$16.29
$68.08
bear $16.29
base $50.20
bull $68.08
margin of safety $26.00
market price* $43.30
KB Home builds single-family homes, from entry-level homes suited to first-time buyers up to larger homes in premium locations, in four regions: West Coast (California, Idaho, Washington), Southwest (Arizona, Nevada), Central (Colorado, Texas) and Southeast (Florida, North Carolina). Most of its cash goes into land and homes under construction, so its reported cash flow swings with how fast it buys land rather than with what it earns. To see through that, we measured what the business generates after stock compensation and the equipment spending it needs to stay in business, but before the cash it ties up in land and inventory. On that basis owner earnings per share (after interest) were $3.76 in 2019, $4.03 in 2020, $6.74 in 2021, $9.63 in 2022, $7.86 in 2023, $8.60 in 2024 and $6.91 in 2025. Non-cash charges, including inventory write-downs ($32.1 million in 2025 versus $4.6 million in 2024), deferred taxes and equity in joint-venture income, are handled automatically through cash flow, and for 2023 to 2025 the figures reconcile exactly to net income plus those items. Because almost all interest is capitalized into inventory (about $114 million incurred in 2025), we value the business before its lenders and then subtract the debt. Future cash is valued at 6% a year. What has changed: the housing cycle is turning against the company. In 2025 net orders fell 11% to 11,596 homes, deliveries fell 9%, and the housing gross margin fell to 18.6% from 21.2% (19.1% excluding inventory write-downs, down 200 basis points). In the first half of 2026 revenue was $2.19 billion versus $2.92 billion a year earlier (down 25%), net income was $60.8 million versus $217.4 million (down 72%, with roughly $33 million and $27 million in the two quarters), and operating cash flow was an outflow of $93 million. Notes payable rose to $1.97 billion at May 31, 2026 from $1.69 billion at November 30, 2025, while the company repurchased $130 million of stock in the half. Two weak quarters in a row are enough for us to lower the near-term path sharply; they are not enough to change what we think the business earns across a full cycle. Separately, the federal tax credit for energy-efficient new homes (worth $13.1 million in 2025 and $19.3 million in 2024) ends for homes delivered after June 30, 2026, and the company expects its 2026 tax rate to be 24% to 26% versus 22.6% in 2025. Our earlier analysis was built on a different basis, and its figures are not comparable with these. Anchor. We used 2019 to 2025, seven years that cover a pre-boom market, the 2021-2022 surge and the normalization since. The median cash earnings before interest over that window is $8.56 per share, about $537 million a year at today's 62.7 million shares, and the lowest year in it was about $496 million. We left out 2016 to 2018 because those years carry unusually large deferred-tax add-backs (deferred taxes of $105 million in 2017 and $192 million in 2018). We do not use the median as the base case because the strong margins of 2021 to 2024 are not repeating in the current results, and the loss of the energy-efficient home credits and the higher tax rate are known, dated headwinds. Growth-path check: cash earnings per share before interest fell 14% in 2025 ($9.97 to $8.56), and owner earnings after interest fell 20% ($8.60 to $6.91). For 2026 we assume a far larger decline, to about $3.7 per current share, so year one does not assume any reversal of the trend. Base case. 2026 brings about $232 million of cash earnings before interest (our estimate of about $100 million in the first half, built from the quarterly figures, plus $130 million in the second half). We then assume $290 million in 2027, $350 million in 2028, $410 million in 2029, and $450 million a year from 2030, flat in dollars through 2046. The $450 million is what the business has already shown it can produce: every year from 2019 to 2025 was at least $496 million, and we take roughly 10% off the lowest of those for the lost tax credits and higher tax rate. We credit no growth from new communities, price increases, margin recovery or share repurchases: land spending and buybacks use the same cash, and nothing signed or under way secures an increase. We stop the story at 20 years, roughly two housing cycles, because nothing we can evidence extends it further, and we credit nothing after that. The business is worth about $5.0 billion ($80.0 per share). It owes about $1.87 billion (notes payable of $1.97 billion less $100 million of the roughly $200 million of cash, which we treat as excess). The difference is about $50.20 per share. Bear case. This follows the scale of the 2007-2010 housing collapse, taken from industry history rather than from the filings, layered onto the current slump: the weak 2026 gets weaker (about $100 million in the second half), then $180 million in 2027, and $120 million to $130 million in each of 2028 and 2029, which is about the size of the company's annual interest bill and roughly a quarter of the normal level. Recovery reaches $250 million by 2031 and settles at $300 million, barely above today's depressed run rate, for 20 years. Rising write-downs of land and inventory are already visible (a seven-fold increase in 2025), and in a collapse they would be large, although they are non-cash and so do not show up in this measure; they would, however, erode book value, which was about $59.65 per share at May 31, 2026. We also assume refinancing costs about $110 million more in present-value terms, because we assume new debt costs about 2.5 points more than the existing coupons. The maturities are about $301 million in fiscal 2027 (including the 6.875% senior notes due June 15, 2027), about $660 million in fiscal 2029 (the Term Loan matures November 12, 2029, and the 4.800% senior notes are due in 2029), and $350 million in fiscal 2030. The business is worth about $3.0 billion ($47.80 per share). It owes about $1.98 billion ($31.50 per share). The difference is about $16.30 per share. In the trough years cash earnings roughly match the interest bill, and liquidity of $1.43 billion at November 30, 2025 (the latest figure we reviewed) covers any gap. Under its credit agreements, the financial test is breached only if both liquidity falls below annual interest and the interest coverage ratio drops below 1.5 times. Bull case. The company's own levers execute: cash earnings before interest of $150 million in the second half of 2026, $380 million in 2027, $480 million in 2028, and then $540 million a year, the seven-year median, for 20 years. The evidence is that the company produced $593 million in 2025, $768 million in 2024 and at least $496 million in every year since 2019. The business is worth about $6.1 billion ($97.90 per share). It owes about $1.87 billion. The difference is about $68.10 per share. What we tried to break. The load-bearing assumption is that the business returns to roughly $450 million a year. The specific facts against it are that by our estimate the first half of 2026 produced only about a third of the prior year's first-half cash earnings, that the gross margin fell 2.6 points in 2025, and that debt is rising while the company is buying back stock. We kept the 2026-2028 path low for that reason and made the bear case close to the interest bill. What supports it is the record: no year since 2019 was below $496 million, and the senior notes were valued at about their carrying amount ($1,333 million of fair value versus $1,332 million carrying value at November 30, 2025). Second-quarter 2026 figures come from the company's reported financial data; we did not review that filing's narrative commentary. How we arrived at the margin-of-safety price. We are moderately sure of the long-run level but not of the next two years, because margins are falling and first-half 2026 profit fell 72%. The worst plausible outcome is a housing collapse like 2007-2010 combined with falling land and home values. The structure would likely survive it for several years: the nearest large maturity is $300 million in June 2027, the senior notes were valued at about their carrying amount, and liquidity was $1.43 billion at November 30, 2025, but book value would be heavily written down and shareholders could be diluted if refinancing failed. If we are wrong about the normal level and it is $350 million rather than $450 million, the shares are worth about $35 each; a shorter life or no margin recovery would push the value toward the bear case of about $16. The price assumes a three-to-five-year holding period, during which the company pays a dividend of about $1 per share a year. Weighing those outcomes, the margin-of-safety price is $26. From that price, the base case implies about 30% a year over three years, a normal level $100 million lower still implies about 20% a year including dividends, and the bear case implies a loss of about 37%, which is the 10th-percentile outcome.