HORTON D R INC /DE/ (DHI) stock intrinsic value fair value estimate & owner earnings analysis

D.R. Horton is the largest homebuilder in the United States by volume. In the year to September 2025 it had revenue of $34.3 billion, and it sells mainly entry-level and first move-up homes across the Sun Belt, with Texas and Florida the biggest states. It also runs a mortgage company that finances most of its customers, a rental-community business, and owns about 62% of Forestar, a lot developer that supplies land to it and to other builders. Its model is light on capital for a homebuilder. It builds largely on speculation, turns inventory quickly, and controls most of its future lots through option contracts with small deposits rather than owning the land. That lets it shrink quickly when demand falls. How we measure earnings. A homebuilder's reported operating cash flow swings wildly because nearly all its cash goes into or comes out of land and homes under construction. So we start from operating cash flow before those inventory, mortgage-loan and payables swings. We then subtract stock compensation, the equipment spending needed to stay in business, and the share of earnings that belongs to Forestar's outside owners. Measured this way, owner earnings per share were $2.61 in fiscal 2016, $4.25 in 2019, $6.37 in 2020, $11.55 in 2021, a peak of $16.68 in 2022, $13.79 in 2023, $14.42 in 2024 and $12.36 in 2025. In every year checked, this figure ties to net income plus non-cash charges with no gap. One important judgement follows. When the business grows, it must tie up more money in land and homes, about 60 cents of inventory for every extra dollar of annual revenue on its recent asset base. We charge that reinvestment against any growth we credit, so growth is not counted as free. The earnings level. Homebuilding is cyclical, so we set the level on the full ten-year span from 2016 to 2025, which runs from the recovery years through the 2021-2022 boom and into today's slowdown. The median of that span in total dollars, before interest, is roughly $3.3 billion a year. We use totals rather than per-share figures because the share count has fallen from 375 million to about 283 million through buybacks. Per-share numbers would partly credit the buybacks a second time. The current year is running below that level. Net income for the nine months to June 2026 was $2.15 billion against $2.68 billion a year earlier. Gross margins remain squeezed by the mortgage-rate buydowns the company pays to keep homes affordable. Last year owner earnings fell 20%. So our base case starts below the normal level and does not assume a return to it for several years. Base case. Business cash flow of about $2.9 billion in the coming year, compared with roughly $4.0 billion in fiscal 2025. That is a steeper decline than last year's actual 20% drop, because the current year's results are tracking lower. It then recovers gradually to $3.3 billion by year four, the median of a full cycle, and holds there. We credit no growth beyond that. Volume gains from share gains, a housing shortage or a rate recovery are not things the business has secured, and growth would in any case require more inventory investment. The business is worth about $54 billion. It owes about $4.3 billion after its cash: homebuilding, rental and Forestar notes of roughly $5.6 billion, plus Forestar's outside owners' stake of about $0.6 billion, less about $1.6 billion of surplus cash. The mortgage company's warehouse lines are left out because they are matched against the mortgage loans they fund. That leaves about $50 billion for shareholders, or about $177 a share. Bear case. This is one full housing downturn of the kind the industry actually lived through in 2007-2011. In that episode D.R. Horton's revenue fell by roughly three-quarters from its 2006 peak and it reported losses for several years as land values were written down. We model a milder but still severe version that reflects today's much lighter land position. Cash flow falls to $2.6 billion next year and $0.3 billion the year after, then turns negative for two years: $1.5 billion and $1.0 billion of losses from land write-offs, forfeited option deposits and price cuts. It is near zero in year five and recovers only to $2.6 billion by year ten, below the cycle median, on the assumption that a shrunken business permanently loses some scale. Debt is manageable through this. In the last downturn the company paid down debt by liquidating inventory, and its notes are spread out, with the largest maturities $0.6 billion in 2026 and $0.5 billion in 2027. The business would be worth about $31 billion. Less $4.3 billion owed leaves about $26 billion, or about $93 a share. That is somewhat below the $84 a share of book equity, which is right for a case where land values are written down. Bull case. This is what the company's own levers deliver if they keep working. Cash flow returns to about $3.9 billion by year three on margins the business earned as recently as fiscal 2024. After that it grows revenue 3% a year, its long-run pace before the pandemic boom, from new communities on lots it already controls, and funds the extra inventory out of cash flow. The business would be worth about $72 billion. Less $4.3 billion owed leaves about $67 billion, or about $238 a share. Change since our previous analysis. Our previous estimate rested on net income and a different method of building the cases, so the earlier figures are not directly comparable. In business terms, the current year's earnings have come in somewhat below last year's, consistent with the softening our earlier view described. Margin of safety price. How sure are we of the earnings level? Fairly sure of the range, not of the timing. Ten years of filings show the business earning between $1 billion and $6 billion a year before interest, and today sits in the middle of that range. The worst plausible outcome is a repeat of 2007-2011: several years of losses, write-downs on the $14 billion of owned land, and forfeited deposits. The business's structure would survive it. Debt is modest relative to $24 billion of equity, three-quarters of its lots are only optioned, and homes and land are hard assets that can be sold. The things we are most likely to be wrong about are how deep and long the next downturn is, and whether the margin squeeze from rate buydowns proves permanent rather than cyclical. If both go against us, the shares are worth around the bear figure of $90-95. Over the holding period, the company pays $1.80 a share a year in dividends and has been retiring around 7% of its shares a year. At $95, the price is no higher than the bear-case value, roughly what a severe downturn leaves. In a median outcome, value of about $177 within four to five years plus dividends implies roughly a 20% annual return. The margin of safety price is $95 a share.