STERLING INFRASTRUCTURE, INC. (STRL) stock intrinsic value fair value estimate & owner earnings analysis

Sterling Infrastructure is a U.S. contractor with three businesses. E-Infrastructure Solutions (about $1.47 billion of 2025 revenue, 23.6% operating margin) prepares large sites for data centers, chip plants and distribution centers and, since buying CEC Facilities Group in September 2025, also installs electrical and mechanical systems. Transportation Solutions ($641 million) builds highways, airports and rail in Utah, Arizona, Colorado, Nevada, Texas and Hawaii for state transportation departments. Building Solutions ($383 million) pours foundations and installs plumbing for homebuilders in Dallas-Fort Worth, Houston and Phoenix. What the business earns. We measure the cash an owner could take out after keeping the business as it is: operating cash flow, less the swings in working capital, less stock pay, less the capital spending needed to maintain the equipment fleet, less the share of profits that belongs to joint-venture partners. On that basis Sterling earned about $1.00 to $1.28 per share in 2017-2019, $2.53 to $2.58 in 2020-2021, $4.02 in 2022, $4.74 in 2023, $6.02 in 2024 and $10.30 in 2025. Our earlier figures for 2023 and 2024 (around $12.75 a share) included more than $450 million of customer prepayments on data-center jobs. That is real cash, but it reflects how fast the backlog grew rather than what the work earns, and it reverses when growth stops. We now remove it, so the previous figures are not comparable. Reported profit and our cash measure line up closely once this is done. The 2024 accounting gain of $91 million from deconsolidating Road and Highway Builders is excluded because no cash came in. Which years count. The business today is not the business of 2016-2021. Most of its profit now comes from mission-critical site and electrical work, where margins went from 22.0% in 2024 to 23.6% in 2025. For the starting level we use the business before interest: about $10.88 per share in 2025, supported by results for the first half of 2026 (operating income $357 million against $161 million a year earlier) and a backlog of $4.23 billion at June 30, 2026 against $3.01 billion at year-end 2025. We do not treat the 2025 level as permanent. Owner earnings grew 71% in 2025, helped by four months of CEC. The base case assumes about 47% growth in 2026 and then slows sharply. That is below last year's rate, and the backlog already signed largely covers it. The main risk is that a large share of profit comes from a few huge customers building data centers. The top four customers were 27% of segment revenue in 2025 and 40% in 2023. Construction has always been cyclical. In 2015-2016, before this mix shift, Sterling lost money, and the warehouse building boom stopped abruptly in 2023-2024. The $555 million of acquired customer relationships and trade names on the balance sheet will run off. That is why every case assumes today's margins fade rather than last. Bear case, about $98 per share. Backlog carries 2026 and part of 2027, then data-center capital spending pulls back the way warehouse spending did after 2022. Sterling has never been through a downturn in its newer mix, so we size the slump from the industry's own record: earnings fall over three years to roughly a third of the 2027 peak, around $3.50 a share, about where they stood in 2022-2023, as competitors bid down margins on a smaller pool of work. Housing stays weak and highway work does not offset the decline. Earnings recover only to about $5 a share, a level the business has already shown it can reach. The business is worth about $3.0 billion. It owes a $269 million term loan due June 2028, about $15 million of current debt and up to $80 million of CEC earn-out. Against that sit about $388 million of cash it controls, so net cash is roughly $24 million and the equity is worth about the same as the business. Even at the bottom, cash flow is many times the roughly $17 million a year of interest, so refinancing the 2028 term loan is not a threat in this case. Base case, about $196 per share. The signed backlog turns into earnings of about $16-17 a share before interest in 2026-2028. Then margins drift back as data-center construction normalizes and the acquired customer relationships wear off, with earnings settling near $10 a share by the mid-2030s. That is roughly where the business already earned in 2025, and it is held there by its position in regional site work and electrical contracting. No new acquisitions, new regions or market growth are credited. Business value about $6.1 billion, net cash about $24 million. Bull case, about $292 per share. Data-center and chip-plant programs already announced by Sterling's customers keep the backlog full through about 2030. CEC's electrical work grows inside Sterling's existing customer relationships, and earnings peak near $22 a share before settling around $16. That would still be well above anything earned before 2025. Business value about $9.1 billion. Margin-of-safety price. We are fairly sure of the next 18 months, because signed backlog is 70% larger than a year ago, and much less sure after that. Today's profit comes from an unusually strong building boom for a handful of hyperscale customers, and this company has not yet been through a construction downturn in its current form. The worst plausible outcome is a sudden halt in data-center spending while housing is also weak. The balance sheet would survive that easily: net cash, a modest term loan not due until 2028, and assets that can be sold, such as equipment and joint-venture interests. What we are most likely to have wrong is how long margins stay above 20% and how deep the next slump is. If margins return to the low teens of Sterling's older work, the shares are worth less than $100. The price assumes a three-to-five-year holding period, and the company pays no dividend, so the whole return has to come from earnings growth and share buybacks of about $70-75 million a year. At $85 a share, the price is about eight times 2025 owner earnings, before most of the signed backlog has been earned. Even in the bear case the shares would be worth more than that, and a 20% annual return over several years needs only a fair part of the base case to happen. That is why the margin-of-safety price is $85.