D.R. Horton, Inc. (NYSE: DHI) — The Lowest-Cost Builder on an Average Street, Priced for the Trough to Be In

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows is deliberately position-free and carries no price target.

Verdict: HOLD / not-a-short at ~$158. Accumulate-on-weakness in the ~$115–130 zone (~1.4–1.6x book), where you are paid for the residual cyclical risk; do not chase toward the high-$170s/$180s (~2.1x+ book, ~17x trough-ish earnings). Low-to-medium conviction. Tag: “A returns machine idling in the wrong gear of the cycle.”

D.R. Horton is a genuinely well-run, best-in-class operator — the largest, lowest-cost US homebuilder, a fortress balance sheet (net debt only ~$3B against ~$24.7B equity), and a ~8%/yr share-count-shrinking buyback funded by real cash flow. But two things keep this a HOLD rather than a buy at ~$158. First, the “cheap 14.7x earnings” headline is a cyclical mirage. Homebuilders look cheapest on P/E at the peak (FY22 printed $16.51 EPS at a ~4x P/E) and most expensive at the trough; today’s ~14.7x sits on cyclically declining earnings ($16.51 → $11.57 FY25 → ~$10.76 TTM, FY26 guide ~$10–11). On the only multiple that behaves sensibly across the cycle — price-to-book at ~1.9x — DHI is above its mid-cycle, full but not extreme. Second, the margin reset may be more than cyclical: home-sales gross margin has collapsed from ~31% (FY22) to ~20% as DHI effectively rents its volume through mortgage-rate buydowns, and at 6–7% mortgages that incentive drag may be the new normal, not a temporary trough.

The framing — grounded in the factor tape — is that DHI is a housing-rate macro bet wearing a quality-compounder costume. Its dominant factor loading is Industry: Home Construction (~2.30); beta is a deceptively low 0.77 and alpha is slightly negative (‑0.03), so the +32% one-year run is sector beta, not idiosyncratic alpha. The market (and Berkshire’s 2024 endorsement) is underwriting “the trough is in” — that is roughly the base case, already priced. Bull-flip: gross margins re-expand toward ~21–22% with falling incentive intensity as the 30-year mortgage eases toward ~5.5% and order growth turns durably positive (proof the margin reset was cyclical). Bear-flip: FY26/27 EPS prints below ~$10 with incentive spend still rising as a share of price even as volume holds — i.e., ~20% gross margin is the new ceiling, and the market is over-capitalizing trough-ish earnings at ~1.9x book. This is not a short — spread-positive ROIC, real housing undersupply, and a balance sheet that lets DHI take share through the downturn rule that out — but the easy money has been made.

📈 Stock Price Action — Five-Year Event Map

Factual price history (AZI five-year CSV). Price moves are FACT; attributed drivers are INTERPRETATION. No recommendation, no price target.

Arc. Over the trailing ~60 months DHI completed a full cyclical round-trip and then some: it bottomed at a $56.87 intraday low (17-Jun-2022) as the Fed’s hiking cycle gutted builder sentiment, ran to an all-time intraday high of $195.88 (19-Sep-2024) on rate-cut euphoria plus the disclosure of a Berkshire Hathaway stake, chopped down through the 2025 affordability slowdown to a ~$120 low (20-Jun-2025), and recovered to close $157.81 on 18-Jun-2026 — roughly ‑19% off the all-time high, inside a 52-week range of ~$120–$183. The price action of the last two years is a housing-rate trade, not a fundamentals trade — DHI tracked the 10-year/30-year-mortgage path far more closely than its own (steadily declining) EPS.

Cycle narrative. (1) 2021 boom: ultra-low rates and pandemic household formation drove record absorption. (2) 2022 crash: the most violent leg — the fastest hiking cycle in 40 years doubled mortgage rates and crushed builder multiples even as DHI printed all-time-peak EPS of $16.51; the multiple compressed exactly when earnings peaked (P/B fell to ~1.2x, P/E to ~4x) — the textbook cyclical inversion. (3) 2022–23 doubling: once the market sensed a rate peak, builders led the recovery as the incentive/forward-commitment machine restored affordability. (4) 2023–24 run to ATH: Fed-pivot optimism plus the Berkshire 13F endorsement (Aug-2024) re-rated the name while the buyback shrank the float into a rising tape. (5) 2024–25 drawdown: “higher-for-longer” reset the multiple as the 10-yr backed up and the affordability ceiling bit. (6) 2025 recovery: renewed cut hopes plus mechanical buyback support; “the earnings trough is in” became consensus. (7) 2025–26 chop: sideways, caught between a soft spring selling season / compressing gross margins and ~8%/yr share-count reduction.

1. Executive Summary

D.R. Horton is the largest homebuilder in the United States by units — a title held since 2002 — closing 84,863 homes at a $370,400 average price in FY2025 across 126 markets in 36 states, generating $34.25B of revenue. It is not a manufacturer: substantially all land development and construction is outsourced to subcontractors, and the company has restructured itself around three capital-efficiency levers — land-light lot control (at FY25, ~575K lots, ~23% owned / ~77% controlled via non-recourse option contracts), a 62%-owned, separately-listed land developer (Forestar) that supplies the lots, and a captive mortgage operation (DHI Mortgage) that finances ~81% of its own buyers and delivers the rate buydowns that close the affordability-constrained, ~65%-first-time-buyer customer base. DHI is, in short, a national land-aggregation, purchasing, and project-management orchestrator wrapped around the most affordable rung of the housing market.

The business is genuinely well-run but sits in a structurally mediocre, deeply cyclical industry. Its competitive advantage is real but narrow: a supply-side, economies-of-scale cost advantage at the entry-level tier (volume discounts on materials, lower subcontractor labor rates, G&A leverage, captive lot supply, captive mortgage). There is no demand captivity — a home is the most infrequent considered purchase there is, switching costs are nil, and the product is a commodity. In Greenwald’s terms this is economies of scale without customer captivity, which is far weaker than scale with captivity. The tell: the high returns are cyclical, not structural — ROIC swung from 25% (FY22) to 10.8% (FY25); ROE from ~36% to ~12–15%; gross margin from 31.4% to 23.7% (home-sales GM ~20% by Q2-FY26). And the land-light model that defenders cite as a moat is being copied by every major peer (DHI and #2 Lennar are both racing to imitate #4 NVR’s asset-light template — proof it is competitively necessary table-stakes, not a defensible edge).

The cycle is the story. DHI is past its post-COVID peak and grinding lower: revenue, EPS, closings, and backlog are all down year-over-year, and the margin compression is driven by mortgage-rate buydowns that DHI uses to rent volume against a ~6.5–7% mortgage affordability ceiling. FY26 guidance (revenue $33.5–34.5B, 86–87.5K homes, gross margin ~19.7–20%, EPS roughly $10–11) implies continued mid-cycle grind, not recovery — recovery is contingent on rate relief the company cannot control.

What redeems the story is capital allocation and balance-sheet quality, which are close to best-in-class for a cyclical. DHI returned essentially all of its FY25 operating cash flow ($3.4B) to shareholders, repurchasing ~$4.35B of stock (share count 324M → 294M, ~8%/yr shrinkage) and paying a small, fast-growing, easily-covered dividend (~13–14% payout), while carrying only ~$3B of net debt against ~$24.7B of equity (21.7% leverage). The incentive plan is anchored on returns and relative TSR — not volume or revenue growth — so the empire-building flag is absent. The honest blemishes: DHI now buys back stock at ~1.9x book (vs. ~1.0–1.5x in 2022–23, when it was far more accretive); it built a ~$3B rental platform at the cycle peak that it is now unwinding; founder/Chairman Donald R. Horton died in May 2024, leaving a professionally-managed company with insider ownership below 1%; and insiders have made zero open-market purchases.

Valuation is the crux. The “cheap 14.7x P/E” is a cyclical artifact — the denominator (earnings) is falling. On price-to-book (~1.9x), the only multiple that behaves across the cycle, DHI is above mid-cycle but below peak-froth — full, not extreme. The market is underwriting a base case of mid-cycle normalization (~$10–12 sustainable EPS, gross margin stabilizing near ~20%, buyback compounding per-share value); that is roughly where the stock trades, leaving a thin margin of safety. The asymmetry from ~$158 is roughly symmetric-to-slightly-unfavorable: the bull needs both rate relief and margin re-expansion; the bear needs only the affordability ceiling to persist.

2. Business Overview

What it is. D.R. Horton (“America’s Builder”) is the #1 US homebuilder by volume, a position it has held since fiscal 2002. In FY2025 (ended 9/30/25) it closed 84,863 homes at a $370,400 average closing price through operating divisions in 126 markets across 36 states (the company’s older “118 markets / 33 states” descriptor is stale). It is an S&P 500 constituent and, since June 2025, dual-lists on NYSE and NYSE Texas. Roughly 84% of FY25 home-sales revenue came from single-family detached homes, the remainder from attached product (townhomes, duplexes, triplexes); homes range ~1,000–4,000 sq ft and ~$250,000 to $1,000,000+ — but the strategic center of gravity is the entry-level / affordable tier: average price ~$361,600 (Q2-FY26) sits roughly 30% below the US new-home average and ~$70K below the median existing home, and ~65% of closings finance first-time buyers.

Five reporting pieces. The company runs ~92 decentralized homebuilding operating divisions aggregated into six homebuilding regions (Northwest, Southwest, South Central, Southeast, East, North), plus three non-homebuilding segments:

• Forestar Group (NYSE: FOR), 62%-owned: a separately-listed residential lot developer that DHI consolidates. In FY25 Forestar sold 14,240 lots, 83% of them to D.R. Horton — functionally a consolidated-but-public land bank that feeds DHI finished lots while sharing development capital/risk with minority holders.
• Financial services (DHI Mortgage + captive title): originated or brokered 68,982 loans in FY25 against 84,863 closings — a ~81% mortgage-capture rate of its own buyers — and sells substantially all originations and servicing to third parties (71% to the GSEs/Ginnie Mae). It earns a ~27% pretax margin but exists to close DHI’s homes and deliver the rate buydowns, not to compete as a standalone lender.
• Rental (single-family + multifamily): ~$3B of inventory, sold in bulk; being deliberately trimmed.

The geographic tilt is Texas/Sun Belt: South Central (22,319 homes FY25) and Southeast (20,390) are the volume core; the Northwest is the highest-ASP region (~$537,500).

The “land-light” architecture (the most important fact about the model). The defining structural choice is capital-light land control. DHI owns roughly 23% of its ~575K controlled lots and options the other ~77% through purchase contracts that are “generally non-recourse,” limiting exposure to forfeited earnest money. Combined with fully outsourced land development and construction (DHI employs no trades), this converts a historically balance-sheet-heavy, cyclical capital sink into something closer to an asset-light orchestrator — the model NVR pioneered and DHI and Lennar are now imitating. In Q2-FY26, 67% of homes closed were on lots developed by Forestar or third parties (up from 64% a year earlier).

Brand architecture — a quiet but telling change. The FY2022 10-K sold homes “primarily under the names of D.R. Horton, America’s Builder, Emerald Homes, Express Homes and Freedom Homes.” The FY2025 10-K drops the sub-brands entirely, describing instead a product tier spectrum (entry-level, move-up, active adult, luxury) under one D.R. Horton brand. This is consistent with the thesis that the “brands” were never a demand-side moat — they were price-point labels, and management has stopped pretending otherwise. (INTERPRETATION.)

Recurring vs. non-recurring revenue. Essentially none of the revenue is contractual/recurring — each home is a one-off sale to a buyer with zero switching cost and the longest repurchase cycle of any consumer purchase. The closest thing to recurring economics is the captive mortgage/title attach, which rides on home-closing volume. Backlog (10,785 homes / $4.1B at FY25, down 14% YoY) provides only a few months of forward visibility.

Verdict: A genuinely national, scaled, vertically-coordinated production homebuilder that has restructured around land-options + outsourced construction + captive finance to lower the capital intensity of an inherently cyclical, low-differentiation product. The business is real and well-run; the question the rest of this memo answers is whether scale at the affordable tier is a durable moat or merely operational excellence in a commodity.

3. Industry Dynamics

Market size and the demand engine. The relevant market is US new single-family construction, gated almost entirely by single-family housing starts — which are in turn gated by mortgage rates, affordability, employment, and household formation. Starts ran ~941K in 2025, ~17% below the 2021 peak and below the ~1.1–1.5M household-formation level (Census/NAHB data; corroborated by Builders FirstSource disclosures). The path: starts boomed on COVID-era sub-3% mortgages, collapsed in 2022–23 as the 30-year ran to ~6.5–7%, and have stagnated since. The cross-read peers corroborate a frozen demand backdrop: existing-home sales are stuck near a 30-year low (~4.0M SAAR), housing turnover is at a ~40-year low (~2.9% of stock), and ~80% of mortgaged homeowners hold rates below the current ~6.3–6.5% — the “rate lock-in” that suppresses both moves and trade-up demand (NAR/Freddie Mac data; corroborated by Home Depot disclosures).

Structure — historically fragmented, now consolidating to the publics. Homebuilding was for decades a fragmented cottage industry of local/regional builders. It is now consolidating hard to a handful of scaled public builders: DHI #1, Lennar #2, then PulteGroup, NVR, Meritage, KB Home, Toll Brothers, Taylor Morrison. The top builders take a steadily larger share of starts because large production builders out-buy, out-land, and out-finance local competitors, and because building-products suppliers prefer to concentrate volume with fewer multi-region builders. This is a genuine, multi-decade share-consolidation tailwind that favors DHI — but consolidation of a commodity industry raises returns only at the margin; it does not by itself confer pricing power, because the product remains undifferentiated and the buyer is rate-driven. (INTERPRETATION.)

Profit pools and cyclicality. Homebuilder economics are violently operating-levered and cyclical. DHI’s own gross margin went 31.4% (FY22 peak) → 23.7% (FY25 total), with home-sales GM ~21.5% (FY25) and ~20.1% in Q2-FY26; ROE went ~36% (FY21/22) → ~12–15% (FY25) and ROIC 25% → 10.8%. The margin compression is overwhelmingly driven by mortgage-rate buydowns and sales incentives — the FY25 10-K states home-sales GM “decreased to 21.5% as we increased sales incentives, such as buydowns of mortgage rates,” and management “expect[s] to maintain an elevated level of sales incentives… and may increase them further.” This is the textbook commodity tell: when demand softens, the builder cuts the effective price (via buydowns) rather than holding price and losing volume — there is no franchise margin to defend.

