Arch Capital Group Ltd. (ACGL): what the price assumes

In the published model solve dated 2026-Q2, anchored at $98.84, Arch Capital Group Ltd. (ACGL) is priced for 10.7% return on equity. boothcheck publishes no house fair value, target price, or buy/sell rating; individual model outputs and user-controlled scenarios are analytical inputs, not Boothcheck targets. Narrative composed 2026-07-25.

Generated: 2026-08-30 · Source: https://boothcheck.com/report/ACGL

Headline

FieldValue
TickerACGL
CompanyArch Capital Group Ltd.
Sector / IndustryFinancial Services
Current price$98.84/sh
CompositionInsurance 46% / Reinsurance 48% / Mortgage 7%

What The Price Assumes (Inversion)

The assumption today's price embeds, recovered by inverting the valuation.

FieldValue
Inversion basisfinancials
Return on equity needed10.7%
Return on equity now18.6%
ROE gap-7.9pp
Price-to-book1.48x

Solve inputs: computed at a 8.5% cost of equity with 4% terminal growth over a 10-year stage, on common book equity (FY2026).

How unusual the bet is: within-range

ReferenceValue
vs own history-0.59σ
cohort percentile (of 78 peers)42
sustained it ~10 years at this level74%
implied end-window share0%

Valuation X-Ray

The price is supported by asset-based and earnings-power value. A value/asset-supported name, not a pure growth bet.

How the valuation models price the stock relative to the market price. Price/FV above 1.0 means the market pays more than that lens defends (expensive); at or below 1.0 the lens can defend the price.

FamilyMedian price/FVModelsReads
Asset0.66x3justifies
Earnings0.70x1justifies
Relative0
Growth0

Families that justify the price: Asset, Earnings

The models below discount at their own flat-beta convention rates (cost of equity 9.3%, WACC 4.7%); the inversion above states its own rate.

Per-Model Detail (n=4)

ModelFamilyFVPrice/FVApplicableMethodology
DCF Perpetual GrowthGrowthno
Bank Fair Value (P/TBV)$246.660.40xyesTBVPS $69.23 × 3.56x (ROE (TTM) 20.1% / CoE 9.3%, g=5.0% (sustainable: 65% retention × ROE, 5% cap; not the terminal-growth assumption))
Relative ValuationRelativenoP/E 11x (static sector reference · 2026-04), scenarios: 9.2x / 11.0x / 12.8x (bear / base = reference held flat / bull), EV/EBITDA 10x
Simple DDMGrowthno
Two-Stage DDMGrowthno
Simple Excess ReturnAsset$150.750.66xyesBV/sh $69.23, ROE (TTM) 20.1%, ke 9.3%
Two-Stage Excess ReturnAsset$220.240.45xyes5yr excess ROE then converge to ke=9.3%
Discounted Future Market CapGrowthnoRev $19.8B, growth 9% (input: historical growth; tapered), Terminal P/S: 1.5x / 1.7x / 2.0x (bear / base = today's held flat / bull, cap 8x)
Peter Lynch Fair ValueRelativenoEPS $13.00, growth 35% (input: historical EPS growth), PEG=0.20 (Undervalued)
Margin TrajectoryGrowthno
Earnings Power ValueEarningsno
Residual IncomeAssetno
Graham NumberAsset$142.300.69xyes√(22.5 × EPS $13.00 × BVPS $69.23) — Graham's conservative floor
EV/EBITDA RelativeRelativeno
FCF YieldEarningsno
SBC-Adj FCF YieldEarningsno
Ben Graham FormulaEarningsnoEPS $13.00 × (8.5 + 2×15.0%) × (4.4 / 5.3%)
ROIC-Justified P/BAssetno
P/Sales SectorRelativeno
PEG Fair ValueRelativenoEPS $13.00 × (PEG 1.5 × growth 25.0% (input: historical EPS growth)) → PE 37.5x
Earnings YieldEarnings$140.540.70xyesEPS $13.00 / required return 9.3% (Rf 4.3% + ERP 5.0%)
Funds From Operations MultipleRelativeno
Clinical Phase NPVGrowthno
MertonAssetno
V5 Mechanicalno

Economic-Unit Decomposition (Sum Of The Parts)

The issuer is a funded financial business. Debt, interest, and cash flows are operating inputs, so industrial EV, net-debt, WACC, and free-cash-flow lenses do not apply; value the common-equity claim with book, earnings, capital, and payout economics.

UnitRoleValuation basisRevenueReported profitValue evidenceStatus
Insurancefinancialequity$7.8bwithheldunresolved standalone equity facts required
Reinsurancefinancialequity$8.1bwithheldunresolved standalone equity facts required
Mortgagefinancialequity$1.2bwithheldunresolved standalone equity facts required

No unit-level total common-equity value is stated. Each financial unit requires supported standalone common equity, normalized earnings, capital adequacy, and payout capacity. Consolidated debt, interest, and cash are operating balances, not an enterprise-to-equity bridge; company-level book, earnings, capital, and payout lenses remain the coherent cross-checks.

Solvency

FieldValue
Share count CAGR (buyback)-1.6%

Deposit/float-funded balance sheet: debt is funding, not corporate leverage, and GAAP operating cash flow follows loan flows. Net-debt, interest-coverage, and cash-burn lenses do not apply. The solvency frame for a financial is regulatory capital and payout capacity (CET1, stress buffer, dividends plus buybacks against earnings).

Bullet Takeaways

Bull Case

Start from an unusual place. Nearly every way of valuing this company off what it already owns and already earns lands above where the stock trades. Book value adjusted for the profitability being earned on it, the conservative asset floors, and earnings multiples applied to trailing profit all sit higher than $103.38. The one approach that lands lower gets there by projecting revenue forward and re-pricing a larger company at the same multiple of sales it carries today. That is a growth test, not a value test. For a business earning a return on equity near 18.6%, the configuration is odd: the market is not paying up for what Arch has demonstrated. It is paying less than the demonstration would support.

The reason to think that demonstration is repeatable rather than lucky is the way the company describes its own operating model. The 10-K says the platform exists to be turned down as well as up: "Our experience as cycle managers is complemented by scalable underwriting platforms enabling us to increase or decrease our business as market conditions demand." That sentence is easy to skim past. It is the whole argument. Insurance companies do not usually get to choose how much business they write, because sales forces, reinsurance treaties and distribution relationships all push toward volume. A company built to shrink on purpose is buying itself the option to sit out a soft market rather than underwrite through it. The same section describes the middle market build-out, where "The MCE platform enhances our U.S. focus on middle market companies using our strategic analytics capabilities and continued focus on customer solutions."

Then there is the mortgage arm, which the revenue weight badly understates. It is 7% of the top line and it does not behave like the rest of the company at all. Look at what its own cohort earns: MGIC converts 59.6% of revenue into net income, Essent 53.6%, NMI 53.8%, Radian 41.1%. Now look at the primary insurance cohort, where Travelers (TRV) converts 15.5% of a $48.9 billion top line, Berkley 12.6% of $14.9 billion, and the best conversion in the group belongs to RLI at 20.8% of $1.9 billion. Mortgage insurance sits in a different economic class because of how its losses get recognized. The 10-K spells the mechanism out: "losses on U.S. primary mortgage insurance business are only recorded at the time a borrower is delinquent on their mortgage". No delinquency, no loss. Arch owns a slice of the most profitable line in the sector and reports it as a rounding error on revenue.

The results follow. Trailing net income of $4.87 billion on $19.8 billion of revenue converts revenue to profit at a rate no name in the primary insurance cohort matches, and the filing attributes part of the mortgage segment's recent favorable reserve development to releases on the U.S. first lien portfolio from the 2020 to 2022 accident years.

Capital allocation is where the argument stops being theoretical. About 86.9% of last fiscal year's earnings went back to shareholders as dividends and buybacks, and the share count has fallen roughly 1.6% a year over the four years to March 2026. That decline is not a policy statement, it is the receipt. The 10-K reports 240.8 million shares held in treasury at an aggregate cost of $6.4 billion as of December 31, 2025, and approximately 455.0 million shares repurchased since the program began. When a company retires shares while trading below what its own book value and profitability support, every dollar spent buys back more earnings power than it would at a full price.

The obvious objection is that the current return reflects a hard pricing market that will not last. It probably does. But a company whose stated design is to write less when rates are unattractive is precisely the one you would want holding capital going into that turn, and the price already assumes a return meaningfully below the current one.

Bear Case

The industry writes its own bear case, and this one is in the risk factors. Property and casualty insurance is "highly cyclical, and we may at times experience periods characterized by excess underwriting capacity and unfavorable premium rates". That is not boilerplate. It is a description of what happens after a stretch of profitable years: capital notices, capacity arrives, and prices fall for everyone still in the water. Arch has just had a stretch of profitable years, and reinsurance is 48% of its revenue, the line where new capacity lands fastest because it requires no distribution network to deploy.

So the question is which return is the real one. Trailing return on equity is about 18.6%. Stretch the lens back across the seventeen-year record behind this company and the average is closer to 14.1%. The gap between those two numbers is the cycle. Capitalize the first and the stock looks cheap; capitalize the second and it looks ordinary. Nothing in the recent results tells you which one the coming years resemble, and the buyer at $103.38 is implicitly taking a view.

Reserve development sharpens the point, because a meaningful part of recent earnings did not come from business written recently. The mortgage segment's favorable development traces to reserve releases on the U.S. first lien portfolio from the 2020 to 2022 accident years, a vintage of loans underwritten into rising home prices. Releases from a fixed pool are finite by construction. Running the other direction, long-tail casualty took $45 million of adverse development from the 2013 to 2020 underwriting years, and casualty is the line where a reserving error takes a decade to surface. The filing lists exactly this exposure among the things that could go wrong: "greater than expected loss ratios on business written by us and adverse development on claim and/or claim expense liabilities".

Now the calibration, honestly stated. The price is not demanding a heroic return. At roughly 1.55 times book it assumes something near 11.2%, which is below what the company earns today and below its own long-run average. That makes this a different kind of bear case than an overvaluation argument, because there is no stretched assumption to attack. The argument is that the market's discount is deliberate and possibly correct. Even the modest return the price asks for has historically slipped: of firms earning at that level, roughly 72% held it for a decade, which is another way of saying about a quarter did not. And the frame is rate-sensitive. A one point rise in the return equity investors demand lifts what the price is asking of the business by about 1.6 points, which is how a stock that looks undemanding today becomes demanding without the company doing anything wrong.

The mortgage arm carries its own structural problem, and it is a competitor that does not need to earn a return. The 10-K states that "Arch MI U.S. and other private mortgage insurers compete with federal and state government agencies that sponsor their own mortgage insurance programs." Private mortgage insurance exists in the space government programs leave open, and that space is set by policy rather than by underwriting skill. Within the private cohort itself the share is moving: Radian grew revenue 14.1% over the year while MGIC went backwards at 1.2%. A high-margin business with a policy-determined addressable market is a good business until the policy changes.

What compresses, then, is not the balance sheet. Leverage is not the question at a company funded by premiums and reserves rather than by borrowing, and the capital return record argues the opposite of distress. What compresses is the earnings being capitalized. Let return on equity drift from 18.6% back toward its own long-run average and the multiple of book that the business supports drifts with it. The premium over book absorbs the difference, and there is no operating problem anywhere in that sequence. That is what makes it hard to hedge against.

Valuation

An insurer is worth the return it earns on the capital it holds, which is why the market reads it against book value rather than against an operating multiple. That reading currently runs about 1.55 times book, and at $103.38 it assumes the company sustains a return on equity of roughly 11.2%. It has lately been earning about 18.6%.

Book value per share is $67.24, and management treats its growth as the company scorecard; the 10-K defines the measure as "total common shareholders' equity available to Arch divided by the number of common shares and common share equivalents outstanding." The filing also states that segment results are judged the same way the market ends up judging the whole: "We measure segment performance for our three underwriting segments based on underwriting income or loss."

