First American Financial Corporation (NYSE: FAF) — An Oligopoly at Peak Margin, Priced for Permanence

Independent Research Note Report date: 2026-07-19 | Price (2026-07-17 close): $73.03 | Market cap: ~$7.46B | 52-week range: $51.35–$73.50 Sector: Financials / Insurance — Specialty (Title Insurance & Real Estate Services) | FYE: December 31 | HQ: Santa Ana, California

The analytical body of this report carries no position: no BUY/SELL recommendation and no price target appears anywhere outside the clearly-labeled Kimi’s Take block immediately below. Valuation in the body is discussed only as embedded expectations and scenarios. This article is the author’s independent analysis and general information, not investment advice.

⚡ Kimi’s Take

Kimi’s own subjective opinion — the author’s independent view, offered as general information only; it is not investment advice. The analytical body below carries no position.

Call: HOLD — trim into strength, do not add at ~$73. Not a short. First American is a genuinely good business: the #2 underwriter in a four-family title-insurance oligopoly with real barriers (title plants, state licensing, lender relationships), a conservative AA-quality $9.4B investment book, eight straight dividend increases, and a well-aligned, pay-for-performance management team. None of that is the debate. The debate is the price of the earnings stream. FY2025 EPS of $6.00 and TTM EPS of ~$6.50 are cyclically elevated — built on a commercial-title boom (Q1 2026 commercial revenue +48%), a $39.8M reserve release, investment income of $621M running at peak portfolio yields, and a title pretax margin (12.1%) nearly triple the 2024 trough. My base-case normalized, mid-cycle EPS power is ~$5.25 (bear ~$3.50, bull ~$7.70). At $73 the market is paying ~11.2x TTM earnings — which looks tame — but ~1.35x book and ~1.0x sales, both near the top of the stock’s own ten-year range. Reverse-engineering the price, the market underwrites ~$6.00 of EPS as durable (implied ROE ~11%); I think ~$5.25 is the honest mid-cycle number. That gap is the whole thesis.

Framing: quality cyclical at a cyclical high — a one-way street, not a falling knife. The Momentum read is unambiguous: price sits within 1% of a five-year high, 21/50/200-day EMAs are stacked bullishly, trailing 3/6/12-month returns are +15.4% / +25.9% / +36.2%, and the dominant factor loadings (Value, Dividend Yield, Financials, Real Estate, with a large negative InterestRate loading and essentially zero Momentum loading) are being helped by the current regime. This is a rate-sensitive value/dividend hybrid riding a cyclical-recovery narrative — extended above all moving averages with 7.3% short interest and 23.7% idiosyncratic volatility, so the cushion is thin if the origination data rolls over. The right posture is to respect the franchise and the tape but refuse to pay peak-margin prices for peak-margin earnings: the directional zone where the risk/reward turns attractive is roughly $55–60 — mid-cycle EPS of ~$5.25 at 11–12x, or ~1.05–1.15x book — versus ~$73 today.

Conviction: medium. Bull-flip trigger: two or more quarters proving the earnings are structural, not cyclical — title pretax margin holding ≥12% with purchase (not just commercial/refi) volumes growing and Endpoint/Sequoia automation showing up as measurable cost-per-order savings. Bear-flip trigger: commercial title mean-reverts (commercial opened orders/day turning negative) or a refi fade that pushes title margin back toward single digits — confirming 2025–26 as the cycle top and re-rating the stock to >15x normalized earnings.

Tag: a fortress franchise at the top of the cycle — own the moat, not the multiple.

📈 Stock Price Action — Five-Year Event Map

Over the trailing five years FAF traced a complete real-estate/rate cycle: from a late-2021 high near $78, through a Fed-hike-driven 41% drawdown to a five-year low of $38.92 (2022-10-21), a choppy 2023–24 recovery weighed down by investment losses and a cyber incident, and a 2025–26 climb to new highs. The 52-week range is $51.35–$73.50; the stock closed 2026-07-17 at $73.03, within ~1% of both the 52-week and five-year highs. Price moves below are facts; attributed causes are interpretation, labeled as such.

The cycle narrative: events 1–3 are the rate shock and its trough — FAF’s earnings are a geared play on mortgage origination volume, and the market repriced that gearing violently in both directions. Events 4–6 are the idiosyncratic years: venture/realized investment losses ($516M in 2022), a cyber incident (Dec 2023), a $345.4M portfolio-rebalancing loss (2024), and an abrupt CEO change (Apr 2025) — each met with a single-digit drawdown, each absorbed. Events 7–8 are the recovery: with MBA forecasting 2026 originations of ~$2.2T (+6–8%) and Q1 2026 showing commercial title up 48%, the tape has fully re-rated the stock back to — and slightly above — its 2021 high, even though earnings remain ~46% below the 2021 peak. The price has round-tripped; the earnings have not.

1. Executive Summary

First American Financial Corporation is the second-largest title insurance underwriter in the United States, with a 23.1% share of an ~$18.5B industry premium pool (ALTA 2025 data) and a two-segment structure: Title Insurance and Services (93.6% of FY2025 segment revenue) and Home Warranty (~6%). Title insurance in the U.S. is a tight oligopoly — four families (FAF, FNF, Old Republic, Stewart) control roughly 75–80%+ of premiums — protected by proprietary title plants, state-by-state licensing and rate regulation, lender relationships, and statutory capital requirements. It is also a geared play on the mortgage origination cycle: FAF’s domestic title orders closed fell ~55% from 1,050,700 (2021) to ~455,500 (2023) as the Fed hiked, and title pretax margin swung from 16.3% (2021) to 4.3% (2024) and back to 12.1% (2025).

The current moment is a cyclical upswing with three distinct tailwinds stacked on top of each other. First, volume: 2025 revenue rose 22% to $7.45B and Q1 2026 revenue rose another 16.2%, driven by commercial title (+48% year-over-year in Q1 2026, to a record quarter) and a refinance rebound (refi opened orders/day +48.1%), while residential purchase — the largest, healthiest demand leg — remains weak (-4.1% opened orders/day). Second, investment income: a $9.4B, AA-quality portfolio (62% agency MBS) produced $621M of net investment income in FY2025, ~8.3% of revenue, at the highest portfolio yields in over a decade. Third, reserve luck: FY2025 included a $39.8M prior-year reserve release (~0.75% of title premiums), and management itself describes the normalized loss rate as closer to 5% than the 3.0% reported provision. Strip the one-offs and FY2025’s reported $6.00 EPS is a ~$5.60 run-rate; normalize the margin and mix for mid-cycle and the base case is ~$5.25.

The balance sheet and capital allocation are genuine strengths. Debt/capitalization is 30.7% (21.9% excluding secured financings), the portfolio is 99% investment grade, GAAP free cash flow was $762.5M in FY2025 (OCF $950.8M less capex $188.3M), the dividend has been raised eight consecutive years ($2.18 paid in FY2025; $2.20 run-rate after the May 2026 raise to $0.55/quarter), and share count is down 9.3% since YE2019. CEO pay is ~90% performance-based, tied to adjusted pretax margin and ROE. Governance is improving (board declassification and supermajority elimination proposed in the 2026 proxy).

The valuation question is whether the market is pricing the cycle correctly. At $73.03 the stock trades at ~11.2x TTM EPS (~$6.50), ~1.35x GAAP book ($53.92 at YE2025), ~0.97x TTM sales, and a ~3.0% dividend yield. The P/E reads mid-cycle; the P/B and P/S read late-cycle against the company’s own history. A justified-P/B framework at the current price implies the market underwrites a durable ROE of ~11% — i.e., FY2025-level earnings treated as largely permanent. The base-case normalized ROE is ~10% on ~$5.25 EPS. The market is pricing a soft landing in which the commercial boom, the refi lift, the reserve releases, and peak investment yields all mostly persist. That is a reasonable scenario; it is not a conservative one, and it is priced into a stock sitting at a five-year high with essentially no margin-of-safety discount. Q2 2026 earnings (release July 22, call July 23, 2026) are the next hard data point.

2. Business Overview

2.1 What the company is

First American Financial Corporation (NYSE: FAF, CIK 0001472787), founded in 1889 and spun off from the old First American Corporation in May 2010, operates two reportable segments:

1. Title Insurance and Services — issues residential and commercial title insurance policies in the U.S. and internationally (Canada, UK), and provides closing and escrow services, title plant data access, appraisals/valuations, lien release, document custodial services, warehouse lending, mortgage subservicing, tax-deferred (1031) exchanges, and banking/trust/wealth management through First American Trust, a federal savings bank.
2. Home Warranty — sells residential home warranty products (renewal-based service contracts covering appliances and systems). This segment was reported as “Specialty Insurance” through FY2022 and renamed effective FY2023 as the legacy P&C book went into runoff.

A small Corporate segment holds venture investments, the runoff P&C book, certain financing facilities, and holding-company costs — structurally loss-making (interest on holdco debt and overhead), which compresses consolidated margins below the title segment’s economics.

2.2 Segment economics through the cycle

The segment record over five years shows a single dominant engine and a small annuity attached to it:

(FACT — 10-K FY2021–FY2025 segment footnotes. FY2021 Home Warranty shown under the legacy “Specialty Insurance” label; Corporate figures include eliminations.)

Three observations matter. First, title is the cycle: its revenue fell 31% from 2021 to 2023 and its pretax income fell 82% — operating leverage works brutally in both directions because direct operations carry a large fixed-cost base (personnel, title plants, branches) against transaction-driven revenue. Second, Home Warranty is a quiet compounder: margins have improved every year from 6.6% to 19.5% as the claims ratio fell from 48.8% (2023) to 41.4% (2025); it is now a ~$86M pretax annuity, small but genuinely accretive and less cyclical than title. Third, 2024 was the earnings trough, 2025 the recovery year — and the 12.1% title margin of 2025 sits almost exactly midway between the 4.3% trough and the 16.3% peak, which is the central fact for the valuation debate below.

2.3 Revenue mix: who actually keeps the premium dollar

Consolidated revenue composition (10-K FY2025, $M):

The single most important line for understanding FAF’s economics is the one that is not revenue: premiums retained by agents — $2,374.0M in FY2025, or 80.2% of gross agent premiums (retention has been stable: 79.5% / 79.8% / 79.7% / 79.8% / 80.2% for 2021–2025; 79.3% in Q1 2026). Independent title agents originate roughly 40% of FAF’s premium volume and keep four-fifths of every dollar. The underwriter’s residual take on the agency channel is essentially a risk-bearing and service fee. This means reported revenue materially overstates FAF’s net economic share of the premium pool — a necessary caveat whenever revenue-based multiples (P/S) are used, and a structural cap on how much of the industry’s profit pool any underwriter can capture. (FACT — 10-K Note 22; INTERPRETATION — economics of the split.)

2.4 The second earnings engine: the investment portfolio

FAF is not just an underwriter; it is a levered holder of insurance float and bank deposits. At December 31, 2025, total debt and marketable equity securities were $8,944.3M fair value (debt $8,466.7M, equity $477.6M), with total investments including other items of $9,394.3M. The debt book is conservative: 95% of the portfolio is debt securities, 72% of those U.S. government-backed or AAA/Aaa, 99% investment grade, average credit quality AA. Composition by fair value: agency MBS $5,260.8M (62.1% of debt securities), U.S. corporates $1,034.8M, municipals $960.1M, foreign corporates $471.2M, agency bonds $261.0M, Treasuries $243.6M, foreign governments $235.2M. Carried alongside is a $273.5M non-marketable/venture portfolio (cumulative unrealized gains $280.1M against $360.8M of losses/impairments — a net value destroyer to date and a source of earnings noise, most infamously the Offerpad stake).

Net investment income of $621.0M in FY2025 (vs. $561.0M in 2024 and $570.0M in 2023) is roughly 8–9% of revenue and a much larger share of pretax profit. The arithmetic is powerful: on ~$8.5B of fixed-income assets, a 100bp change in portfolio yield is worth ~$85M pretax, ~$0.65/share. The flip side is rate sensitivity: a hypothetical instantaneous +100bp parallel shift would reduce the debt portfolio’s fair value by $418.3M (4.9%, implying ~4.9-year modified duration); +200bp costs $834.1M (9.8%). That sensitivity runs through AOCI — $(496.4)M at YE2024, improving to $(256.7)M at YE2025 as rates fell, then re-widening to $(322.5)M at March 31, 2026. Management’s own rule of thumb on the Q1 2026 call: each 25bp Fed cut costs ~$15M of annualized investment income — offset to date by deposit growth at First American Trust (average deposits $6.8B, +19% YoY, 29% from non-captive sources including $1.4B ServiceMac and $300M 1031 exchange deposits). (FACT — 10-K Note 3, Item 7A; Q1 2026 10-Q and call.)

Verdict (Business Overview): FAF is a high-cycle, mid-ROE oligopoly underwriter with a second, rate-driven earnings engine in its investment book, plus a small home-warranty annuity. Through-cycle ROE is likely low-double-digits, not the high-teens of a wide-moat franchise; 2025’s 16.3% ROE is a recovery-year figure, not a run rate. (INTERPRETATION — argued from the segment margin cycle and the normalized EPS work in the valuation section.)

3. Industry Dynamics

3.1 Structure: a regulated oligopoly taxing real-estate transactions

The U.S. title insurance industry wrote ~$18.5B of premiums in 2025, up 13.8% from $16.2B in 2024, and paid $667M in claims (~3.6% of premiums — an extraordinarily low loss ratio by insurance standards, reflecting the retrospective nature of the product: title claims arise from pre-existing defects in records, not from future events). Fitch reports sector operating margin improved to 11.4% in 2025 from 10.3% in 2024. (FACT — ALTA via HousingWire, May 2026; Fitch via Beinsure, March 2026.)

Market share by individual underwriter (ALTA 2025 full-year): First American Title 23.1% (#2), Fidelity National Title 14.5%, Old Republic Title 14.0%, Chicago Title 13.1%, Stewart Title 10.9%. Because FNF owns Fidelity National Title, Chicago Title, Commonwealth and Alamo, the FNF family is the largest combined player at roughly 27–30%; the Big Four families (FAF, FNF, Old Republic, Stewart) control roughly 75–80%+ of premiums, and the top seven approach 85%. Concentration has been stable for years — new entrants cannot replicate a 50-state title plant network, agency relationships, ratings-agency capital, and lender/GSE acceptance. (FACT — ALTA data; note family roll-ups require manual aggregation and Stewart’s ownership reporting remains unresolved.)

3.2 The demand driver: the mortgage origination cycle

Title premium volume is a transaction tax on real estate: every purchase and every refinance generates a policy. Demand therefore tracks mortgage originations with a lag, and 2022–2024 was a historic trough — the 30-year fixed rate went from ~3% (2021) to ~6–7%, refinance volume collapsed, and FAF’s domestic orders closed fell from 1,050,700 (2021) to ~455,500 (2023).

The forward data points, as of this report:

• MBA baseline (Oct 2025): 2026 total single-family originations $2.2T, +8% vs. $2.0T in 2025; purchase $1.46T (+7.7%); refinance $737B (+9.2%); 30-year rates seen at 6.0–6.5% through 2026–27.
• MBA April 2026 update: 2026 total trimmed to $2.19T (+6.7%), with refis revised up to +11% ($769B) and purchase growth trimmed.
• iEmergent Q2 2026: $2.235T total first-lien originations (+15% YoY); purchase $1.40T (+4%); refinance $834B (+38%).
• Home sales: NAR +4.0% to 4.22M existing-home sales; MBA +3.1% total home sales to 4.90M.

(FACT — MBA, iEmergent, NAR forecasts; INTERPRETATION follows.) Two readings of the same data: the recovery is real but modest and rate-capped — 2026 purchase volumes remain far below 2020–21, and every forecast is hostage to the 6.0–6.5% mortgage-rate assumption. Through the Marathon capital-cycle lens, the industry has been through a severe capacity-shrinking downturn (originators exited, headcount was cut), so the first leg of recovery brings outsized margin expansion to survivors — which is exactly what FAF’s 2025–26 results show. But the cycle is early-to-mid, not secular: there is no structural force making 2026’s volumes the new floor rather than a waypoint.

3.3 Commercial title: the current cycle’s differentiator

The standout demand leg of 2025–26 is commercial. MBA sees commercial real estate finance volumes +24% in 2026, and FAF’s Q1 2026 commercial revenue grew 48% to a record $271M quarter (closed orders +9%, ARPO +36%). Commercial title is less rate-sensitive, carries materially higher fees per file, and is lumpier. It is also precisely the kind of cyclical strength that flatters consolidated margins near a peak: if CRE financing slows, ARPO and title margin compress quickly because direct-operations costs are sticky. (FACT — Q1 2026 call; INTERPRETATION — the margin-quality implication.)

3.4 Regulation: state-centric, double-edged

Rate-setting varies by state — from promulgated rates, to filed-and-approved, to a few states with no filing requirement. State regulators license insurers, approve policy forms, set reserve and capital requirements, and can compel price reductions; FAF’s own risk factors warn that regulatory changes “may compel the Company to reduce its prices” or restrict increases and acquisitions. Federally, RESPA Section 8 (12 CFR §1024.14) prohibits kickbacks and unearned referral fees in settlement services, enforced by the CFPB — a perennial wild card for the referral relationships (brokers, lenders, homebuilders) through which title business actually flows. State-level RESPA-adjacent legislation continues to surface (e.g., North Carolina S 577, “Title Insurance Kickbacks Clarification,” filed March 2025). ALTA itself notes title insurance costs have declined 7.8% nationally since 2004 — evidence that regulation and scale competition, not pricing power, govern the industry’s pricing. (FACT — 10-K Item 1/1A; CFPB; ALTA.) A check for major 2025–26 class-action “excess profit” litigation in New York and California came up empty in this review and remains an open item (see Open Questions).

The regulatory frame is genuinely double-edged: it caps upside pricing but also prevents price wars, deters entry, and protects incumbent returns. The industry’s ~3% loss ratios and stable shares are the fingerprints of a business where competition happens on service, splits, and technology rather than headline price.

3.5 Technology: disruption has failed so far; incumbents are absorbing it

The digital-disruption threat has so far validated the oligopoly rather than breaking it. Radian launched titlegenius (digital title/closing) in 2021 — and announced in September 2025 that it would sell its title and real-estate services business and pivot to specialty insurance. The incumbents are instead absorbing the technology themselves: FAF’s AgentNet Assist (AI document analysis), Endpoint (digital escrow — launched December 2025, industry’s first AI-powered escrow closing), the SEQUOIA title production system (running live refinance orders, targeting “instant title” on purchases), and First American Data & Analytics (Esri/ArcGIS partnership announced July 2026). Endpoint was at 153 orders opened / 47 closed with 30% automation as of the February 2026 call, against a long-term 80–90% automation target; management has disclosed a ~100bp title-margin drag from running legacy and new platforms in parallel. (FACT — Radian/Business Wire; The Title Report; Q4 2025/Q1 2026 calls.) The likely trajectory is that technology lowers incumbent unit costs rather than enabling entrants — margin-accretive if executed, but with benefits back-loaded over ~2 years and currently immaterial in volume. (INTERPRETATION.)

Structural verdict: a decent, not great, industry — genuine Greenwald barriers (scale, title plants, licensing, agent networks) protecting a low-loss, transaction-driven profit pool, with state rate regulation capping pricing power, an agency channel that keeps ~80% of agent premiums out of underwriter hands, and inescapable beta to the origination cycle. A comparison with Assurant (AIZ) is instructive: FAF’s title core resembles AIZ’s lender-placed niche (regulated, scale-driven, captive distribution) more than its moat-less warranty businesses — good-but-not-great, with the industry’s best economics accruing to scale underwriters that own data and direct distribution. 2025–26 is an early-to-mid-cycle recovery with favorable operating leverage; nothing in the structure makes the current margin level permanent.

4. Competitive Position

4.1 The moat mechanisms, named and weighed

FAF’s competitive advantages are real and identifiable, and each one traces to a financial outcome that would deteriorate without it:

1. Title plants — proprietary, geographically indexed compilations of public records, maps, documents and prior policies, built over decades (FAF traces to 1889). They lower search/examination cost per order, improve underwriting accuracy (hence the ~3% loss provision versus what an under-capitaled entrant would bear), and cannot be assembled quickly by a challenger. Digital public-records access and AI search are eroding the exclusivity of this advantage at the margin, but the incumbents — FAF included — are the ones deploying that AI on top of their own plants (see the technology discussion in Industry Dynamics).
2. State licensing and rate-filing history — 50-state licensing, statutory capital, approved policy forms, and standing with state regulators. Friction for entrants; also a ceiling on pricing (see Regulation above).
3. Lender/GSE relationships and ratings — national lenders and the GSEs require financial-strength ratings and operational scale from title counterparties; FAF’s AA-quality investment book and national footprint keep it on every panel. The June 2026 Fannie/Freddie verification of ACI Sky Workbench for UAD 3.6 is a small but concrete example of incumbency compounding into the appraisal-adjacent process.
4. Scale economies in the agency channel — FAF can offer independent agents better technology, underwriting support, and data than smaller underwriters, which is how it defends the ~40% of its premium volume that agents originate.
5. Distribution breadth — both a direct channel (company-owned offices doing search/exam/closing, capturing full margin) and an agency channel (asset-light, ceding ~80% of premium). Old Republic skews agency; FNF/FAF/Stewart skew direct; the blended model gives FAF reach without full fixed-cost exposure.

4.2 Pressure tests

Pricing power — weak by design. Competition “is based primarily on service, quality, price, relationships and the ease of access and use of products” (10-K). Rates are state-regulated; ALTA says title costs have fallen 7.8% since 2004; agents negotiate splits and routinely represent multiple underwriters. There is no evidence FAF can raise price without losing volume. The moat protects returns on existing share, not the ability to expand price.

Agency channel economics — a structural leak. With agents retaining 80.2% of agent premiums (2025), the underwriter is a wholesale risk-bearer in 40% of its book. This caps consolidated margins, means revenue overstates economic share, and creates a wholesale price floor: underwriters compete for agents on splits and service, preventing true oligopoly rent extraction even with four-family concentration.

Cyclicality — the binding constraint. Title pretax margin: 16.3% (2021) → 4.3% (2024) → 12.1% (2025). ROE: 39.2% → 3.6% → 16.3%. Compare Old Republic — the most conservative, agency-heavy peer — whose ROE held a much tighter band (30.0% / 12.3% / 10.4% / 14.6% / 16.2% for 2021–2025) and whose trough margins stayed in double digits. FAF’s direct-operations leverage gives it higher peaks and deeper troughs. A moat that does not protect earnings through the cycle is a narrower moat than the market’s current pricing implies. (FACT — ROIC.ai peer data; INTERPRETATION — the comparison.)

Demand-side captivity — limited. End consumers do not choose title insurers; lenders, agents, and brokers do. Switching costs at the point of decision are low. What captivity exists is distribution captivity (agent relationships, lender panels, escrow integration) — real, but contestable on service and splits.

Share stability — supportive but unresolved longitudinally. Point-in-time ALTA data shows FAF a stable #2 (22.2% in 2024, 23.1% in 2025). A 10-year share trend for FAF vs. FNF vs. ORI vs. Stewart was not compiled for this report and remains an open question (see Open Questions); the durability claim currently rests on the structural barriers plus two years of stable share, not on demonstrated longitudinal share defense.

4.3 Verdict

Durable but narrow. In Greenwald terms this is a scale/captive oligopoly with strong supply-side cost advantages and regulatory barriers but limited demand-side captivity. The moat is genuine — it explains stable share, ~3% loss ratios, and adequate mid-cycle returns — but it is narrow: pricing is regulated, agents capture most of the premium they originate, the consumer is not captive, and earnings remain hostage to the origination cycle. Management’s AI/automation claims (Endpoint, Sequoia, AgentNet Assist) are efficiency reinforcements, not standalone moats, and are unproven at scale (30% automation on a trivial order base today). The correct mental model: FAF’s moat protects its seat at the table and its cost position; it does not protect its margin level from the cycle.

5. Growth History and Forward Opportunities

5.1 The historical record: cyclical, not secular

Five-year consolidated record (FACT — 10-K FY2021–FY2025; Q1 2026 10-Q):

Revenue in 2025 was still 19% below 2021; EPS was 46% below 2021. There is no secular growth story in the top line — there is a deep cycle with a modest structural uplift from mix (average revenue per order up from $2,718 in 2021 to $3,961 in 2025, driven by purchase/commercial mix and home-price appreciation, not unit pricing) and from investment income.

5.2 Order trends: the operating heartbeat

Domestic direct title orders closed: 1,050,700 (2021) → 695,900 (2022) → 455,500 (2023) → 468,800 (2024) → 531,900 (2025). Average revenue per order closed: $2,718 → $3,498 → $3,651 → $3,914 → $3,961. Q1 2026: 119,900 orders closed (+8.8% YoY) at $4,229 ARPO (+12.9% YoY). (FACT — 10-K MD&A; 10-Q.)

The ARPO story deserves skepticism. Rising revenue per order is mix, not pricing: the 2021 denominator was swollen with low-fee refinance orders; as refi died, the surviving mix (purchase, commercial) mechanically lifted the average. Q1 2026’s +12.9% ARPO was driven by commercial mix. Treat ARPO as a cyclical composition effect that will mean-revert if/when low-fee refi volume returns in size — not as evidence of pricing power. (INTERPRETATION.)

5.3 Forward growth lanes, in order of materiality

Lane 1 — Cyclical origination recovery (the dominant driver). MBA/iEmergent forecasts (see Industry Dynamics) imply mid-to-high single-digit origination growth in 2026 with refi growing faster off a tiny base. Q1 2026 already shows the shape: refinance opened orders/day +48.1% (but refi is only ~8% of direct revenue), commercial +3.5% opened orders/day with +48% revenue, purchase -4.1%. A genuine volume recovery in purchase — the healthiest leg — has not yet arrived; that is upside if rates break below 6%, and the biggest gap in the recovery if they do not.

Lane 2 — Commercial title. Record Q1 2026 ($271M revenue, +48%); management expects a record 2026. Commercial is high-fee and less rate-sensitive, but lumpier and tied to CRE transaction financing — a strength today and a mean-reversion candidate tomorrow. The exact residential/commercial premium split is not disclosed (an open question — see Open Questions), so sizing this lane’s durability is judgmental.

Lane 3 — Non-title fee businesses. Mortgage subservicing (called out as a driver of FY2025’s +12.0% information-and-other revenue growth), tax-deferred 1031 exchanges, warehouse lending, data & analytics (Esri partnership, ACI Sky UAD 3.6 verification), and First American Trust banking/wealth ($5.6B fiduciary/custody assets administered, $6.2B deposits at YE2025). These monetize the same transaction infrastructure and captive float; strategically sensible, small in a $7.5B revenue base, and not separately disclosed by revenue/margin. Home Warranty, at $442.9M revenue and a 19.5% pretax margin with a falling claims ratio (48.8% → 41.4% over three years), is the best of the ancillaries — a genuine, improving annuity.

Lane 4 — Automation (Endpoint, Sequoia, AgentNet Assist). Early and unproven at scale: Endpoint at 153 opened orders and 30% automation as of February 2026, with an 80–90% long-term automation target and management flagging a ~100bp margin drag during the parallel-run period. If automation delivers, it converts fixed cost to variable and structurally raises through-cycle margin — the single most credible structural margin story in the company. It is also the least evidenced: no cost-per-order, adoption, or ROI KPIs are disclosed. Directionally correct, currently immaterial, worth tracking on every call. (FACT — Q4 2025/Q1 2026 calls; INTERPRETATION — the margin implication.)

Verdict (Growth): near-term growth is mostly the cycle, not the company. Structural levers (commercial share, subservicing, data, warranty, automation) are real but individually small; together they plausibly add a point or two of through-cycle growth on top of GDP-ish housing turnover. Nothing here justifies extrapolating 2025–26 growth rates.

6. Financial Quality

6.1 The earnings cycle, restated honestly

Consolidated results swing with the origination cycle, amplified by operating leverage and punctuated by investment-portfolio noise:

TTM (through Q1 2026): revenue ~$7.71B, net income ~$673M, diluted EPS ~$6.50. (FACT — 10-K/10-Q.)

6.2 Underwriting quality: loss ratios, reserves, and the release

Title claims are long-tailed but small: provisions for policy losses and other claims ran ~3.3% (2021), 3.4% (2022), 3.25% (2023), and 3.0% (2024, 2025) of title premiums and escrow fees. The FY2025 3.0% decomposes into a 3.75% current-policy-year ultimate loss rate plus a 0.75% prior-year reserve release of $39.8M. Total loss reserves stood at $1,170M at YE2025 (title claims $1,151M), essentially flat against payments ($77.8M provision vs. $77.7M payments in Q1 2026). An approximate title combined ratio for FY2025 is ~88% (operating expenses ~85.6% of segment revenue plus the ~2.3% loss provision) — consistent with an oligopoly earning adequate, not extraordinary, underwriting margins.

Two quality flags. First, management itself normalizes higher: on the Q4 2025 call the company described the normalized loss rate as closer to 5% and said releases are expected to slow or stop as the current-year rate sits near/below normalized levels. The 3.0% reported provision is therefore a favorable, possibly trough, provisioning level — the reserve release was ~$0.31/share of after-tax 2025 EPS that will not reliably repeat. Second, a release flattered Q4 2025 specifically: a one-time $13M Canada reserve release plus an $11M net prior-year decrease in that quarter alone. Home warranty claims ratios, by contrast, have genuinely improved (48.8% → 46.4% → 41.4%, 2023–2025). (FACT — 10-K MD&A, Q4 2025 call; INTERPRETATION — the normalization.)

6.3 The investment book: quality high, rate risk real

The portfolio (see Business Overview) is $9.4B, 95% debt securities, 72% government-backed/AAA, 99% investment grade, AA average quality, ~4.9-year implied duration. Credit quality is not the issue; duration and accounting volatility are. AOCI swung from $(496.4)M (YE2024) to $(256.7)M (YE2025) to $(322.5)M (Q1 2026) — reported book equity moves materially with rates even when operating earnings do not. The 2024 strategic rebalancing crystallized the problem: FAF sold $2.8B of securities in unrealized-loss positions, realizing a $345.4M loss in the Title segment (pre-announced as ~$342M in the September 2024 8-K), against consolidated FY2024 net investment losses of $401.6M. The stated purpose — repositioning into higher yields — is now producing the $621M investment-income line that flatters current earnings. Credit the trade; but recognize that 2024’s pain and 2025’s gain are the same duration bet wearing different rate regimes. (FACT — 10-K Note 3/20, Item 7A, 8-K 2024-09-09.)

6.4 Quality of earnings: the adjustments that matter

Severance has run $5.0M / $34.7M / $12.6M / $8.3M / $11.7M (2021–2025) — treat as partly recurring restructuring. Effective tax rate: 24.3% (2025), 19.8% (2024, flattered by credits/valuation-allowance reversals), 22.9% (Q1 2026). (FACT — 10-K MD&A, 8-Ks.)

The QoE verdict: 2024 GAAP earnings ($1.26 EPS) were artificially depressed; 2025 GAAP earnings ($6.00) are artificially flattered. Stripping 2025’s reserve release (~$0.31) and insurance recovery (~$0.11) gives a ~$5.60 run-rate before any margin/mix normalization; a fuller mid-cycle normalization lands at ~$5.25 (see Valuation). Free cash flow quality is genuinely good — GAAP FCF of $762.5M in FY2025 (OCF $950.8M less capex $188.3M) versus net income of $622M, a >120% conversion, aided by float dynamics and minimal reinvestment needs. (A non-GAAP ~$881M FCF figure from data providers could not be reconciled to the primary statements; this report uses the GAAP number throughout.) Note also that Q1 2026’s balance-sheet expansion (assets to $17.9B, cash to $2.4B) is bank-deposit-driven (deposits $7.1B vs. $5.3B at YE2025) — non-operating growth that must be netted in any enterprise-value work. (FACT — 10-K/10-Q.)

6.5 Balance sheet

YE2025: total assets $16.2B; cash $1.39B; investments $9.39B; goodwill $1.82B (Mother Lode-era); deposits liability $5.29B; notes and contracts payable $1,545.4M ($650M 2.40% due 2031, $450M 4.00% due 2030, $450M 5.45% due 2034); secured financings $906.5M; equity $5.52B; debt/capitalization 30.7% headline, 21.9% excluding secured financings; undrawn $900M revolver. Regulatory dividend capacity from insurance subsidiaries: $382.0M maximum in 2026 plus $113.6M of permissible loans/advances — comfortably covering the ~$224M annualized dividend. GAAP book value per share: $53.92 at YE2025, $53.77 at March 31, 2026 (equity attributable to FAF ÷ basic shares). Statutory surplus and RBC ratios by subsidiary are not disclosed — an analytical gap, not necessarily a red flag. (FACT — 10-K Note 9/14, Item 7; 10-Q.)

Verdict (Financial Quality): clean underwriting accounting, conservative credit book, strong cash conversion — but earnings are cyclically positioned and carry identifiable one-off flattery in 2025 (reserve release, insurance recovery, peak investment yields) just as 2024 carried identifiable one-off damage. Reported numbers at this point in the cycle overstate run-rate earning power by roughly 10–25%.

7. Capital Allocation

7.1 The record

Dividends — the backbone. Declared dividends per share have risen every year since at least FY2019: $1.66 (2019), $1.77 (2020), $1.92 (2021), $2.03 (2022), $2.08 (2023), $2.12 (2024), and $2.18 paid in FY2025 ($223.0M total). The May 2026 declaration of $0.55/quarter sets a $2.20 run-rate — eight consecutive annual increases, maintained straight through the 2023–24 earnings trough when the payout ratio briefly exceeded 100% of GAAP earnings. That is a deliberate signal: management runs the dividend off through-cycle earning power, not the current year. Cash dividends paid: $188.0M (2019) rising to $223.0M (2025). (FACT — 10-K cash-flow statements; Business Wire 2026-05-12. Note: an interim $2.16 FY2025 figure from a data feed was superseded by the primary-source $2.18 paid.)

Buybacks — opportunistic, recently modest. Gross repurchases: $139.0M (2020), $99.2M (2021, ~1.7M shares), $440.7M (2022), $72.7M (2023), $68.5M (2024), $122.3M (2025, ~2.1M shares at ~$58.54 average), and $33.5M in Q1 2026 (~0.6M shares at ~$55.83). Share count: 112.5M (YE2019) → 102.0M (YE2025) → 102.1M (Mar 2026), a net 9.3% reduction since 2019. Remaining authorization: $266.5M under the July 2025 $300M program. The 2022 burst — $441M near the top of the rate cycle, just before rate-driven investment losses widened — was poorly timed in hindsight; the 2023–25 pattern of buying weakness at ~$55–58 was well timed. Buybacks are opportunistic, not mechanical. (FACT — 10-K/10-Q; INTERPRETATION — the timing assessment.)

M&A — dormant after Mother Lode. The one material deal of the period: Mother Lode Holding Co., acquired May 2022 for ~$300M cash (preliminary goodwill $203M, intangibles $37M). Since then: zero acquisition cash outflow in 2024 and 2025; FY2025 title-segment acquisitions of just $10.5M. Goodwill stands at $1.82B (YE2025) against $5.5B of equity — meaningful but not alarming. Stated capital priorities (Q1 2026 call): (1) reinvest in the business/technology, (2) disciplined M&A, (3) return excess capital. Debt-to-capital target ~20% over the cycle (22.5% at Q3 2025 on the company’s adjusted basis; 21.9% excluding secured financings at YE2025).

Balance-sheet allocation. The dominant “allocation” is the investment portfolio itself: $9.4B, 99% investment grade, AA average — conservative to a fault, with the 2024 rebalancing the one active portfolio decision of the period (cost: $345.4M of realized losses; benefit: today’s ~$621M income line). Cash returns in 2025 totaled $345M (dividends $223M + buybacks $122M), ~36% of the $950.8M operating cash flow — meaningful retained capital even in a recovery year. (FACT — 10-K; Q1 2026 call.)

7.2 Incentives and governance

Executive pay is tightly and sensibly aligned. The annual incentive plan is 50% adjusted pretax margin / 50% adjusted ROE — cycle-aware metrics, not revenue growth. The 2025 payout was 142% of target (adjusted pretax margin 10.6% vs. 9.7% target → 138%; adjusted ROE 10.6% vs. 9.5% target → 146%) on a 0–200% curve. Long-term incentives are 50% performance RSUs on 3-year relative TSR vs. the S&P MidCap 400, 50% time-vested RSUs. CEO Seaton’s 2025 compensation: $8.28M total, ~90% performance-based on an actual basis. Ownership guidelines (CEO 6× salary), anti-hedging/pledging policies, clawbacks, and >93% say-on-pay support (2024) complete a clean governance picture. The 2026 proxy proposes eliminating supermajority voting and declassifying the board (annual elections phased in by 2029) — unambiguously shareholder-positive if approved. (FACT — DEF 14A 2026.)

One footnote: 2025 included leadership-transition grants to Seaton/Wajner/Cornehl, and DeGiorgio’s $9.62M summary-comp figure is inflated by separation-related acceleration — disclosed, modest in context, but worth noting as a one-year distortion. Insider ownership is modest (~3.5% for directors and officers as a group); principal holders are Vanguard 10.3%, BlackRock 9.2%, FMR 9.0%, Boston Partners 5.3%.

7.3 Insider signal

The only genuine open-market insider purchase in 24 months: director Parker S. Kennedy bought 59,841 shares for ~$4.05M in February 2026 at $67.17–$67.92 — a long-tenured director putting new capital to work ~8% below today’s price. Against it: open-market sales by the executive chairman (Gilmore, 110,145 shares at $63.59, Feb 2025), the incoming CEO (Seaton, 50,000 shares at $64.28–65.27, Feb 2025), and smaller sales by other officers/directors at $65–71. Most officer dispositions are mechanical tax withholdings (F-codes), and 10b5-1 plan flags could not be parsed — so the discretionary-sale signal is genuinely mixed. Net: modestly positive (Kennedy’s buy is the single most informative transaction), far from overwhelming. (FACT — Form 4s; INTERPRETATION — the weighting.)

7.4 Verdict

Grade: A−. A patient capital-return compounder: eight straight dividend increases funded through the trough, a 9.3% share-count reduction since 2019, a fortress balance sheet, disciplined-to-dormant M&A, and pay tied to margin and ROE rather than empire-building. Deductions: the poorly timed 2022 buyback burst, dormant reinvestment optionality, and modest insider ownership. The dividend’s sustainability is well supported — $382M of 2026 subsidiary dividend capacity against a ~$224M annualized payout, and GAAP FCF of $762.5M covering the dividend ~3.4x.

8. Changes and Headwinds — Last Two Years

Timeline of material events, 2024-07 to 2026-07 (all FACT — 8-Ks, 10-K, proxy, Business Wire, transcripts):

• 2024-09-09 — Strategic investment-portfolio rebalancing announced via 8-K; ~$342M pretax realized loss expected in Q3 2024 (final: $345.4M Title-segment realized losses on $2.8B of proceeds; consolidated FY2024 net investment losses $401.6M).
• 2024-09-24/30 — Issued $450M 5.45% senior notes due 2034; repaid $300M 4.60% notes at maturity in November 2024.
• 2024-10-23 — Q3 2024 earnings: $308M net investment losses in the quarter from the rebalancing.
• 2024-12 — Endpoint launched; the industry’s first AI-powered escrow closing, in one office.
• 2025-02-21 — FY2024 10-K filed: EPS $1.26, the trough year, with the rebalancing loss and a $39.5M software impairment inside it.
• 2025-04-15 — CEO transition: Ken DeGiorgio terminated without cause effective April 10; CFO Mark E. Seaton named CEO; Matt F. Wajner promoted to CFO. The stock fell ~8.6% over April 1–15 on the shock, then stabilized; the transition has since proven orderly (DeGiorgio received separation benefits per his employment agreement; no 2025 bonus).
• 2025-07-23 — New $300M share repurchase authorization (July 2025 program; $266.5M remaining at Q1 2026).
• 2025-09-19 (industry) — Radian announces exit from title insurance (sale of its title/real-estate services business) — a structural validation of the oligopoly (see Industry Dynamics).
• 2025-10-23 — Q3 2025: adjusted EPS $1.70; Sequoia running live refinance orders.
• 2026-02-12 — Q4 2025: adjusted EPS $1.99; Endpoint live; one-time benefits of $28M ($13M Canada reserve release, $15M insurance recovery).
• 2026-02-13/17 — Director Parker Kennedy’s ~$4.05M open-market purchase (see Insider signal above).
• 2026-03-30 — 2026 proxy: proposals to declassify the board (phased through 2029) and eliminate supermajority voting; board shrinking from 11 to 9.
• 2026-04-22/23 — Q1 2026: revenue +16.2%, adjusted EPS $1.33 vs. ~$1.06 consensus, commercial +48%; stock +19% over the following 20 days.
• 2026-05-12 — Quarterly dividend raised to $0.55 ($2.20 run-rate).
• 2026-06-08 — ACI Sky Workbench verified by Fannie Mae/Freddie Mac for UAD 3.6 (required for UCDP submissions from Nov 2, 2026).
• 2026-07-08 — Q2 2026 earnings scheduled: release July 22 after close, call July 23, 2026.

Standing headwinds. (1) Cyber incident aftermath: the December 2023 breach (unauthorized activity disclosed 2023-12-22; investigation concluded May 2024; class action received June 2024) was never cost-quantified in filings — remediation and litigation costs are embedded, undisclosed, and the matter is a live contingency. (2) Residential purchase demand remains weak (-4.1% opened orders/day in Q1 2026): the recovery is narrow — commercial and refi-led — and a rate re-spike would remove both. (3) Platform-transition drag: running Endpoint/Sequoia alongside legacy systems costs ~100bp of title margin during the parallel period; execution risk is real even if the destination is attractive. (4) New-CEO execution: Seaton is a credible insider (former CFO) and the first year has gone well, but the strategic imprint — technology pace, M&A appetite, commercial underwriting discipline at the top of the CRE cycle — is still being written. (5) Investment-income normalization: three Fed cuts over the past year; each 25bp costs ~$15M annualized income; management expects 2026 title-segment investment income roughly flat with 2025 only because deposit growth offsets rate cuts — a treadmill, not a tailwind, from here.

9. Risk Analysis (Risk Matrix)

Aggregate assessment. The catastrophic-loss risk is low: a low-loss-ratio insurance book, a 99%-investment-grade portfolio, moderate leverage, and statutory dividend capacity make permanent impairment unlikely outside a true housing depression. The realistic downside is earnings and multiple compression in tandem — the classic cyclical trap where the P/E denominator falls as the multiple rises. Risks 1–3 and 11 are the same risk seen from different angles and constitute the dominant exposure. Risks 6–7 are self-inflicted-wound risks (portfolio and cyber) that have each cost the company nine figures in the last four years and deserve more weight than their “one-off” labels suggest.

10. Valuation Discussion (Embedded Expectations)

No price target; no recommendation. This section asks only what the current price must assume.

10.1 The multiple stack

At the 2026-07-17 close of $73.03 (basic shares ~102.1M; market cap ~$7.46B):

10.2 Own-history context: the divergence is the message

FAF’s trailing multiples against its own ~10-year history (calendar-year data): P/E 10.2x last vs. ~15.4x average (range 6.0–49.5x, the high an artifact of trough earnings); P/B 1.58x last vs. ~2.2x average (range 1.08–3.87x); P/S 0.85x last vs. ~0.91x average (range 0.45–1.22x); EV/Sales 1.02x vs. ~0.94x average. A data-provider percentile ranking of the stock’s own history puts P/E around the 63rd percentile but P/B and P/S at/near the very top of their historical distributions (mid-90s and ~100th percentile respectively; the exact P/S convention behind the percentile was not reconcilable, but the direction is corroborated by the primary-source series). (FACT — ROIC.ai annual multiples; AZI valuation_index, convention-flagged.)

Read together, this is the fingerprint of a cyclically elevated margin year: the P/E looks unremarkable because the E is inflated; the price relative to revenue and book is at the top of the historical range because revenue is being produced on temporarily wide margins (12.1% title pretax vs. 4.3% in 2024) and book has been rebuilt by recovering earnings and AOCI. The market is not paying a peak P/E — it is paying a peak price per dollar of revenue, which only makes sense if current margins persist.

10.3 Comps

(FACT — ROIC.ai at 2026-07-17; yields approximate.) FAF is not an outlier cross-sectionally: it trades at a P/B discount to every title peer and to AIZ, at a P/E in line with ORI (the cleanest through-cycle comp, agency-heavy and less operationally levered) and below STC/ITIC. FNF’s consolidated multiples are distorted by its annuity business and are not comparable. The relative-valuation verdict is neutral — the market is not singling FAF out for a premium. The warning comes from the company’s own history (P/B and P/S), not from relative valuation. (INTERPRETATION.)

10.4 Normalized earnings power

FY2025 reported EPS $6.00 less the reserve release (~$0.31 after tax) and the corporate insurance recovery (~$0.11) is a ~$5.60 clean run-rate. Normalizing further for mid-cycle volumes, margins, and mix (assumption-based scenario framework; share count ~103M, tax 24%):

Base-case mechanics: title revenue ~$6.9B, title pretax ~$742M (10.75%), home warranty pretax ~$76.5M, corporate loss ~($110M) → consolidated pretax ~$708M, net ~$538M, EPS ~$5.25 — a through-cycle ROE of ~10% on the ~$53 book. The bear case is a modest pullback; a 2024-style hard trough implies EPS toward $2.50–3.00. The bull case requires the MBA recovery to persist, commercial to stay strong, and margin to approach the 2021 peak. (ASSUMPTION — all scenario inputs; commercial mix undisclosed, so ARPO/margin assumptions are judgmental.)

10.5 What the price embeds

Using a justified-P/B framework (P/B = (ROE − g)/(r − g)) with cost of equity r ≈ 9% and long-run growth g ≈ 3%: at ~1.35–1.37x book, the market underwrites a durable ROE of ~11% — i.e., ~$5.90–6.00 of EPS at the current book base, essentially FY2025’s reported level treated as permanent. That is above the base-case normalized ~$5.25 (~10% ROE), far below the 16.3% printed in 2025, and nowhere near the 39.2% of 2021. In plain terms: the market is pricing a soft landing — the title cycle stays healthy, the commercial boom mostly holds, investment income stays near $620M, and the reserve/mix tailwinds mostly persist — with only mild normalization. It is not pricing a deep trough, and it is not pricing a 2021-style peak.

The asymmetry this creates, stated as scenarios rather than targets: if normalized EPS is ~$5.25 and the market applies a mid-cycle multiple to it, the stock carries meaningful downside from $73 (roughly a fifth to a third, depending on the multiple one assigns a ~10% ROE franchise); if the bull case proves out ($7.50+ EPS on a sustained commercial/purchase recovery), the stock is ~10x earnings with room to re-rate, and today’s price looks cheap in hindsight; if the bear case merely normalizes 2025’s one-offs and mix, EPS of ~$3.50 puts the stock above 20x normalized earnings — the “cheap on peak earnings” trap. The load-bearing assumption for the long side is that title pretax margin holds at/above 12% while revenue grows; the load-bearing assumption for the short side is that a top-of-history P/S multiple survives a return to mid-cycle margins.

11. Variant Perception

Consensus / what the market believes. The tape, the factor profile, and the April 2026 reaction (+19% on a $1.33-vs-$1.06 adjusted EPS beat) tell a consistent story: the market has re-rated FAF as a cyclical-recovery beneficiary — commercial title strength, refi optionality if rates fall, a ~3% growing dividend, and “only 11x earnings.” Sell-side framing (e.g., Zacks’ post-Q1 coverage) emphasizes the beat and commercial strength. Short interest is moderate at 7.3% of float (4.99 days to cover) — skeptical money exists but is not crowded. Factor positioning confirms the identity: positive Value (+0.20), Dividend Yield (+0.20), Real Estate, Home Construction, Banks, Insurance loadings; negative InterestRate (-0.52) and Growth (-0.20); Momentum ≈ 0 — the market owns FAF as a rate-sensitive value/dividend hybrid, and the current regime (Value, Low Vol, Dividend Yield, Financials all working) is helping it. Notably, the rally is not a momentum-factor trade and R² is only ~50% — half the stock’s variance is idiosyncratic. (FACT — FactorsToday, Benzinga; INTERPRETATION — the synthesis.)

Bull case. The 2026 recovery broadens: MBA’s +6–8% origination growth materializes, purchase demand joins commercial and refi, commercial delivers a record year, investment income holds ~$620M+ on deposit growth, Endpoint/Sequoia start converting fixed costs to variable, and title margin pushes toward 13–14%. EPS approaches $7.50+; at $73 the stock is ~10x that with a 3% growing dividend — cheap. Kennedy’s ~$4M February buy at ~$67–68 and eight straight dividend raises are the insider/management corroboration. The five-year high is an all-time-high breakout, not a ceiling.

Bear case. 2025–26 is the cycle top being mistaken for a new plateau. The recovery is narrow (purchase orders still negative); commercial title is a lumpy, late-cycle CRE phenomenon that mean-reverts; the reserve release ($39.8M) and insurance recovery ($14.8M) do not repeat; management’s own ~5% normalized loss rate reasserts; Fed cuts bleed investment income ~$15M per 25bp; and a top-of-history P/B and P/S compress toward mid-range even on flat earnings. Normalized EPS is $3.50–5.25, not $6.50; the stock re-rates to 14–20x normalized earnings. Every prior cyclical peak in this stock (2021) was followed by a 40–50% drawdown; the factor model shows no momentum-factor sponsorship to arrest a reversal.

Key assumptions on which the variant rests. (1) Commercial title’s durability — unverifiable from disclosure (no segment split); (2) whether purchase demand recovers under 6.0–6.5% mortgage rates; (3) the pace of Fed cuts versus deposit-growth offset on investment income; (4) whether automation delivers measurable cost-per-order savings within ~2 years; (5) the true mid-cycle title margin — 10.75% (the base case) versus the ~12% the price requires.

Where this view differs from the tape (interpretation): the disagreement is not about the franchise or the cycle’s direction — it is about how much of 2025’s margin is permanent. The market’s embedded ~$6.00 durable EPS treats roughly three-quarters of the recovery-year margin expansion as structural; a close reading of the segment history, the reserve math, and the mix drivers says about half is.