Regulation. The sector is gated by local zoning and land-entitlement (the binding supply constraint and a real barrier favoring incumbents who already hold entitled lots), building codes, and — critically — GSE/FHA mortgage finance, since the affordable buyer depends on Fannie/Freddie/Ginnie conforming and FHA loans (71% of DHI’s originations flow to the GSEs/Ginnie). Entitlement difficulty is double-edged: it constrains supply (supporting prices) and advantages scaled incumbents with land-acquisition teams and option pipelines, but it does not create demand captivity.

Barriers to entry — modest, mostly scale/land, not demand. A determined entrant cannot easily replicate DHI’s land pipeline, subcontractor relationships, purchasing scale, or capital access — real frictions. But customers face essentially zero switching costs, and there is no demand captivity; new entrants and small local builders compete every cycle at the community level. Barriers exist on the cost/supply side and are weak-to-absent on the demand side — which, per Greenwald, caps how durable any returns advantage can be.

Capital-cycle position (Marathon). Homebuilding ran the canonical Marathon boom (2020–22): record profitability, builders flooding spec inventory and land-buying — followed by the rate-driven bust (2023–25). The current phase shows clear late-bust signatures: gross margins down ~1,100 bps from peak, builders delaying starts to work down spec inventory, per-start de-contenting, and DHI’s backlog down 14% YoY. But the important Marathon nuance is that capital has NOT been violently destroyed and supply has NOT been forced out — the public builders entered this downturn with fortress balance sheets, and the land-light model lets them throttle starts rather than dump owned land at a loss. The result is a managed, shallow down-cycle — incentive-driven margin compression instead of a 2007–09-style solvency wipeout. This is double-edged: it caps both the downside (no capitulation low) and the mean-reversion upside for survivors (no washed-out competitors to harvest). The capital cycle here is being dampened by balance-sheet strength and asset-light flexibility, not running its full punitive course.

Verdict: Structurally a MEDIOCRE industry that has gotten incrementally better — but not good. Demand is exogenous, rate-gated, and deeply cyclical; the product is a commodity with no demand captivity; returns mean-revert hard (DHI’s own ROE roughly halved peak-to-trough). The genuine structural improvement is supply-side: consolidation to scaled publics, the land-light model dampening the capital cycle, and chronic under-building vs. household formation providing a long-run demand floor. Net: a cyclical, capital-intensity-reduced industry where the scaled, low-cost, affordable-tier leader can earn adequate-through-cycle returns — but not the durable 20%+ ROICs of a true franchise. A good house on an average street.

4. Competitive Position

The honest baseline: homebuilding is a commodity. Apply Greenwald’s three-step test. (1) Landscape: the top is countable on one hand (DHI, Lennar, Pulte, NVR, then a tier of regionals) — suggesting some barriers — but it remains contestable by local builders community-by-community. (2) Profitability test, through-cycle: DHI’s ROIC swung 25% (FY22) → 10.8% (FY25); a true franchise delivers sustained 15–25% ROIC across a full cycle, and DHI’s trough is below that floor. (3) Source: there is no proprietary technology (home construction is the definition of “in the long run, everything is a toaster”) and no demand captivity — a homebuyer has zero habit, zero switching cost, and high search willingness on the largest purchase of their life. The demand side offers no moat.

Where a real advantage exists: scale-driven cost advantage at the affordable tier. DHI’s own 10-K names the mechanism precisely: scale provides “volume discounts and rebates from national, regional and local materials suppliers and lower labor rates from certain subcontractors,” “greater access to and lower cost of capital,” and “enhanced leverage of our G&A activities.” This is a genuine supply-side / economies-of-scale cost advantage — the strongest of Greenwald’s three types when combined with captivity, but here it stands largely alone. The advantage is real and measurable: DHI builds the cheapest home, buys materials and subcontractor labor more cheaply than a local builder, runs even-flow production for higher inventory turns, and — via Forestar — has captive finished-lot supply. At the entry-level price point specifically, scale + the captive mortgage (which delivers the buydown that closes the affordability-constrained buyer) is a defensible, if narrow, cost moat. SG&A around ~8–9% of revenue even in a down year reflects the G&A leverage.

Pressure-test #1 — durable or cyclical ROIC? Largely cyclical. The 25% → 10.8% ROIC collapse maps almost one-for-one to the gross-margin collapse, which maps to rate-driven incentives. The normalized through-cycle return — homebuilding pretax ROI on inventory of ~17.6% TTM (~20% FY25 per management) — is genuinely above commodity (≤8%) and reflects the scale/land-light advantage, but it is not a stable 20%+ franchise return. DHI earns an above-commodity but below-franchise return through the cycle (~15–18% inventory ROI) — exactly what a real-but-narrow scale cost advantage in a no-captivity commodity should produce.

Pressure-test #2 — market-share stability. DHI has held #1 since 2002 and is steadily gaining aggregate US share, with community count +11% YoY. Two-decade dominant-firm longevity plus rising share passes Greenwald’s stability test and is the strongest single piece of moat evidence. But it is share gain in a commodity (raises returns modestly), not share defended by captivity (would raise them a lot).

Peer triangulation — the NVR benchmark. The cleanest test of “is land-light a moat or just a better operating model” is NVR, the asset-light land-option pioneer, which earns the highest ROE in the group (frequently 30%+ through-cycle) precisely because it owns almost no land and avoids the downturn impairment. DHI (23% owned) and Lennar (also going land-light, having spun out its land bank Millrose) are converging toward NVR’s model — which proves the point: the land-light advantage is replicable operating discipline, not a proprietary moat. The whole industry is copying it. Where DHI does have a genuine edge over NVR is absolute scale — purchasing power and the affordable-tier volume machine that NVR, at ~1/5 the volume, cannot match. PulteGroup runs a higher-margin move-up mix at lower volume; Lennar is DHI’s only true scale peer, pursuing the identical playbook.

Verdict: A REAL BUT NARROW supply-side / economies-of-scale COST advantage at the affordable tier — NOT a demand-side franchise moat. The mechanism (purchasing scale, lower subcontractor labor cost, G&A leverage, captive Forestar lots, captive mortgage delivering the affordability buydown) is genuine, financially visible, and durable enough to keep DHI the lowest-cost producer at the entry-level rung and to support an above-commodity ~15–18% through-cycle inventory ROI. But there is zero demand captivity, the product is a commodity, the 20%+ peak ROICs are cyclical not structural, and the land-light edge is being actively replicated by every major peer. This is the classic Greenwald case of economies of scale without customer captivity — which he explicitly warns is far weaker than scale with captivity. The right characterization: operational excellence plus a narrow scale-cost advantage in a commodity industry — not a wide-moat compounder.

5. Growth History and Forward Opportunities

The revenue arc — boom, plateau, roll-over. FY20 $20.3B → FY21 $27.8B → FY22 $33.5B → FY23 $35.5B → FY24 $36.8B → FY25 $34.25B, guided to $33.5–34.5B and 86–87.5K closings in FY26. Diluted EPS tells the cyclical story even more starkly: $6.41 (FY20) → $16.51 (FY22 peak) → $11.57 (FY25), with TTM ~$10.76 still falling. On a multi-year view the company is past the cyclical peak and grinding lower — FY25 closings (84,863) were down ~5% YoY from 89,690 and home-sales revenue fell ~7%.

Volume vs. price. Growth here is fundamentally a volume story constrained by an affordability ceiling. ASP is being held down (~$361,600 in Q2-FY26, deliberately below market) and effective price cut further via buydowns — so there is no price-led growth; mix/incentive is a price headwind. The growth lever is units, gated by how affordable management can keep the monthly payment. Community count +11% YoY is the genuine organic expansion signal (more selling locations is how a builder grows volume), but it is being run into a soft demand market — hence backlog ‑14% and orders +11% YoY on lower per-unit value (order ASP $366,300, ‑2% YoY).

Organic vs. acquired. DHI’s growth is overwhelmingly organic — community-count expansion, geographic infill, and aggregate share gain, fed by the Forestar lot pipeline rather than transformational M&A. This is a positive on quality (no goodwill-laden empire-building) and consistent with the consolidation tailwind — but it is still cyclical commodity volume.

Forward opportunities — four, of varying quality.

1. Aggregate share gain (highest conviction). The multi-decade consolidation to scaled publics is real; DHI is the prime beneficiary as the largest, lowest-cost, most affordable builder. This continues largely regardless of cycle.
2. Land-light scaling + Forestar (medium). Pushing owned-land below 23% and routing more lots through Forestar/third parties raises capital efficiency and ROIC durability — but, as established, it is table-stakes the whole industry is adopting, so it is ROIC defense, not a growth engine.
3. Rental monetization (low/neutral). Trimming the ~$3B rental book is sensible capital discipline, not growth; it frees capital for buybacks and the core.
4. First-time-buyer demographic tailwind vs. affordability ceiling (the swing factor). Bull case: a large millennial/Gen-Z household-formation wave hitting prime first-home age against a chronically under-built housing stock — structural demand DHI is best-positioned to serve. Bear case: the affordability ceiling — at ~6.5–7% mortgages and stretched price-to-income, the entry-level buyer DHI depends on is exactly the cohort priced out, which is why buydowns eat the margin. These forces are in direct tension and the resolution is rate-path-dependent — i.e., largely exogenous. (OPEN QUESTION: does the demographic floor reassert as rates ease toward ~6%, restoring volume and relieving the incentive drag — or does affordability stay broken, keeping DHI in a low-margin, high-incentive grind?)

Verdict: MEDIUM-QUALITY growth — high-quality in structure, low-quality in current cyclical posture. The growth is organic, share-gain-driven, and self-funded (no value-destructive M&A), with a genuine long-run demographic + under-supply tailwind and the best competitive position to harvest it. But the company is past its cyclical peak and contracting (revenue, EPS, closings, backlog all down YoY), the only price lever points down (deeper buydowns), and the central forward driver (first-time-buyer demand) is hostage to an affordability ceiling DHI cannot control. Growth resumes meaningfully only if rates ease; absent that, the story is defending share and ROIC in a managed down-cycle, not growing.

6. Financial Quality

Income statement — operating leverage in reverse. The multi-year arc shows the cycle plainly (FY ends Sep 30):

The pattern is unambiguous: revenue plateaued FY23–24 and rolled over in FY25, while margins and returns compressed every year since the FY22 peak. The gross-margin slide (31.4% → 23.7%, and ~20% by Q2-FY26) is the central earnings-quality fact — it is rate-buydown/incentive-driven, i.e., the cost of buying affordability for the entry-level buyer. SG&A as a percent of revenue is rising (9.2% in Q2-FY26 vs. 8.9% a year earlier) as the denominator (ASP × volume) shrinks — negative operating leverage. The effective tax rate is steady (~23–24%). Net margin compressed from ~17.5% (FY22) to ~10.5% (FY25). One nuance for run-rate: Q2-FY26 home-sales GM of 20.1% included a ~40bps benefit from a favorable litigation outcome and low warranty costs; normalized ~19.7%.

Cash flow — the cyclical signature and an honest read of “FCF.” Homebuilder operating cash flow is dominated by inventory swings, which makes it counter-cyclical to reported earnings. In the boom, cash is consumed building inventory (FY21 OCF just $534M, FY22 $562M against $5–6B of net income — the cash went into a $5.2B FY22 inventory build); as the cycle turns and inventory is harvested, OCF surges (FY23 $4.30B, FY25 $3.42B). Because DHI has minimal true capex (~$100M/yr of D&A; it owns almost no PP&E — net fixed assets are only ~$86M), operating cash flow ≈ free cash flow. FY25 FCF was ~$3.42B (~$11/share). The earnings-quality caveat: this “FCF” is real but is a function of where you stand in the inventory cycle — it looks weak at the bottom of demand (cash sunk into land/WIP) and strong on the way down (harvest). Stock-based compensation is small and non-distorting (~$131M, ~0.4% of revenue). There are no aggressive accounting tells; warranty/litigation reserves and impairments are the items to watch, and impairments have been minimal this cycle (the land-light model is designed to avoid them).

Balance sheet — a fortress by homebuilder standards. At FY25: cash $2.99B, total debt $6.03B → net debt ~$3.0B against ~$24.7B equity (net-debt/equity ~12%; homebuilding leverage 21.7% at Q2-FY26 vs. a ~20% target). Inventory is the dominant asset at $25.3B; the cash-conversion cycle is long (~330 days) because land/WIP is inherently slow — but the land-light model (77% optioned) keeps owned inventory and impairment risk down. Liquidity is ~$6B; only ~$600M of senior notes mature in the next 12 months. This low leverage is what lets DHI keep buying back stock and taking share through a soft patch rather than retrenching — the supply-side discipline that historically lets the strongest builders win downturns.

Returns and unit economics. Even at a cyclical trough, returns are spread-positive: FY25 ROIC ~10.8%, ROE ~12–15%, homebuilding pretax ROI on inventory ~17.6–20%. Per-home economics: ASP ~$361–370K at ~20% home-sales gross margin = ~$72–74K gross profit per home, before SG&A and the captive-mortgage contribution. The captive financial-services segment (~27% pretax margin) adds a high-return, low-capital earnings layer that scales with closings.

Verdict: High-quality operator, cyclically-deteriorating economics, fortress balance sheet. Economics do not durably improve with scale in the franchise sense — they improve with the cycle, and we are on the wrong side of it (margins and returns compressing, EPS past peak). But the cash conversion is genuine, the balance sheet is conservative, capex is trivial, and the accounting is clean. This is a financially well-managed cyclical, not a structural compounder — and the reader must not confuse the two.

7. Capital Allocation

The model: aggressive, returns-disciplined, FCF-funded capital return — a clear STRENGTH. DHI runs one of the most shareholder-friendly programs in homebuilding, funded by genuine free cash flow generated as the cycle turned, not by leverage. In FY25 it generated $3.4B of operating cash flow and returned all of it to stockholders; the proxy’s framing is blunt — “Fiscal 2025 stockholder distributions increased by $2.6 billion or 118% from the prior year.” FY25 repurchases were ~$4.35B, taking the share count from ~324M to ~294M (~8%/yr reduction) — a rate of float shrinkage few large-caps in any sector match. The cadence continues: in H1-FY26 DHI repurchased 10.4M shares for $1.6B (~$154/sh), with $1.7B remaining on the authorization; FY26 guidance is ~$2.5B of buybacks. The board also authorized $500M of debt-securities repurchases (July 2024).

Marathon / supply-side lens. DHI exhibits the rare combination the Capital Returns framework prizes: a shrinking share count alongside (cyclically) above-cost-of-capital returns, with management not plowing every dollar into balance-sheet growth at the top of the cycle. Instead it shifted to the land-light model ($26B of lots controlled via options at 9/30/25) that structurally lowers capital intensity and lifts inventory turns and ROIC. This is the single biggest capital-efficiency lever and is genuinely value-accretive — it converts a capital-heavy land-banking business into a higher-turn, option-based one, freeing the cash that funds the buyback.

Dividend. $0.45/qtr ($1.80 annualized), grown from $0.70/yr in FY20; FY25’s ~$1.60/sh was a ~13–14% payout on ~$11.57 EPS — almost trivially covered, with enormous room to grow. DHI deliberately keeps the payout low and channels the bulk of returns through buybacks (more tax-efficient and optionality-preserving for a cyclical). Sustainability is not in question even through a downturn. Dividend yield is modest (~1.1%).

Incentive alignment. The proxy is reassuring. The annual bonus is based on consolidated pre-tax income with a hard cap and downward-only Committee discretion (50% cash / 50% equity). The long-term PSUs (majority of target equity, 3-yr performance) vest on a basket of return and shareholder-return metrics — pre-tax income, TSR relative to the S&P 500, ROA relative to S&P 500 companies, pre-tax ROA relative to a peer group, stock price, and EPS growth — each ranked against peers. Crucially, the plan rewards returns and relative performance, not raw revenue or unit growth: the empire-building flag is absent, and there is no volume-only growth metric.

The honest weaknesses. (1) Buying back stock above book and near multi-year highs. H1-FY26 repurchases averaged ~$154/sh ≈ ~1.8–1.9x book, versus the ~1.0–1.5x of 2022–23 — far less accretive, and embedding reinvestment risk if the cycle rolls over (DHI is a steady, heavy repurchaser, not a contrarian buy-low one). (2) Rental detour: a ~$3B rental platform built near the cycle peak and now being unwound — capital that wandered into a lower-return, longer-duration adjacency. (3) Forestar minority leakage: consolidating a 62%-owned listed vehicle is coherent with the land-light thesis but adds complexity and a ~38% economic minority interest.

Verdict: Management has allocated capital intelligently — close to best-in-class for a cyclical. Low leverage, ~8%/yr float shrinkage funded by real OCF, a low/fast-growing/covered dividend, a capital-light land strategy that structurally lifts ROIC, and a returns-and-relative-TSR incentive plan. The legitimate weaknesses — buybacks at ~1.9x book near a cyclical earnings high, and the partly-reversed rental detour — temper but do not overturn the verdict.

8. Changes and Headwinds — Last Two Years

1. Founder Donald R. Horton died — May 2024 (the defining change). The single most material change is the passing of founder/Chairman Donald R. (“DR”) Horton in May 2024. DR had already ceased serving as an executive officer effective October 1, 2023, remaining Chairman and Strategic Advisor; in that same transition David Auld moved to Executive Vice Chair and Paul Romanowski was promoted to CEO and appointed a director, so succession was pre-planned and the operational transition orderly. Governance/ownership consequences: the Horton Family Limited Partnership and affiliates hold ~6.91% (DR’s sons Ryan and Reagan Horton relate to ~7.4% as co-trustees); top institutions are Vanguard ~12%, Capital World ~10.5%, BlackRock ~9.8%. DR’s sons’ “R&R” entities now act as a related-party land seller/banker to DHI, but the volumes are de minimis (FY25: 54 acres for $5.4M plus a ~$1.1M accrual fee, trivial against $5.9B of total FY25 land/lot purchases). The real consequence is the loss of founder ownership alignment — all officers and directors now hold just ~0.66% (CEO Romanowski <0.1%).

2. CEO/Chairman succession completed; board refreshed. Paul Romanowski (CEO since Oct-2023; FY25 total comp ~$16.3M), Michael Murray (COO), and Bill Wheat (CFO) form a stable, long-tenured operating team; David Auld is Executive Chairman. Three new independent directors were added in August 2024 (including Barbara R. Smith, former Chairman/CEO of Commercial Metals), strengthening board independence post-founder.

3. Strategy shifts. Deeper land-light/Forestar integration ($26B lots controlled via options; 67% of closings on lots developed by others) — strengthens the ROIC story; rental rationalization (trimming the ~$3B platform) — a small cleanup of a peak-cycle detour; escalating rate-buydown/incentive intensity to defend volume and take share in a soft market — pressures margin but defends closings; market-share aggregation as weaker builders retrench — the classic strongest-builder downturn playbook.

4. Berkshire Hathaway homebuilder stake (2024). Berkshire disclosed homebuilder positions (including DHI, LEN, NVR) in 2024 via 13F — a sentiment/validation data point only (later trimmed), not thesis-determining and not evidence of any particular investor’s position.

Headwinds. The dominant headwind is the affordability ceiling at ~6.5–7% mortgages, which forces the buydown spending that is compressing gross margin and is the question mark over whether the ~20% margin is a cyclical trough or a structural reset. Secondary: soft spring selling season, ‑14% backlog, negative operating leverage on a falling ASP, and the macro/rate uncertainty management itself flags.

Verdict: Net thesis-NEUTRAL to mildly strengthening. The founder’s death was culturally significant but operationally well-managed; the board is more independent; the land-light deepening and disciplined capital return strengthen the structural ROIC story. The genuine new watch-items are the loss of founder ownership alignment, the (immaterial) related-party land channel, and the margin pressure from escalating incentives — a cyclical, not structural, headwind on current evidence.

9. Risk Analysis

Catastrophic-loss / total-loss risk: Very low. A fortress balance sheet (net debt ~$3B, ~$6B liquidity), land-light flexibility to throttle starts, and clean accounting make a solvency event implausible absent a 2008-scale housing depression — and even then DHI would be a survivor/share-taker, not a casualty. The realistic downside is valuation (a de-rate toward ~1.3–1.4x book on a deeper/longer downturn), not impairment of the enterprise.

10. Valuation Discussion (Embedded Expectations)

Read the cyclical-multiple inversion first. For a homebuilder, P/E is a trap near the cycle inflection. Builders look cheapest on P/E at the peak (high E, low multiple) and most expensive at the trough (low E, high multiple). DHI’s own history makes it vivid:

So the own-history valuation percentiles (P/E 94.7th, P/S 88.4th, P/B 63.2nd, composite 82.1st) must be read with the inversion in mind. The P/E 94.7th-percentile reading is misleadingly rich — the numerator (price) is mid-range; the rising ratio is driven by cyclically declining EPS. The P/B 63.2nd percentile (~1.9x) is the better cyclical anchor: versus a ~1.0x trough to ~2.4–2.9x froth range, DHI is above mid-cycle but well below peak-froth — full but not extreme. P/S is elevated partly because post-COVID net margins (the sales the market capitalizes) are structurally higher than the pre-2020 base.

Comp set (most-recent-fiscal-year-end basis; relative ranking holds).

DHI is neither the cheapest nor the most expensive. On P/B it sits mid-pack (above KBH/NVR-book, near LEN/TOL, below PHM); on EV/EBITDA it is at the top of the group (~12x) alongside Lennar — the market pays a scale/quality premium for the two largest builders over PHM/TOL/KBH (~7–10x). The premium is defensible on #1 scale, the most aggressive buyback, and the lowest-cost/highest-turns model — but it is a premium, so the margin of safety on price is thin. (OPEN QUESTION: is DHI’s EV/EBITDA premium to PHM — which has better returns and a cheaper multiple — justified by durability, or is it Berkshire-halo plus index-weight inertia?)

Embedded expectations — what is ~$158 / ~1.9x book / ~14.7x trough-ish E pricing? At ~$43B market cap / ~$46B EV:

• Earnings: the market is not pricing a return to FY22 peak EPS ($16.51) — that world puts the stock far higher — nor a deep downturn (which would compress P/B toward ~1.2–1.4x). It is pricing a mid-cycle normalization: roughly $10–12 of sustainable EPS (around the FY26 guide) at a normal ~14–16x, with the buyback doing the heavy lifting on per-share growth.
• Margins: embedded is an assumption that gross margin stabilizes near ~19.5–20% (down from ~26%+ peak) rather than collapsing further — i.e., that buydowns are a cyclical drag, not a permanent reset. This is the crux assumption.
• Returns: at ~1.9x book the market underwrites a sustainable forward ROE comfortably above cost of equity (a ~13–16% ROE is roughly consistent with ~1.9x book). If ROE settles at ~12–13% the multiple is fair-to-slightly-full; high-teens makes it cheap; ~9–10% makes it too high.
• The buyback is the swing factor. ~8%/yr float shrinkage means even flat aggregate net income produces ~8% EPS growth — the market is partly paying for the compounding, not just operating earnings. The risk: repurchasing at ~1.9x book is modestly accretive, far less than the sub-1.5x repurchases of 2022–23.

Scenario zones (embedded-expectations, NOT price targets):

• BEAR — affordability ceiling holds, margins reset lower: mortgages stay 6.5–7%+, incentives prove structural, EPS drifts to a trough ~$7–8, and the multiple de-rates toward ~1.3–1.4x book as the market re-prices builders as deep cyclicals. A meaningfully lower zone than today.
• BASE — mid-cycle normalization: EPS ~$10–12, gross margin stabilizes ~19.5–20%, multiple holds ~1.7–1.9x book / ~14–16x normalized E; buyback drives mid-single-digit-plus EPS growth. Roughly where the stock sits — the base case is largely already priced.
• BULL — rate relief re-accelerates volume: the 30-yr mortgage falls toward ~5.5%, incentive intensity drops, margins re-expand toward ~21–22%, EPS recovers to $13+, and the multiple re-rates to ~2.0–2.2x book on a perceived new up-leg plus the Berkshire halo.

Net (no rec): DHI is a quality cyclical priced at a full-but-not-extreme level on the right (P/B) anchor, with the “expensive P/E” headline being a cyclical artifact. The market underwrites the base case; the margin of safety on price is thin, and the buyback — not multiple expansion — is the most reliable per-share-value lever. The asymmetry at ~$158 is roughly symmetric-to-slightly-unfavorable: the bull needs both rate relief and margin re-expansion; the bear needs only the affordability ceiling to persist.

11. Variant Perception

Consensus. DHI is the best-in-class, largest, lowest-cost US homebuilder — a “quality cyclical” with a land-lighter, high-turns, entry-level model, a fortress balance sheet, a disciplined ~8%/yr buyback, and a Berkshire endorsement. The narrative: “the earnings trough is in, structural housing undersupply is a multi-year demographic tailwind, and rate relief is a call option on an earnings inflection.” A defensive-ish homebuilder — and the factor tape agrees (beta 0.77).

Strongest bull case. (1) Structural undersupply — the US is ~3–5M units short after a decade of post-GFC underbuilding; millennial/Gen-Z household formation is a secular demand floor. (2) Land-light, high-ROIC model — options rather than owns land, lifting turns and capital efficiency; ROIC ~10.8% is spread-positive even at a trough. (3) Relentless buyback shrinking the float — ~8%/yr means EPS compounds even with flat net income; over a decade DHI has quietly become a per-share compounding machine. (4) Rate relief = double benefit — every 50–100bps off the mortgage disproportionately restores entry-level affordability, reducing incentive spend and re-expanding margins while lifting volume.

Strongest bear case. (1) Peak-cycle earnings/margins still mean-reverting down — the EPS arc ($16.51 → $11.57 → ~$10.76 TTM, guide ~$10–11) is still descending, and “the trough is in” has been asserted prematurely before. (2) Affordability ceiling at 6–7% mortgages — the buydown machine masks a real demand problem; DHI is renting volume via forward commitments that structurally suppress margin. (3) Buying back stock no longer cheap on book — ~1.9x repurchases are far less accretive than the ~1.0–1.5x of 2022–23. (4) Incentive-driven margin erosion may be structural — if 6%+ mortgages are the new normal, ~20% gross margin is the new ceiling, not a trough, and the market is over-capitalizing it. (5) The P/E inversion cuts both ways — “cheap 14x” sits on trough-ish earnings.

The assumptions that matter most (and what falsifies each side):

1. Are ~20% gross margins a cyclical trough or a structural reset? Falsifies bull: margins slide below ~19% with incentive spend rising as a % of price even as volume holds. Falsifies bear: margins re-expand toward 21–22% as rates ease and incentive intensity falls.
2. Is the EPS trough actually in? Falsifies bull: FY26/27 EPS prints below ~$10 with declining orders. Falsifies bear: EPS stabilizes ~$10–11 and order growth turns durably positive.
3. Does rate relief materialize and flow through? Falsifies bull: mortgages stay 6.5%+ through 2026–27. Falsifies bear: the 30-yr falls toward ~5.5% and both absorption and margin improve.
4. Is the buyback still value-accretive? Falsifies bull: management keeps repurchasing at 2x+ book into a deteriorating return profile. Falsifies bear: book value/share compounds and per-share metrics outrun aggregate stagnation.

Factor-positioning read. DHI carries a dominant Industry: Home Construction loading of ~2.30 with SmallSize ~0.9–1.0 and Market ~0.80 — i.e., overwhelmingly a housing-beta vehicle, not an idiosyncratic-alpha name. Beta is a low 0.77 and alpha is slightly negative (‑0.03) — the recent strength (y1 +31.6%, Sharpe 0.76; latest quarter ~+19% raw) is sector beta, not stock-specific outperformance. Factor twins are the entire group (PHM 0.98, LEN 0.97, KBH 0.96, TOL 0.95, NVR 0.94, plus ITB/XHB ETFs). The low beta + negative alpha says the market is treating builders as defensive-ish lately — a regime quirk, since builders are deep cyclicals (lifetime max drawdown ‑85% in the GFC) — which is itself a hint that consensus may be offsides on how cyclical the earnings still are. rs_12m +33% but rs_peak ‑18% (off the high): a strong-trend name that has not reclaimed its peak — momentum present but not extended.

Where consensus may be wrong (the variant). The crowd reads “14x earnings, Berkshire owns it, housing shortage” as cheap quality. The variant read is that (a) the 14x is on trough-ish earnings (P/E inversion), (b) the only sane cyclical anchor — P/B at ~1.9x — is above mid-cycle, © the margin reset may be structural not cyclical, and (d) the buyback is now far less accretive than the market remembers. None of this makes DHI a short (spread-positive ROIC, fortress balance sheet, real undersupply), but it argues the easy money has been made, and the name is now a housing-rate macro bet priced near fair value, masquerading as a margin-of-safety compounder.

12. Fact vs. Interpretation Table

13. Open Questions

1. Margin: cyclical trough or structural reset? Is the ~20% home-sales gross margin the bottom of a cycle, or the new ceiling at a 6%+ mortgage regime? This single question drives the bull/bear split.
2. Is the EPS trough actually in, or do FY26/FY27 prints come in below ~$10 as orders/ASP keep softening?
3. Rate path: does the 30-yr mortgage ease toward ~5.5–6% (restoring entry-level affordability and relieving the incentive drag), or stay 6.5%+ through 2026–27?
4. Buyback discipline at ~1.9x book: will management keep repurchasing heavily into a deteriorating return profile, or pivot to opportunistic timing?
5. Forestar/related-party governance: does the R&R land channel grow from immaterial, and is the 62%-owned consolidated structure the right long-run form?
6. Is DHI’s EV/EBITDA premium to PHM (which has better returns and a cheaper multiple) justified by durability — or Berkshire-halo + index inertia?