The methods largely agree here, which is rarer than it sounds. Valuing the company on its book value plus the profitability earned on it, on the earnings power of trailing profit, and on multiples drawn from comparable insurers all produce numbers above today's quote. The lone approach landing lower projects revenue forward from $19.8 billion at 9% a year and then values the bigger company at the same multiple of sales it carries now, held flat; the price sits about 21% above where that arithmetic finishes. Read as a pattern rather than as a scoreboard, that says today's level is supported by what the company already owns and already earns, and turns demanding only when the test becomes whether revenue compounds from here.

Against its own cohort the multiple of book sits in the lower half of the group. That is worth pausing on, because on the profitability the multiple is supposed to reflect the company does not sit in the lower half of anything: $4.87 billion of net income on $19.8 billion of revenue converts at a rate above every primary insurance peer in the set, RLI's 20.8% included.

What has to be true is therefore modest in absolute terms and specific in kind. The business needs to hold a return on equity near 11.2%. It currently produces 18.6% on that measure, and across its seventeen-year record the average is closer to 14.1%. History offers no guarantee at either level: among firms earning at the assumed rate, about 72% sustained it for a decade. The requirement also moves with rates, roughly 1.6 points for every point of change in what equity investors demand, which is how an insurance stock re-rates without a single underwriting decision changing.

The balance sheet frame for a company like this is capital return capacity rather than leverage, since premiums and reserves fund the assets and debt is not the constraint. On that measure the record is direct: about 86.9% of last fiscal year's earnings left as dividends and buybacks, and the share count has declined roughly 1.6% a year over the four years to March 2026. The 10-K puts 240.8 million shares in treasury at an aggregate cost of $6.4 billion as of December 31, 2025, with approximately 455.0 million shares repurchased since the program's inception.

The gap between 18.6% earned and 11.2% assumed is the entire valuation question, and by itself it is not a claim that the market is wrong. It is a statement about how much of the recent return the market expects to survive the next soft market.

Catalysts

Second quarter 2026 results are due after the close on Tuesday, July 28, 2026, with the conference call the following morning. Consensus for the quarter is $2.45 of Q2 FY2026 diluted EPS against $2.58 in the year-ago quarter, on Q2 FY2026 revenue of $4.59 billion, a 3.5% decline year over year. A softer expected top line at a company whose reinsurance segment is nearly half of revenue is consistent with the pricing turn the risk factors describe, and the line item to read first is prior period development by segment rather than the headline.

Capital structure work happened in the quarter. The company priced cash tender offers for 5.144% notes due 2043 and 5.031% notes due 2046 issued by subsidiaries, repurchasing up to $417,851,000 of aggregate principal, with early settlement completed June 18, 2026 and the offers expiring July 1, 2026. Retiring long-dated fixed-rate debt is a quieter form of the same capital discipline the buyback record shows.

Leadership at two of the three underwriting segments changed hands, with Jerome Halgan promoted to chief executive of Arch Global Reinsurance Group and Michael Schmeiser to chief executive of Arch Global Mortgage Group, both reporting to president Maamoun Rajeh. Sell-side positioning is cautious rather than negative: sixteen firms carry a consensus recommendation of hold, with eight buys, seven holds and one sell, at a consensus target of $109.82, while UBS reiterated a buy and lifted its target to $120 from $114 on July 8, 2026. The street's central number sits close to the current price, which is consistent with a market that credits the current return without extrapolating it.

Peer Cohorts (Per Segment, With Filing Citations)

Insurance (reported)

Reinsurance (reported)

Mortgage (reported)

Methodology Note

Fundamentals sourced from SEC EDGAR filings. Current price from Databento. The priced-in inversion and valuation x-ray are computed by the boothcheck engine; narrative composed by AI from the structured data.

Sources

company earnings-date announcement, July 2026 · company results-date announcement, July 2026 · Zacks Consensus Estimate, July 2026 · company tender offer announcements, June 2026 · company management announcement, 2026 · brokerage consensus compiled July 2026 · UBS research note, July 8, 2026

View the full interactive ACGL report on boothcheck