12. Fact vs. Interpretation

13. Open Questions

1. Residential vs. commercial title premium split by year — not disclosed; sizes the durability of the current margin engine. IR request or quarterly-file inference needed.
2. Longitudinal market-share trend (10+ years) for FAF vs. FNF vs. ORI vs. Stewart — point-in-time ALTA shares only; the durability claim rests on two years of data plus structure.
3. Endpoint/Sequoia unit economics — cost per order, automation-rate trajectory, legacy decommissioning schedule; the ~100bp parallel-run drag’s end date.
4. Portfolio duration and reinvestment runoff by quarter — the 10-K gives only a year-end sensitivity snapshot; investment-income scenarios are rule-of-thumb.
5. Statutory surplus / RBC by insurance subsidiary — only aggregate 2026 dividend capacity ($382M + $113.6M loans) is disclosed.
6. NY/CA “excess profit” / rate class-action litigation — flagged by the industry sweep as un-researched; status and exposure unknown.
7. Cyber-incident total cost and litigation status — never quantified in filings; class action received June 2024.
8. Outcome of the 2026-05-12 annual-meeting votes on board declassification and supermajority elimination (8-K Item 5.07 not extracted).
9. Mortgage subservicing, 1031 exchange, and trust/wealth revenue and margins — aggregated inside “information and other” and Corporate.
10. Trigger for the 2026-03-19 −9% single-day drop — no company-specific catalyst identified.
11. FNF/ORI/STC title combined ratios — peer underwriting margin-of-safety not compiled.
12. DeGiorgio separation terms — full acceleration/severance values not reconciled to the employment agreement.

14. What Must Be True

For the bull case (EPS → $7.50+, today’s price cheap):

• 2026 originations grow ≥ MBA’s +6–8% and purchase opened orders turn positive by H2 2026. Falsification test: two consecutive quarters of negative purchase opened-orders/day with total closed orders below ~525k annualized.
• Commercial title holds at/near record levels through 2027. Falsification: commercial opened orders/day turning negative YoY or commercial revenue growth decelerating below ~10%.
• Title pretax margin sustains ≥12% (adjusted) while revenue grows. Falsification: adjusted title margin printing below ~10.5% in any 2026 quarter with revenue still growing.
• Investment income stays ≥$600M despite Fed cuts. Falsification: two quarters of YoY declines larger than the ~$15M-per-25bp rule of thumb implies after deposit offsets.
• Endpoint/Sequoia show measurable cost-per-order or margin contribution. Falsification: automation KPIs still anecdotal (<5% of order volume) by mid-2027.

For the bear case (EPS → $3.50, multiple re-rates higher):

• The 2025 margin was mix-and-release flattered, not structural. Falsification of the bear case: title margin holds ≥12% through a quarter in which reserve releases are zero and commercial growth is flat.
• Mortgage rates stay ≥6.5% or re-spike, freezing purchase demand. Falsification: sustained sub-6% rates with purchase orders inflecting positive.
• Commercial mean-reverts with CRE financing stress. Falsification: CRE origination volumes (MBA CREF) grow double digits again in 2027.
• Investment income erodes with cuts. Falsification: title-segment investment income grows YoY through a 100bp cumulative cutting cycle.
• The loss rate normalizes toward management’s ~5%. Falsification: current-policy-year ultimate holds ≤3.75% for two more policy years.

The hinge variable both sides share: the durability of the 12.1% title pretax margin. Every scenario — multiple, normalized EPS, embedded expectations — reduces to whether that number is a waypoint to 14%+ or a peak above a 9–10% mid-cycle.

15. Source Appendix

The full source appendix — SEC filings (10-K FY2021–FY2025, the 10-Q series, material-event 8-Ks, DEF 14A 2026, Forms 3/4/5, with EDGAR links and accession numbers), transcripts of the Q3 2025, Q4 2025, and Q1 2026 earnings calls, industry sources (ALTA, MBA, iEmergent, Fitch), and data providers (ROIC.ai, AZI Trading, FactorsToday, Benzinga) — follows as Appendix B, and every factual claim in this report traces to it. All sources cited are public.

Facts, interpretations, assumptions, and open questions are labeled throughout. The analytical sections take no position and carry no price target; the single opinion block is Kimi’s Take at the front of this report. This article is the author’s independent analysis, published for general information only — it is not investment advice.

APPENDIX A — Standard Diligence Questionnaire

First American Financial Corporation (NYSE: FAF) Date: 2026-07-19 | Companion to the research report above

This appendix answers a standard diligence questionnaire in full. Where a standard question does not map to a title-insurance / specialty-insurance business model, the correct sector analog is used and the mismatch is noted. Every figure traces to the public primary sources listed in Appendix B (10-K FY2021–FY2025, Q1 2026 10-Q, 2026 DEF 14A, 8-Ks, Form 4s, ALTA/MBA industry data). Labels: Fact = sourced from filings or verified data; Interpretation = analyst judgment on facts; Assumption = estimate not directly sourced. No BUY/SELL recommendation and no price target appear here.

1. General — What thoughtful questions have other investors asked about this company?

The questions that matter for FAF, synthesized from the evidence base assembled for this report:

1. Is the title cycle at a trough or mid-cycle? Neither. Fact: domestic title orders closed fell from 1,050.7K (FY2021) to ~455K (FY2023 trough) and recovered to 531.9K (FY2025); MBA forecasts 2026 originations of ~$2.19–2.2T, +6–8% y/y, with 30-year rates ~6.0–6.5% through 2026–27. Interpretation: 2025–26 is an early-cycle recovery year — the Marathon capital-cycle read is that capacity exited in 2022–23 and cannot re-enter overnight, so the first leg of recovery brings margin expansion. The cycle is not at peak: 2026 purchase volumes remain well below 2020–21 levels and refi is only beginning to recover from historic lows.
2. Is the 12.1% FY2025 title pretax margin sustainable? Fact: title pretax margin swung 16.3% (2021 peak) → 4.3% (2024 trough) → 12.1% (2025); Q1 2026 adjusted pretax title margin was 10.4%. Interpretation: 12.1% sits above the base-case normalized margin used in this report (~10.75%) and below the 2021 peak; sustainability depends on commercial title strength and operating leverage persisting. Management describes a ~100 bps drag from running legacy and new platforms in parallel, suggesting modest structural upside as legacy systems decommission — but that is a management claim, not yet evidenced at scale.
3. What is normalized EPS? Fact: FY2025 reported diluted EPS $6.00; TTM ~$6.50. Interpretation (assumption-based): base-case normalized EPS power is ~$5.00–$5.50 (midpoint ~$5.25), after stripping the $39.8M reserve release and $14.8M insurance recovery and normalizing margin/mix; bear ~$3.50; bull ~$7.70.
4. Is the investment income tailwind durable? Fact: net investment income was $621.0M in FY2025 (8.3% of revenue) on a $9.4B, 99% investment-grade, AA-average portfolio; Q1 2026 investment income +12.8% y/y despite three Fed cuts, on ~19% deposit growth at First American Trust. Management’s rule of thumb: ~$15M annualized investment income lost per 25 bps Fed cut. Interpretation: durable near-term (deposit growth and duration are offsetting cuts; FY2026 title investment income guided roughly flat) but structurally rate-sensitive — a sustained cutting cycle compresses reinvestment yield.
5. How does FAF compare to FNF/ORI/STC? Fact: FAF is the #2 underwriter (ALTA 2025 share 23.1%; FNF family ~27–30% combined; Big Four ~75–80%+ of premiums). FY2025: FAF ROE 16.3% vs ORI 16.2%, FNF 10.5%; FAF P/B ~1.35–1.37x vs ORI ~1.79x, STC ~1.85x. Interpretation: ORI is the closest through-cycle comp (similar ROE, more agency-heavy); FNF is not a clean comp (F&G annuities distort it); STC is smaller and more volatile. FAF trades at a P/B discount to peers — the valuation warning comes from the company’s own history (P/S), not from relative valuation.

2. Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Fact: FY2025 EPS $6.00 is nearly 5× the FY2024 trough ($1.26) but still 46% below the FY2021 peak ($11.18); Q1 2026 EPS +70% y/y. Interpretation: earnings are mid-recovery, off the trough but above normalized mid-cycle power. The standard “cyclical high/low” question maps to the mortgage origination cycle: 2022–24 was the volume trough (30-year rates went from ~3% to 6–7%), and 2025–26 is the recovery. Earnings are cyclically elevated relative to normalized even though volumes are below peak, because margins and mix (commercial, ARPO) have recovered faster than order counts.

External environment or internal actions? Predominantly external. Fact: order volume tracks MBA origination volumes; average revenue per order rose from $2,718 (2021) to $3,961 (2025) mainly on mix shift toward purchase/commercial, not pricing power (state-regulated rates; ALTA notes national title costs declined 7.8% since 2004). Internal actions — cost discipline through the trough, Endpoint/Sequoia automation, commercial share gains, bank deposit growth — are real but second-order. Interpretation: FAF is a beta play on housing/CRE transaction volume with modest self-help levers.

How stable are revenues? Not stable at the consolidated level; two-speed underneath. Fact: total revenue swung $9.21B (2021) → $6.00B (2023) → $7.45B (2025). Title premiums are episodic — a transaction tax on real estate closings. The stabilizers: Home Warranty (~6% of segment revenue) is a renewal-based annuity (claims ratio 41.4% FY2025, improving from 48.8% FY2023; pretax margin 19.5%), and net investment income ($570–621M/yr 2023–25) is annuity-like. Interpretation: ~85% of the economics remain transaction-cyclical; the annuity pieces smooth but do not transform the earnings profile.

Outlook for products/services? Market size — growing or shrinking, domestic or international? Fact: U.S. title premium pool ~$18.5B in 2025, +13.8% from $16.2B in 2024 (ALTA); Fitch reports sector operating margin improved to 11.4% in 2025. Growth lanes: (1) cyclical origination recovery (MBA +6–8% 2026); (2) commercial title (Q1 2026 commercial revenue +48% to a Q1 record $271M; management expects a record 2026; MBA CREF +24%); (3) non-title fee income — mortgage subservicing, data/analytics (Esri/ArcGIS partnership July 2026), trust/wealth/banking ($6.8B balance sheet, $6.2B deposits at YE2025). Primarily domestic U.S. with Canada/UK operations. Interpretation: the core market is mature and GDP-ish through the cycle; near-term growth is cyclical recovery, not secular expansion. Home warranty competes with American Home Shield et al. and lacks title-style barriers.

