PulteGroup and the Price of Cyclical Earnings
What the business is
A house may be built on a foundation, but a homebuilder’s economics rest largely on the land beneath it. Buy the land sensibly, control inventory, and maintain discipline through the cycle, and the structure can endure. Overpay for dirt and chase volume, and even the finest kitchen counters offer little comfort.
PulteGroup is among the three largest US homebuilders. It acquires or controls land, develops communities, builds homes, and recognizes most revenue when those homes close. Its three principal brands address different stages of the buyer’s life: Centex serves first-time buyers, Pulte Homes targets move-up demand, and Del Webb focuses on active-adult communities for buyers aged 55 and older. A mortgage-finance subsidiary provides ancillary income.
In 2025, the company delivered 29,572 homes and generated $16.7 billion of housing revenue. The two central profit levers are straightforward: what PulteGroup pays to control land, and how effectively it manages home prices and buyer incentives.
Its land-light approach offers some protection from the industry’s cyclicality. PulteGroup controlled roughly 244,000 lots, with 59% held through option contracts rather than outright ownership. Options can reduce the balance-sheet consequences of a downturn, although they do not make poor demand disappear. They merely allow the company to encounter it wearing a lighter backpack.
The moat is real but narrow. Scale helps with land acquisition, materials purchasing, and mortgage integration, while Del Webb has a differentiated position supported by an aging population and buyers who are relatively less sensitive to mortgage rates. Large builders have also been gaining share.
Still, housing remains close to a commodity. Pricing power weakens when demand slows, and the industry responds with incentives and price reductions. PulteGroup’s first-quarter 2026 gross margin fell to 24.4% from 27.5%, while average selling prices declined 5%. I therefore view the moat as holding or widening modestly, helped by scale and land discipline, but not becoming broad. Artificial intelligence may improve design or land analysis, yet those tools are likely to spread across the industry rather than create a durable advantage.
What the numbers say
Reported annual owner earnings are approximately $2.21 billion. That is substantial earning power, but the thesis treats it as a high-cycle figure rather than a dependable mid-cycle base. Housing profitability remains tied to rates, incentives, land costs, and the economic cycle. Applying an ordinary multiple to unusually strong earnings is an old way to make a cyclical business look cheap.
The current valuation card is:
- Market cap (price-adjusted, 2026-09-23 close): $22.33 billion
- Owner-earnings yield (price-adjusted): 9.89%
- Share price (2026-09-23 close): $118.71
- 10-year Treasury: 5.11%
The resulting spread is +4.78 percentage points. On the surface, a 9.89% owner-earnings yield against a 5.11% Treasury is attractive enough to merit attention. The difficulty lies in the numerator: if current owner earnings represent a favorable part of the housing cycle, the apparent yield overstates normalized earning power.
The mechanical valuation record supplied with the screen lists an intrinsic-value range of $1.4–$1.6 billion. Against that range, the price-adjusted market capitalization stands at a +1495% premium to the low end and a +1295% premium to the high end. Whatever generosity one grants the current earnings, those figures do not describe a margin of safety.
Capital allocation is the brighter part of the ledger. Since 2013, PulteGroup has retired more than 69% of its issued shares. In 2025 it repurchased 10.6 million shares, equal to 5.2% of shares outstanding, for $1.2 billion at an average price of $112.76. The company also invested more than $25 billion in land during 2021–2025 while returning $6.2 billion to shareholders. Its five-year average return on equity was 24.6%.
That record suggests management understands per-share value and balance-sheet resilience. It does not, however, make the purchase price ceremonial. Good capital allocators can create value for years; outside shareholders can still pay too much to participate.
What Wall Street believes vs what I see
The available headlines convey a cautious but not despairing mood. One emphasizes earnings durability and buybacks amid housing headwinds. Another says housing legislation helps without solving affordability. A third asks whether the second quarter can prove that the worst is over. The common theme is that PulteGroup is a capable operator facing a difficult environment rather than a broken business.
I largely agree on the business and remain less comfortable with the price. The market appears willing to treat recent owner earnings as reasonably durable, supported by structural housing undersupply, existing homeowners’ mortgage lock-in, and increasing share for large builders. Those forces could make today’s profitability more persistent than older housing cycles would suggest.
But the operating evidence still shows weak pricing power when demand cools. Incentives across the industry have exceeded 5% of order value, and PulteGroup’s margin contraction demonstrates that scale does not repeal affordability. The company can manage the cycle better than smaller competitors without escaping the cycle.
The owner-earnings arithmetic therefore deserves two readings. The first is the visible 9.89% yield, comfortably above the 5.11% Treasury. The second asks whether the earnings supporting that yield are normal. The thesis says they are not, and the absence of a valuation discount makes that distinction consequential. A superficially modest earnings multiple is not necessarily a bargain when profits are doing more than their mid-cycle share of the work.
Verdict
Verdict: watch.
The 9.89% owner-earnings yield exceeds the 5.11% Treasury, satisfying the rubric’s watch condition. Circle is unavailable because the $22.33 billion market capitalization is nowhere near the required discount to the intrinsic range; it instead carries a +1495% premium to the low end. PulteGroup’s land discipline, buyback history, and 24.6% five-year average return on equity make the company worth following, but cyclical earnings and weak pricing power argue against treating the headline yield as permanent.
I would change the verdict if the market capitalization moved into the rubric’s required discount territory, or if sustained operating evidence showed that current earning power had become a defensible mid-cycle level.
This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer