book and 0.81x basic tangible book. That's a meaningful discount, consistent with the market demanding either a lower normalized ROE than the 9% to 11% range used here, or a substantial risk premium for catastrophe volatility, or both. It is not a distressed valuation if the company preserves capital and earns roughly its cost of equity, but it does embed real skepticism about sustained above-cost returns. No precise reverse-DCF is used, since a reliable GBP cost-of-equity curve and forward capital schedule aren't available from public filings.

Bear, base and bull scenarios

Double-pessimism control: the Bear case combines a lower tangible-book multiple with weak ROE deliberately, because a major catastrophe year can impair both book value and investor confidence at the same time; no separate additional catastrophe haircut is stacked on top.

The probability-weighted value is 533p (390p × 25% + 545p × 55% + 680p × 20%), about 15% above the current 462p price.

Sensitivity: price to fully diluted TNAV

How the buy price is set

A 30% margin of safety is required for a reinsurer this exposed to single-event tail risk: results can shift sharply after one hurricane, earthquake or specialty loss event, and the H1 2026 combined ratio isn't a stable earnings guarantee. Applied to the 545p Base value, that gives a maximum buy price of 382p, with a practical buy zone of 350p to 385p assuming capital stays intact and pricing remains adequate. At 462p the shares trade about 21% above that ceiling.

Dividend

The H1 2026 interim dividend was $0.18/share, about 13p at the company's reporting equivalence. Combined with a similar prior payment, the approximate annual rate is $0.36, implying a yield of roughly 5.9% at 462p, assuming that rate continues. H1 2026 buybacks of $38.9 million exceeded the $28.7 million of dividends paid, so repurchases, not the dividend, are the larger form of capital return right now. The dividend is meaningful but not guaranteed: management should prioritize capital adequacy, retrocession and underwriting opportunity over defending a fixed payout through an adverse catastrophe period.

Reasons to own CRE

• Discount to book. Shares trade at a clear discount to fully diluted tangible book value.
• Real turnaround. H1 2026 returned to positive comprehensive income and a 7.8% ROE after H1 2025 losses.
• Improving underwriting. Undiscounted combined ratio improved to 92.6% from 122.1% a year earlier.
• Deliberate mix shift. Growing excess-of-loss and Casualty business while trimming lower-margin quota share.
• Reduced tail risk. Enhanced retrocession lowered the modeled net North Atlantic windstorm PML.
• Conservative portfolio. AA average credit quality, short duration, no equity or alternative assets.
• Growing book value. Tangible book per share has grown materially over the past year, and buybacks below intrinsic value add to that.
• Income while waiting. The interim dividend offers a return while the market re-rates the underwriting story, if it does.

What could go wrong?

• Catastrophe risk. A severe Atlantic hurricane, earthquake, wildfire or specialty loss can materially increase claims and reduce tangible book value; this is the main thesis breaker.
• Not-yet-normalized results. The current combined ratio benefited from a benign cat environment; FY2025 was 101.5% and H1 2025 was 122.1%.
• Softening pricing. Weaker reinsurance rates and broader terms can reduce risk-adjusted returns; H1 risk-adjusted rate change was minus 6%.
• Competitive capacity. High global reinsurance capital could pressure renewals, especially in Property and Specialty.
• Reserve uncertainty. Long-tail Casualty reserves in particular carry real estimation risk.
• Retrocession cost or gaps. Protection could become more expensive, unavailable, or less valuable than modeled.
• Investment volatility. Rising rates or wider credit spreads can create unrealized losses even as reinvestment yields improve.
• Capital return tradeoff. Buybacks and dividends consume capital that could otherwise support underwriting capacity or absorb losses.
• Scale. As a smaller reinsurer, Conduit has less diversification and scale than global incumbents.

Management execution

Management showed active cycle management in H1: Property premiums down 9.3%, Specialty down 4.7%, Casualty up 21.2%, alongside increased retrocession protection. Capital returned in the half was $67.6 million, split between $38.9 million of buybacks and $28.7 million of dividends. Tangible net assets per share, including dividends paid, rose 8.4% since year-end 2025 and 23.2% year over year. The caveat is that the 92.6% combined ratio mostly reflects lower catastrophe losses rather than a proven structural shift, and the ROE methodology was amended in 2026 with prior periods re-presented, which complicates casual comparisons. Assessment: constructive capital and underwriting actions, moderate confidence pending performance through a less benign catastrophe period.

Stock Analyza scorecard

Business quality score is 69/100: improving underwriting and a solid capital position, offset by catastrophe volatility and a softening market. The valuation score is 70/100: the price sits below fully diluted tangible book, but the latest half benefited from unusually benign catastrophe experience.

Confidence score

The interim report is detailed and company-reviewed. Confidence is limited by catastrophe randomness, a one-year recovery comparison, softening pricing, and general uncertainty in long-run reinsurance returns.

What I would watch from here

Also watch capital return discipline: buybacks below estimated value with buffers intact are green, capital distributions that compromise underwriting capacity are red. Review after each half-year and full-year report, and immediately after significant catastrophe events, renewal seasons, reserve updates or retrocession changes.

The three most important thesis breakers

1. A major catastrophe or reserve deterioration that materially erodes TNAV.
2. Combined ratio drifting back above 100% on a sustained basis as pricing softens.
3. Retrocession protection proving insufficient or unaffordable relative to what's modeled.

Adversarial review

The first draft's arithmetic mostly reconciled: the price-to-book multiples, the primary scenario values, and the triangulation all check out. One number was wrong. The probability-weighted scenario value was stated as 527p, but the draft's own scenarios and probabilities (390p × 25% + 545p × 55% + 680p × 20%) actually sum to 533p. The corrected figure is used throughout this article; it remains meaningfully above the current price, so the Watchlist conclusion, driven by the margin-of-safety rule rather than the expected value, is unchanged.

The strongest Bear counter-thesis: the H1 2026 combined ratio is not a normalized measure, and a return to 2025-style loss experience, or a single major catastrophe, could erase a significant portion of the current book-value discount. Rate declines and ample global reinsurance capacity could also compress future margins just as ceded protection costs rise. The strongest Bull counter-thesis: a disciplined reinsurer can simply decline underpriced business, grow higher-margin excess-of-loss and Casualty lines, hold an AA short-duration portfolio, and compound tangible book value through favorable underwriting and buybacks; if returns stay above the cost of equity, the current discount could close on its own.

Audit limitations: precise peer-period price-to-tangible-book comparisons and the exact catastrophe PML dollar figures weren't obtained; the diluted GBP book value relies on the company's own rounded USD/GBP equivalence rather than an independent FX spot estimate; and a full independent actuarial reserve review isn't possible from public filings alone. These are disclosed limits that don't change the Watchlist conclusion, but they keep the confidence grade at C. The publication gate status is PASS WITH WARNING.

Final verdict: hold at 462p, watchlist for new money below 382p

Conduit has recovered from a difficult prior period, trades below tangible book, and is actively strengthening its retrocession and portfolio mix. But H1 underwriting benefited from benign catastrophe losses while market pricing is softening, so the shares don't yet offer the full catastrophe-risk margin of safety this business needs. The main thesis breaker is a major event or reserve deterioration that materially erodes TNAV and capital resilience.

Verdict: WATCHLIST at 462p. Triangulated Base value 545p, upside 18.0%, Confidence C (16/25). Required margin of safety 30%, maximum buy price 382p.

A move into the 350p to 385p buy zone without thesis deterioration, or a demonstrated combined ratio in the mid-to-high 90s through an average or worse catastrophe year, would meaningfully improve the case here.

Sources

Primary sources include Conduit Holdings' 2026 Interim Report, its Q1 2026 trading update, and its official results archive. Market price comes from LSE market data.

The reference market price is 462p, the September 25, 2026 LSE close. Balance sheet figures are as of June 30, 2026. Forward combined ratio and ROE are estimates. GBP figures derived from USD amounts use the company's own published $7.56 / £5.70 equivalence rather than a separately sourced FX spot rate.