3. Business Quality & Competitive Moat

Is the industry getting more or less competitive? Structurally stable. Fact: four families (FAF, FNF, ORI, STC) hold ~75–80%+ of U.S. title premiums; top seven ~85%. Digital challengers failed to scale — Radian exited title entirely in Sept 2025 to pivot to specialty insurance. Interpretation: Greenwald supply-side barriers are holding; technology is lowering incumbent unit costs rather than enabling entrants. The moat is not widening (AI erodes the physical title-plant edge over time) but is not under assault.

How profitable is the business (ROE/ROIC)? Fact: ROE: 39.2% (2021), 7.1% (2022), 5.8% (2023), 3.6% (2024), 16.3% (2025); ROIC ~12% (2025, ROIC.ai). Title combined ratio approximated at ~88% FY2025; loss provision only ~3.0% of premiums/escrow fees (retrospective risk — claims arise from pre-existing defects). Interpretation: through-cycle ROE is low-double-digits (~10–11% normalized), not the high-teens of a wide-moat franchise. The “mid-teens” 2025 print is a recovery-year figure, not run-rate.

How profitable is the industry — competitors, barriers? Fact/Interpretation: industry operating margin ~11.4% (2025, Fitch). Barriers: (1) proprietary title plants — decades-built, geographically indexed property-record databases; (2) state-by-state licensing, rate filing, statutory capital; (3) agent relationships and closing/escrow integration; (4) ratings-agency financial-strength requirements. These explain the persistent ~85% top-player share.

Can the business be easily understood? Yes — it insures clean title on real-estate transactions and invests the float. Complexity sits in the agency-channel economics and the bank/investment portfolio, not the core product.

Foreign low-cost labor risk? Low. Title search/exam is jurisdiction-specific and license-bound; automation (AI) is the real labor-cost vector, and incumbents own it.

Do brands matter? Modestly. Interpretation: brand matters with lenders/GSEs and agents (trust, ratings, service), not with end consumers, who rarely choose the underwriter. “First American” is a B2B brand, not a consumer moat.

Nature of competition? Switching costs? Competition is on service, quality, price, relationships, and product access — at two levels: for agents (premium splits and support) and for direct business (lender/realtor relationships). Fact: agents are independent, write for multiple insurers, and retained ~80% of agent premiums in FY2025 ($2,374.0M of $2,959.4M) — a stable ~79.5–80.2% retention band for five years. Interpretation: switching costs are moderate and relationship/operational (integration, data, service), not contractual; the agency channel is the moat’s main pressure point — underwriters compete wholesale for agents, which caps oligopoly pricing power. State rate regulation is double-edged: it caps upside but also prevents price wars.

4. Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Interpretation: the title plants are the classic case — internally developed over decades, carried at little or no value, yet they are the core cost/underwriting advantage and would cost enormously to replicate. Brand, agent network, and the First American Data & Analytics dataset (now monetized via the Esri partnership) are similarly unbooked intangibles. On the booked side: goodwill $1,819.3M (up from $1.379B at YE2020, mostly Mother Lode) and a $9.4B investment portfolio at fair value with a net unrealized loss of $190.0M (AOCI −$256.7M net of tax at YE2025, −$322.5M at Q1 2026).

Off-balance-sheet liabilities? Fact/Interpretation: the standard corporate question maps to insurance analogs: (1) claim reserves — on balance sheet ($1,166.3M total known + IBNR reserve at Q1 2026), and the real risk is adverse development, though recent development has been favorable ($39.8M release FY2025); (2) agent relationships — contingent, unbooked: agents owe remitted premiums and FAF bears underwriting risk on policies agents issue; (3) secured financings payable $906.5M (YE2025) — on balance sheet but functionally leverage against the portfolio; (4) bank deposits $5,292.7M (YE2025) rising to $7,082.0M (Q1 2026) — a liability that funds the asset surge and should be netted, not treated as operating leverage. Litigation is ordinary-course; the 2024 cyber class action is disclosed but unquantified.

How conservative is the accounting? Reasonably conservative, with normalization caveats. Fact: FY2025 title provision was 3.0% (3.75% current-year ultimate less 0.75% prior-year release of $39.8M); management describes the normalized loss rate as closer to 5% and expects releases to slow or stop. FY2024 absorbed $401.6M of consolidated net investment losses (including ~$345.4M of Title-segment realized losses on the ~$342M pre-announced Q3 2024 rebalancing, $2.8B of proceeds) plus a $39.5M software impairment. Management reports adjusted EPS ($1.33 vs GAAP $1.21 in Q1 2026) excluding investment losses and intangible amortization — a conventional, defensible adjustment set. Interpretation: reserve releases and a ~7% portfolio yield flatter 2025 earnings just as 2022/2024 investment losses depressed those years; through-cycle earnings require normalizing both directions. GAAP FY2025 earnings are cleaner than 2022/2024 but still carry ~$0.42/share of one-off benefit (reserve release + insurance recovery).

How CapEx-hungry is the business? Low — the analog of a data/tech company, not an industrial. Fact: FY2025 operating cash flow $950.8M vs capital expenditures $188.3M (GAAP FCF $762.5M); capex is ~2.5% of revenue and is predominantly technology (title production systems, automation, data). Interpretation: capital intensity is the tech budget, and it is currently elevated by running legacy and new platforms in parallel (~100 bps margin drag per management); there is no heavy fixed-asset reinvestment cycle.

5. Capital Allocation & Management

How much FCF, and how is it used? Fact: GAAP free cash flow (OCF − capex): FY2025 $762.5M ($950.8M − $188.3M); prior years ~$1.16B (2021), ~$713M (2022), ~$294M (2023), ~$833M (2024, ROIC-defined). Deployment FY2025: dividends $223.0M + buybacks $122.3M = ~$345M returned (~36% of OCF), the balance retained. Stated priorities: (1) reinvest in business/technology, (2) disciplined M&A, (3) return excess via dividends/buybacks; debt-to-capital target ~20% through the cycle (21.9% ex-secured-financings at YE2025; 30.7% headline). Regulatory dividend capacity from insurance subs: $382.0M in 2026 plus $113.6M of permissible loans — comfortably covering the ~$224M/yr dividend.

Significant acquisitions recently? Fact: M&A is essentially dormant: Mother Lode Holding Co. (May 2022, ~$300M cash, ~$203M goodwill) was the last material deal; FY2025 acquisitions totaled $10.5M; acquisition cash outflow was zero in FY2024 and FY2025. Interpretation: consolidation posture, not empire-building.

Dividend history? Fact: eight consecutive annual increases; declared DPS $1.66 (2019) rising each year; $2.18/share paid in FY2025 ($223.0M); quarterly raised to $0.55 in Q1 2026 = $2.20 annualized run-rate, ~3.0% yield at $73.03.

Buying back shares? Fact: share count down 10.5M (−9.3%) from YE2019 to YE2025 (102.0M; 102.1M at Q1 2026). Gross repurchases: $440.7M (2022), $72.7M (2023), $68.5M (2024), $122.3M (2025, avg ~$58.54), $33.5M Q1 2026; $266.5M remained on the $300M July 2025 authorization at Q1 2026. Interpretation: buybacks are opportunistic, not mechanical; the 2022 burst (~$441M) was poorly timed — executed near the top of the rate cycle just before investment losses widened — value-destructive in hindsight but not reckless.

Issuing large amounts of new shares to insiders? SBC/dilution? Fact: no net dilution — share count falls every year despite equity comp. Officer equity grants are routine (e.g., Seaton +127,874 shares granted over 24 months, partially offset by −49,664 tax-withholding dispositions); net insider flow is modestly negative through vesting-related sales. Interpretation: SBC is fully absorbed by buybacks; dilution is a non-issue.

Compensation policy? Fact: 2025 AIP = 50% adjusted pretax margin / 50% adjusted ROE (targets 9.7% / 9.5%; actuals 10.6% / 10.6% → 142% payout); 50% of LTI is PRSUs on 3-year relative TSR vs the S&P MidCap 400; ~90% of Seaton’s 2025 pay ($8.28M total) was performance-based; ownership guidelines 6× salary for the CEO; clawback, anti-hedging/pledging; Say-on-Pay >93%. Interpretation: pay is cycle-aware and margin/return-focused rather than growth-at-any-cost — genuinely well aligned. DeGiorgio’s $9.6M 2025 SCT figure is separation-inflated (no AIP payout).

Insider behavior? Fact: the only open-market purchase in 24 months: director Parker S. Kennedy bought 59,841 shares for ~$4.05M at $67.17–$67.92 in Feb 2026 (via Kennedy Enterprises L.P.). Officer open-market sales: Gilmore 110,145 @ $63.59 (Feb 2025), Seaton 50,000 @ ~$64–65 (Feb 2025), plus smaller Doti/McCarthy/Cornehl sales; most officer dispositions are mechanical F-code tax withholdings. Interpretation: mixed/neutral — one long-tenured director put ~$4M of new capital to work near $67–68 (a genuine bullish data point), while no named executive bought on the open market (a soft negative; 10b5-1 status of the sales is unparsed).

Motivations of management? Interpretation: Seaton (CFO-to-CEO, appointed April 2025 after DeGiorgio’s termination without cause) is running a turnaround-and-modernization agenda: Endpoint digital escrow (launched Dec 2025; 153 orders opened/47 closed as of Feb 2026, 30% automation, 80–90% long-term target), Sequoia title production (live on refi; “instant title” ambition for purchase), and commercial/data-analytics expansion. His incentives (margin + ROE + relative TSR) push toward profitable-cycle management rather than volume chasing. The motivation read: prove the digital efficiency thesis and compound through the recovery.

6. Valuation & Market Data

ADR, MLP, or K-1 issuer? Fact: no — a plain-vanilla NYSE-listed Delaware C-corp common stock; a standard 1099 filer, no partnership/K-1 complications.

Dividend policy? Fact: progressive dividend, eight straight annual increases; $2.18/share paid FY2025, $2.20 run-rate ($0.55 quarterly), ~3.0% yield at the 2026-07-17 close of $73.03; payout ~36% of FY2025 EPS, well inside the $382M regulatory dividend capacity.

How profitable is the business? Via EPS/ROE (the correct insurance analogs): FY2025 EPS $6.00, TTM ~$6.50, ROE 16.3% (2025) with a through-cycle base nearer ~10–11%; normalized EPS ~5.00–$5.50. Valuation markers: P/E ~11.2x TTM, ~12.2x FY2025, ~13.9x base-case normalized; P/S ~1.0x (market cap ~$7.46B / TTM revenue ~$7.71B) — the primary-source figure, not the AZI 1.26x convention; P/B ~1.35–1.37x on GAAP BVPS $53.92 (YE2025) / $53.77 (Q1 2026); EV/Sales ~1.1x. Interpretation: P/E reads mid-cycle; own-history P/S and P/B percentiles read late-cycle — the market prices FY2025-level earnings as roughly durable (implied durable ROE ~11.2% on a justified-P/B frame).

Is net income diverging from cash from operations? Fact: FY2025 OCF $950.8M vs net income $622M — OCF exceeds NI by ~1.5x; 2024 OCF $898M vs NI $131M; 2023 OCF $354M vs NI $217M. Interpretation: the divergence runs in the healthy direction and is structural for an insurer: premium/float and working-capital timing, non-cash investment marks (unrealized venture/AOCI moves hit NI but not OCF), reserve accruals, and bank-deposit flows. There is no accruals red flag; if anything, GAAP NI understates cash generation in recovery years, though investment timing (realized losses in 2024) cuts both ways.

7. Risks & Downside

What would cause the stock to decline? Ranked by evidence:

• Mortgage-cycle reversal (the core beta). Purchase orders/day were −4.1% to −6% y/y in Q1 2026 even in a recovery; a rate re-spike toward 7% would crush the nascent refi rebound (+48–57% y/y but only ~8% of direct revenue) and stall purchase. The InterestRate factor loading is −0.52 — rising rates are a direct headwind.
• Margin compression / normalization. If 2025’s 12.1% title margin mean-reverts toward 8–9% (commercial mix fades, reserve releases stop as management expects, platform dual-running costs persist), normalized EPS ~$3.50 and the current multiple re-rates upward on a falling E — the “cheap on peak earnings” trap the valuation work flags.
• Investment-income fade. A sustained Fed cutting cycle (~$15M/yr per 25 bps) without offsetting deposit growth; plus portfolio marks — a +100bp rate shock takes ~$418M (−4.9%) off portfolio fair value and hits book via AOCI.
• Reserve/underwriting adverse development. Recent years enjoyed releases ($39.8M FY2025; $11M prior-year + $13M Canada in Q4 2025); normalized loss rate ~5% vs the 3.0% provision means provisions could rise ~$100M+/yr if experience normalizes.
• Cyber. The Dec 2023 incident (systems isolated; investigation concluded May 2024; class action filed June 2024) was never cost-quantified — residual litigation/remediation tail risk in a data-centric business.
• Regulatory. State rate actions compeling price reductions, RESPA/kickback enforcement, and class-action “excess profit” scrutiny (flagged, not yet located in NY/CA) could cap pricing or worsen customer-acquisition economics.

Risk of a catastrophic loss? Chance of a total loss? Interpretation: total-loss risk is low. This is a regulated insurer with a $9.4B, 99% investment-grade, AA-average portfolio; $5.5B of equity; $382M of subsidiary dividend capacity; an undrawn $900M revolver; and a claims model (retrospective title risk, ~3% loss provisions) with no catastrophe-tail exposure of the P&C kind. The realistic bear case is a deep cyclical drawdown, not impairment: the 2022 rate shock produced a −41% peak-to-trough decline (5-year max drawdown −43.7%, 10-year −50.1%) and EPS fell ~89% peak-to-trough (2021→2024) — yet the dividend was raised every year through it. Equity-like volatility with bond-like survival characteristics; the downside scenario is a 2024-style trough (EPS toward $2.50–3.00), not zero.

8. Recent News & Events

Has the business environment changed recently? Two-year timeline (2024-07 → 2026-07):

• 2024-09-09 — Strategic investment-portfolio rebalancing announced; ~$342M pre-tax realized loss expected in Q3 2024 (final: ~$345.4M Title-segment realized losses on $2.8B of proceeds; consolidated FY2024 net investment losses $401.6M). The income-statement damage is now behind; portfolio yield is the benefit.
• 2024-12 — Endpoint digital escrow launched in one office; industry’s first AI-powered escrow closing.
• 2025-04-15 — CEO change: Ken DeGiorgio terminated without cause (effective Apr 10); CFO Mark E. Seaton named CEO; Matt Wajner promoted to CFO. Transition complete and stable; ~90% of new CEO pay performance-based.
• 2025-10-23 — Q3 2025: adjusted EPS $1.70; Sequoia running live refinance orders; AI engine deployed.
• 2026-02-12 — Q4 2025: adjusted EPS $1.99; one-time benefits $28M ($13M Canada reserve release + $15M insurance recovery).
• 2026-02-13/17 — Director Parker S. Kennedy open-market purchase: 59,841 shares, ~$4.05M at $67.17–$67.92 (Form 4 filed 2026-02-18).
• 2026-03-30 — 2026 proxy filed: board declassification proposal (annual elections phased in through 2029) and elimination of supermajority voting; board shrinking 11 → 9. Vote outcome at the 2026-05-12 meeting not yet confirmed in the filings reviewed.
• 2026-04-22 — Q1 2026 beat: revenue $1,838.0M (+16.2%), GAAP EPS $1.21, adjusted EPS $1.33 vs ~$1.06 consensus; adjusted pretax title margin 10.4%; commercial revenue +48% to a Q1 record; stock rallied ~19% over the following 20 days. Q2 2026 earnings scheduled July 22–23, 2026 — the next catalyst.
• 2026-05/06 — $0.55 quarterly dividend maintained; ACI Sky Workbench verified by Fannie Mae/Freddie Mac for UAD 3.6 (required Nov 2026) — a data/analytics credential with the GSEs.

Significant acquisitions? None in two years (FY2025 deals $10.5M total); the posture is organic investment + capital return.

Change in accounting policies? None identified. The notable reporting changes are the FY2023 segment rename (Specialty Insurance → Home Warranty, reflecting the P&C exit/runoff) and management’s adjusted-EPS convention (ex investment gains/losses and intangible amortization), which is stable and conventional.

New markets, facilities, management? New CEO/CFO (above); growth vectors are commercial title (record 2026 expected), mortgage subservicing, data & analytics (Esri/ArcGIS partnership, July 2026), and the federal savings bank (deposits +34% to $7.1B in Q1 2026, 29% non-captive). No new geography of consequence; Canada/UK remain ancillary.

All figures trace to the public primary sources listed in Appendix B (SEC filings, earnings-call transcripts, industry data, and public data services). Key reconciliations applied throughout: dividend $2.18 paid FY2025 / $2.20 run-rate; GAAP FCF $762.5M; P/S ~1.0x (primary-source); GAAP BVPS $53.92 (YE2025) / $53.77 (Q1 2026); 2024 rebalancing wording ($342M 8-K estimate / ~$345.4M Title-segment realized losses / $401.6M consolidated net investment losses). This appendix carries no BUY/SELL recommendation and no price target.

APPENDIX B — Source Appendix

First American Financial Corporation (NYSE: FAF) — Report date 2026-07-19

Company: First American Financial Corporation (NYSE: FAF, CIK 0001472787)

All sources below are public.

(a) SEC Filings (via EDGAR, CIK 0001472787)

Annual Reports (10-K)

Quarterly Reports (10-Q)

All available on EDGAR (CIK 1472787); the full 10-Q series back to 2021 was reviewed.

Proxy Statements (DEF 14A / PRE 14A / DEFR14A)

Key 8-Ks referenced

Insider Filings (Forms 3/4)

Key Form 4s (full series available on EDGAR):

• Parker S. Kennedy open-market purchases (59,841 shares, ~$4.05M): Form 4 filed 2026-02-18, EDGAR accession 000119312526057792.
• Parker S. Kennedy stock award: Form 4 filed 2026-02-23.

Other registration statements and reports reviewed

• S-3ASR (2024-09-19, 2025-02-21)
• S-8
• Annual reports to shareholders (FY2023–FY2025)
• N-PX

(b) Transcripts

Q1 2026 earnings press release (8-K Ex. 99.1): https://www.sec.gov/Archives/edgar/data/1472787/000119312526170438/faf-ex99_1.htm

Transcript provenance: Alpha Spread (third-party transcription). Not a primary SEC source; quotes were cross-checked against the 8-K/10-Q for GAAP figures. Non-GAAP items (adjusted EPS $1.33, adjusted pretax title margin 10.4%) are management-defined and appear in the press release/transcript only.

© Public Data Services

(d) News / Trade Press

• ALTA press release, “Title Insurance Premium Volume Increased 7% in 2024,” 2025-05-15 — industry premiums $16.2B; top-10 underwriter market shares (FAF 22.2% in 2024). https://www.alta.org/news-and-publications/press-release/ALTA-Reports-Title-Insurance-Premium-Volume-Increased-7-in-2024
• Zacks, “FAF Tops Q1 Earnings and Revenue Estimates on Commercial Strength,” 2026-04-24 — Q1 2026 beat ($1.33 adj. EPS vs $1.06 consensus). https://www.zacks.com/stock/news/2907773/faf-tops-q1-earnings-and-revenue-estimates-on-commercial-strength
• Business Wire releases (FAF investor relations): $0.55 dividend declaration 2026-05-12; Q2 2026 earnings-call scheduling 2026-07-08; ACI Sky UAD 3.6 verification 2026-06-08; CRE AI-adoption survey 2026-05-12.
• MBA origination forecast (Oct 2025): total single-family originations +8% to ~$2.2T in 2026. https://www.mba.org/news-and-research/newsroom/news/2025/10/19/mba-forecast--total-single-family-mortgage-originations-to-increase-8-percent-to--2.2-trillion-in-2026
• iEmergent Q2 2026 mortgage forecast update. https://www.iemergent.com/insights/q2-2026-mortgage-forecast-update
• Alpha Spread Q1 2026 earnings-call summary (commercial +48%). https://www.alphaspread.com/security/nyse/faf/investor-relations/earnings-call/q1-2026

(e) Regulatory / Public References

• ALTA (American Land Title Association) — title-industry premium volume and market-share statistics; primary industry data source for market-share claims.
• MBA (Mortgage Bankers Association) — mortgage origination forecasts used to frame the title-order cycle.
• CFPB / RESPA — regulatory backdrop cited in the 10-K risk factors (settlement-services regulation, anti-kickback/fee rules); no FAF-specific enforcement action located.
• State insurance regulation framework (rate promulgation/filing regimes) — described in 10-K FY2025 Item 1.

Claims Not Primary-Sourced