14. What Must Be True

For the BULL case (DHI compounds from here):

• The ~20% gross margin is a cyclical trough and re-expands toward 21–22% as rates ease and incentive intensity falls. Falsification test: if, over the next 3–4 quarters, gross margin slides below ~19% with incentive spend rising as a share of price even while volume holds, the margin reset is structural and the bull case breaks.
• The EPS trough is in (~$10–11) and order growth + community count convert into per-share earnings growth, aided by the ~8%/yr buyback. Falsification test: two-plus consecutive quarters of declining orders with EPS printing below ~$10.
• Rate relief materializes enough to restore entry-level affordability. Falsification test: the 30-yr mortgage holds ≥6.5% through 2026–27.

For the BEAR case (DHI de-rates / earnings keep falling):

• The affordability ceiling persists at 6–7% mortgages, buydowns prove structural, and EPS drifts toward a ~$7–8 trough while the multiple de-rates toward ~1.3–1.4x book. Falsification test: gross margin re-expands toward 21–22% with falling incentive intensity, and order growth turns durably positive — the cyclical-trough thesis, confirmed.
• The buyback at ~1.9x book proves value-neutral-to-destructive across the cycle. Falsification test: book value/share compounds and per-share metrics demonstrably outrun aggregate net-income stagnation over a full cycle.

APPENDIX A — Standard Diligence Questionnaire

Supplemental diligence questionnaire. Fact / Interpretation / Assumption labels where it matters.

General

What thoughtful questions have other investors asked about this company? Whether the post-COVID gross-margin reset (31%→20%) is cyclical or structural; whether “the earnings trough is in”; whether buying back stock at ~1.9x book is still accretive; how durable the land-light ROIC advantage is once every peer copies it; what Berkshire saw in 2024 (and what its trimming implies); and whether the founder’s death changes the culture/discipline. The central debate is margin durability under a 6%+ mortgage regime.

Cyclicality & Earnings Nature

Cyclical high or low? Past the high and grinding lower — EPS $16.51 (FY22 peak) → $11.57 (FY25) → ~$10.76 TTM; FY26 guide ~$10–11. Margins and returns compressing. (FACT.) External or internal drivers? Overwhelmingly external — mortgage rates, affordability, household formation. DHI’s internal actions (land-light, buyback, share gain) optimize within an exogenously-set demand envelope. (INTERPRETATION.) Revenue stability? Low — homebuilding is deeply cyclical; revenue plateaued FY23–24 and fell ~7% in FY25; only a few months of backlog visibility (10,785 homes / $4.1B, ‑14% YoY). Product/market outlook & size? Large and structurally under-supplied long-run (~3–5M unit deficit), but demand is rate-gated near-term. US single-family starts ~941K (2025), ~17% below the 2021 peak. Domestic-only; no international.

Business Quality & Competitive Moat

Industry more or less competitive? Consolidating to scaled publics (DHI #1) — less fragmented over time, but still contestable community-by-community; the product stays a commodity. (FACT/INTERPRETATION.) Business profitability (ROIC/ROE)? Cyclically high-to-now-moderate: ROE ~36% (FY22) → ~12–15% (FY25); ROIC 25% → 10.8%; homebuilding pretax ROI on inventory ~17.6–20%. Spread-positive even at trough. (FACT.) Industry profitability / barriers? Mediocre through-cycle (returns mean-revert hard); barriers are supply/scale/land-based, not demand-based. Entry-level tier is the most affordability-sensitive. Easily understood? Yes — build affordable homes, finance the buyer, return cash. Undermined by foreign low-cost labor? No — homes are built on-site locally; not tradeable/importable. Input materials (lumber, etc.) and tariffs are a cost factor. Do brands matter? Minimal. The FY25 10-K dropped the sub-brands (Express/Emerald/Freedom) — they were price-point labels, not a demand moat. (FACT/INTERPRETATION.) Nature of competition? Price/affordability and location at the community level; scale players compete on cost and land access. Switching costs? Effectively zero — a home is the most infrequent considered purchase; no captivity.

Financial Condition & Balance Sheet

Assets not fully on the balance sheet? The ~77% optioned lot pipeline (controlled, not owned) is the key off-balance-sheet control; the value of the land-acquisition platform and subcontractor relationships is intangible/unbooked. Off-balance-sheet liabilities? Lot-option deposits (largely non-recourse, limiting exposure); mortgage-warehouse facilities (financial services). Generally modest and disclosed. How conservative is the accounting? Clean — minimal impairments (land-light by design), small SBC (~0.4% of revenue), trivial capex. Watch warranty/litigation reserves. (FACT/INTERPRETATION.) CapEx-hungry? No true PP&E capex (net fixed assets only ~$86M; ~$100M/yr D&A). The “capex” is inventory — land/lot/WIP — which is large ($25.3B) but cyclically harvestable. OCF ≈ FCF.

Capital Allocation & Management

FCF generation & use? FY25 OCF ~$3.42B (~$11/share) ≈ FCF; all of it returned to shareholders ($4.35B buyback + ~$0.49B dividends, funded partly by the cash harvest). Philosophy: low leverage, low/fast-growing dividend (~13–14% payout), bulk of returns via buyback. FY26 guide: ~$2.5B buyback, ~$500M dividends, OCF ≥$3B. (FACT.) Significant acquisitions? No transformational M&A; small bolt-on land/builder deals ($40–270M/yr). Forestar (62%-owned) is the captive land developer. The ~$3B rental platform (built near peak) is being unwound. (FACT.) Buying back shares? Aggressively — ~8%/yr share-count shrinkage (324M→294M FY25). Caveat: now at ~1.9x book vs. ~1.0–1.5x in 2022–23 (less accretive). (FACT/INTERPRETATION.) Issuing shares to insiders? Routine equity comp (small SBC); no large dilutive issuance. Director/management compensation? CEO Romanowski FY25 ~$16.3M. Incentives anchored on pre-tax income (capped), relative TSR vs. S&P 500, ROA/pre-tax ROA vs. peers, EPS growth — returns-based, no volume/empire metric. (FACT.) Management motivations? Returns-and-discipline-aligned by plan design, but skin-in-the-game is now low (all insiders ~0.66%; CEO <0.1%) post-founder. Zero open-market insider buys.

Valuation & Market Data

ADR / MLP / K-1? No — a US C-corp common stock (NYSE: DHI; also NYSE Texas since 2025). Standard 1099 dividend. Dividend policy? $0.45/qtr ($1.80 annualized), ~1.1% yield, ~13–14% payout, grown steadily from $0.70/yr (FY20). Easily covered, large room to grow. Profitability? Cyclically moderate now (net margin ~10.5% FY25, off ~17.5% FY22 peak); spread-positive returns. Net income vs. cash from operations diverging? Yes, by design — homebuilder OCF is inventory-cycle-driven and inversely tracks earnings (weak when building inventory in the boom, strong when harvesting on the way down). Not a red flag; understand the cycle. (FACT/INTERPRETATION.)

Risks & Downside

What would cause the stock to decline? Higher-for-longer mortgage rates; a structural (not cyclical) margin reset; recession/unemployment; EPS prints below ~$10; multiple de-rate toward ~1.3–1.4x book. Catastrophic-loss risk? Very low — fortress balance sheet (net debt ~$3B, ~$6B liquidity), land-light flexibility, clean accounting; DHI would be a downturn survivor/share-taker. Total-loss risk? Negligible absent a 2008-scale housing depression — and even then DHI is a survivor, not a casualty. The realistic downside is valuation, not solvency.

Recent News & Events

Business environment changed recently? Yes — affordability ceiling at ~6.5–7% mortgages forcing escalating rate-buydowns; soft spring 2026 selling season; backlog ‑14%; FY26 guide implies continued grind. (FACT.) Significant acquisitions / accounting changes? None material; rental rationalization ongoing. Recent management/strategy changes? Founder/Chairman Donald R. Horton died May 2024 (succession pre-planned: Romanowski CEO since Oct-2023, Auld Executive Chairman); 3 new independent directors Aug-2024; deeper land-light/Forestar; NYSE Texas dual-listing (Jun-2025); Berkshire homebuilder stake disclosed 2024 (later trimmed).

APPENDIX B — Source Appendix

Report date 2026-06-19. Primary sources prioritized. Third-party aggregated data (third-party aggregators and a quantitative factor model) reconciled to filings; treated as cross-checks, not primary.

Primary — SEC Filings (EDGAR, CIK 0000882184)

• FY2025 Form 10-K (period ended 2025-09-30; filed 2025-11-19) — Item 1 Business (markets/states, brand/tier structure, land strategy, Forestar, mortgage capture, subcontracted construction); Item 7 MD&A (segment revenue/margins, gross-margin bridge, incentives/buydowns, backlog). https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0000882184
• Form 10-Q Q1-FY2026 (period ended 2025-12-31; filed 2026-01-22).
• Form 10-Q Q2-FY2026 (period ended 2026-03-31; filed 2026-04-23) — repurchases (10.4M shares / $1.6B H1; $1.7B remaining), leverage 21.7%, book value/sh $82.91, Forestar 62%-owned/38% NCI, liquidity/maturities.
• FY2021–FY2024 Form 10-Ks (filed 2021-11-18, 2022-11-18, 2023-11-17, 2024-11-19) — multi-year revenue/EPS/margin trend; FY22 brand-architecture diff.
• DEF 14A proxy (filed 2025-12-10) — CEO letter (FY25 OCF $3.4B all returned; distributions +118%; ROE 14.6%, ROA 10.0%, pretax ROI on inventory 20.1%; top-20% S&P 500); executive comp (Romanowski ~$16.3M); PSU metrics (pre-tax income, relative TSR vs S&P 500, ROA/pretax-ROA vs peers, EPS growth); beneficial ownership (Horton Family LP 6.91%; insiders 0.66%; Vanguard 11.97%, Capital World 10.46%, BlackRock 9.76%); R&R related-party land transactions.
• DEF 14A proxy (filed 2024-12-13) — passing of founder Donald R. Horton (May 2024); Auld → Executive Chairman; Oct-2023 Romanowski CEO transition.
• Form 8-Ks (2024-2026) — earnings (2.02), board refresh 2024-08-28 (5.02/5.03; Barbara R. Smith), financings/notes (1.01/2.03), annual-meeting results (5.07), mortgage-warehouse facility amendment (2026-05-12).
• Form 3/4/5 corpus (~416 Form 4 + 4 Form 4/A + 4 Form 3, trailing ~5 yrs) — zero open-market purchases (code P); recent activity 10b5-1-planned sales (~$144–157) + routine vesting (M/F/A/G). Donald R. Horton individual Form 4s (CIK 0000900824) through Apr-2024.

Primary — Company Disclosure

• Q2-FY2026 earnings call transcript (2026-04-21) — pretax income $867M, revenue $7.6B, pretax margin 11.5%, diluted EPS $2.24 (vs $2.58 PY); orders +11% YoY / 24,992; cancel rate 16%; ASP $361,600 (~30% below US new-home avg; ~$70K below median existing); 65% first-time buyers; home-sales GM 20.1% (19.7% normalized); FY26 guidance (rev $33.5–34.5B, 86–87.5K homes, OCF ≥$3B, buyback ~$2.5B, div ~$500M, tax 24.5%); 575K lots (23% owned/77% controlled); 67% of closings on Forestar/3rd-party lots; rental $3B; financial services pretax margin 26.8%; Forestar 94K lots. (Company earnings call; management commentary validated against filings.)
• D.R. Horton Investor Relations — investor.drhorton.com (supplemental data, investor presentation).

Quantitative Cross-Checks (third-party; reconciled to filings)

• Aggregated fundamentals (third-party) — income statement, balance sheet, cash flow (FY2020–FY2025); profitability ratios (ROE/ROA/ROIC, margins); enterprise value (EV ~$46B today; FY25 EV/EBITDA 12.3x); valuation multiples (P/E, P/B, EV/EBITDA trend); per-share data; comp multiples for LEN/PHM/NVR/KBH/TOL. Note: one third-party book value/share field appeared erroneous (~$100.6 FY25); reconciled to filing equity ex-minority $24.19B/294.5M ≈ $82, matching management’s $82.91 and AZI’s $81.75.
• Own-history valuation percentiles (third-party aggregator) — P/E 94.7th, P/B 63.2nd, P/S 88.4th, composite 82.1st; latest price $157.81, ttm_eps $10.76, book $81.75/sh, P/E 14.67x, P/B 1.93x, P/S 1.39x (as of 2026-06-18).
• Price history (market data) — 5-yr OHLCV for the price event map: 5-yr high $195.88 (2024-09-19), 5-yr low $56.87 (2022-06-17), 52-wk high $182.90 (2025-09-08), 52-wk low ~$120 (2025-06-20), close $157.81 (2026-06-18).
• Financial news — sector coverage incl. Berkshire’s homebuilder positioning and homebuilder earnings roundups.
• Factor model (third-party) — stock-info (beta 0.77, alpha ‑0.03, market cap $43.2B, rs_12m +33%, rs_peak ‑18%); loadings (Industry: Home Construction ~2.30 dominant, SmallSize ~0.9–1.0, Market ~0.80; R² ~0.84); leaderboard (y1 +31.6%/Sharpe 0.76, m6 +9%, lifetime max DD ‑85%, y5 max DD ‑44%); related stocks (PHM 0.98, LEN 0.97, KBH 0.96, TOL 0.95, NVR 0.94, ITB/XHB, BLDR 0.88, FOR, IBP).

Industry Context (public sources)

• Builders FirstSource (BLDR) public disclosures and US housing data — single-family starts ~941K (2025, ‑17% vs 2021 peak); builder consolidation favoring scaled publics; spec-inventory destocking; DHI named a top production-builder customer.
• Home Depot (HD) public disclosures and NAR/Freddie Mac data — existing-home sales ~4.0M SAAR (30-yr low); turnover ~2.9% (~40-yr low); ~80% of mortgages below current rate (rate lock-in).
• Lowe’s (LOW) and Sherwin-Williams (SHW) public disclosures — housing-demand / repair-remodel backdrop.

Frameworks

• Greenwald & Kahn, Competition Demystified — barriers to entry; economies of scale without customer captivity; market-share-stability & ROIC tests.
• Chancellor (Marathon), Capital Returns — supply-side capital-cycle analysis; the managed/dampened down-cycle in homebuilding.

Note on third-party estimates: aggregated data and statistical factor estimates are not primary sources. For US-filer facts, EDGAR filings are primary and govern where any discrepancy arises.