PulteGroup: A Good Capital Allocator, Priced for a Boom That's Ending
Before the Gates
PulteGroup is one of the larger American homebuilders, and the question this piece examines is the one the gates are built for: is it a good business run by good people, available at a good price? The first two answers are mostly yes. The third is where the thesis stalls.
Gate 1 — Circle of Competence
The business is easy to describe. PulteGroup buys or options land, develops it, and sells newly built homes across three brands aimed at different life stages: Centex for first-time buyers, Pulte Homes for move-up demand, and Del Webb for 55+ active-adult communities. Revenue is recognized mostly at closing, with a mortgage subsidiary contributing on the side. In 2025 the company delivered 29,572 homes for $16.7 billion in home sale revenue and $2.2 billion in net income. Profitability rests on two levers: what it paid for the land (the largest cost variable) and how it manages prices and incentives. Around 244,000 lots are controlled, 59% of them via options — a land-light structure that trims balance-sheet risk.
Information is fully available: 10-Ks, quarterly calls, lot positions, net new orders, incentive levels — everything a cycle analyst needs. Predictability is the conditional part. Demand for new American homes a decade out is analyzable: structural undersupply, mortgage lock-in among existing owners pushing buyers to new builds, and boomer aging favoring Del Webb. But which year earns what is hostage to rates and the cycle. Ten-year average earnings are analyzable; next year’s earnings are a probabilistic bet. Any valuation must therefore rest on normalized, mid-cycle earnings — the moment you multiply current earnings, you have left the circle. This is a borderline case by design: land and capital barriers qualify it as a stable industry, but the cycle keeps it on the edge.
Gate 2 — Moat
Two sources of advantage are identifiable. First, a partial intangible: Del Webb, the dominant brand in active-adult communities for decades, uncopied by competitors, riding demographic tailwinds, with buyers who are less rate-sensitive (a high share pay cash). Second, scale-based cost advantage in land acquisition, materials, and mortgage attachment, with market share consolidating toward large public builders post-pandemic.
Pricing power, though, is weak — and that is the fundamental limitation of the business. In the 2025–2026 demand slowdown the industry has leaned on incentives exceeding 5% of orders and price cuts; PulteGroup is no exception. Q1 2026 gross margin fell to 24.4% from 27.5%, with ASP down 5%. Can it raise prices without praying? Only in an up-cycle. Only Del Webb shows relative price defense.
The moat’s direction is steady to slowly widening — share concentration among large builders, land-light survivability, demographic tailwinds — but its absolute width is narrow. Houses are close to commodities; differentiation comes from land location and cost discipline. Verdict: a weak pass, with the narrowness to be reflected in the discount rate at Gate 4.
Gate 3 — Management
This is the company’s strongest gate. Under CEO Ryan Marshall (since 2016), capital allocation follows an explicit order: reinvest in high-return projects, dividends, buybacks with excess cash, debt repayment. Having reflected on the post-crisis era of top-line revenue but subpar ROIC, management chose returns over volume — and executed. Since 2013, more than 69% of shares issued have been retired, an industry-leading record. In 2025 alone, 10.6 million shares (5.2% of the float) were bought back for $1.2 billion at an average of $112.76 per share. From 2021–2025, over $25 billion went into land while $6.2 billion returned to shareholders. Debt-to-equity stands at 11.2%; on a net basis, −3.0% — net cash, which in a cyclical industry is precisely the Buffett standard.
The retained-earnings test passes: the stock is up 6.5x under Marshall, with five-year average ROE of 24.6% (standard deviation a stable 4.5 points). Candor checks out — margin pressure and incentive expansion are disclosed in numbers, Q2 2026 margin guidance was explicitly framed as the trough of the year, and the company is expanding build-to-order share rather than defending volume into a downturn. Verdict: pass.
Gate 4 — Price
Owner earnings for 2025 are $2.21 billion — roughly net income, since maintenance capex is trivial in homebuilding. But this is the trap: $2.2 billion is cycle-peak profit. EPS ran $7.4 → $11.0 → $11.7 → $14.7 from 2021 to 2024 on the pandemic housing boom; 2025 declined year over year; Q1 2026 net income fell 34%; margins are normalizing industry-wide. Applying the historical mid-cycle gross margin of 20–22% implies a 30–35% earnings haircut. Normalized owner earnings: $1.4–1.6 billion.
A cyclical business with a narrow moat earns a low multiple — 10–12x — on normalized earnings. That yields $14 billion at the floor and $19.2 billion at the top, centered around $16–17 billion. The market cap is $23.9 billion, with a superficial P/E of 12.05. That 12x is an illusion built on peak earnings; on normalized profit the real multiple is 15–17x. The headline owner-earnings yield of 9.2% looks generous against a 4.3% ten-year yield, but on normalized earnings it compresses to 5.9–6.7%. The current price sits roughly 40–45% above the center of intrinsic value. There is no margin of safety.
The target buy price: a market cap of $11–12 billion, about $56–60 per share — a 30% discount to normalized intrinsic value. Comparable valuations were reached in the 2022 and 2024 drawdowns, so it is reachable in a deepening downturn. Verdict: fail, on price.
What Would Change the Verdict
The falsification scenarios run both ways. If structural housing scarcity plus falling rates make today’s ~24% margins the new normal — and $2.2 billion the true mid-cycle figure — then the current price is fair to cheap, and waiting is an omission risk. Conversely, if the downturn deepens, option abandonment fees and inventory writedowns could push earnings below the $1.4 billion floor, and a structural incentive war would make the 20–22% margin assumption itself optimistic. The monitors: gross margin (below 20% for two quarters forces a downward revision; 26%+ recovery in a falling-rate environment triggers an upside review), net new orders (Q1 2026 was +3%, still sound; double-digit declines signal deepening), incentive rates (5%+ and climbing erodes pricing power), the land-light option share, and the discipline of buyback pricing.
Verdict
Watch. Gates 1 through 3 pass — management is among the best in the industry — but a business earning cycle-peak profits, with a narrow moat and cycle-dependent pricing power, is wearing a deceptively cheap 12x multiple. At $23.9 billion against a normalized center of $16–17 billion, this is the classic trap of normal multiples on peak earnings. The watch becomes interesting if a deepening downcycle delivers the fear: a market cap near $11–12 billion, around $56–60 per share, would pair excellent capital allocation with a genuine margin of safety.
This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer