Booz Allen Hamilton Holding Corporation BAH

Key data

The business

Booz Allen sells the U.S. government its brains. About 98% of revenue comes from federal agencies, and 95% of it as the prime contractor, not a sub: cleared engineers and analysts embedded inside the Pentagon, the intelligence community, and civil agencies, doing cyber, data, and increasingly AI work that outsiders are not allowed to see. The moat is the security clearance and the incumbency. You cannot spin up 35,000 cleared technologists overnight, and once Booz Allen holds a program it tends to hold it for years, which is why total backlog sat at $38.2B in March, more than three years of sales. The newer story is a deliberate shift from selling hours to selling products: the Defy Security deal closed this year, and in August the company issued notes to buy Ultra's intelligence unit, calling itself now "an advanced technology company" that builds commercial-grade tools rather than a pure staffing shop.

The numbers

The government slowdown is real and it is in the print. Revenue fell for the first time in years in FY26, and the quarters show where it turned.

The Q1 FY27 optics are worse than the business. Net income fell 27% year over year, but that is entirely tax: pretax income rose 16% and operating income rose 8.6% on 4% lower revenue, so operating margin widened from 9% to 10%, while the tax line swung from a $55M benefit to a $53M charge. That watch-item from earlier this month held rather than broke, the earnings dip was optical, not operational. Cash backs it up: operating cash flow jumped to $281M from $119M a year ago.

Revenue compounded about 13% a year from FY22 to FY25 and EPS more than doubled. FY26 broke the streak: revenue down 6.4%, EPS down to $6.90, trailing earnings now $6.37. Consensus sees $6.41 for FY27 and $6.72 for FY28, essentially flat, which means the market is pricing a trough that never recovers. At 11 times that trough against a 12-year history of 17 to 22 times, one of two things is true: the decline is permanent, or the multiple is wrong. The variant here is that the underlying business is still growing, pretax up 16%, margins expanding, cash conversion improving, and the products pivot adds a growth line the 11x multiple gives you free. The print that settles it is a return to positive organic revenue growth with book-to-bill back above one.

Three-quarters of that backlog is priced options, work the government has negotiated but not yet ordered, which is real coverage but soft in a budget fight.

Management

The tell worth watching is Horacio Rozanski's $2.0M open-market purchase in October 2025, made near $107 and now underwater by about a third, the only insider buy on the tape against two token sales whose 10b5-1 status is not disclosed. That is conviction that has cost him so far. Capital allocation has shifted: buybacks slowed to $598M in FY26 from $812M, and to just $72M last quarter, while cash fell to $540M and the company took on debt to fund M&A. So the message is consistent, management is spending on acquisitions and its own stock rather than hoarding cash, with net debt at about 2.9 times EBITDA and the dividend covered.

How it fails or surprises you

The budget knife goes deeper. Revenue already fell 6.4% in FY26 on federal cost-cutting, and with 98% of sales tied to one government, a second leg of cuts or a contract descoping takes revenue down another 5% to 8% before the products pivot is large enough to matter. The first sign is funded backlog, $4.3B, ticking lower quarter over quarter.

Margin was the whole story, and it fades. The case for owning it leans on operating margin rising 9% to 10% on falling revenue. If that was cost discipline pulled forward rather than structural, and revenue keeps sliding, decremental margins bite and EPS follows the top line down. Watch operating margin holding double digits against a lower revenue base next quarter.

The re-rate (right tail). At 11 times versus a 17 to 22 times history, the market is pricing permanent decline. If organic growth merely turns positive and Defy and Ultra show the products line compounding, the multiple normalizes toward 15 times on $6.70 of earnings, roughly $100, before any earnings growth. The print is one quarter of positive organic revenue with book-to-bill above one.

Closing thoughts

If organic revenue turns positive over the next two quarters and book-to-bill holds above one, the 11x multiple normalizes toward 17-22x, roughly $100 before any earnings growth. If revenue keeps sliding, you own a shrinking contractor at a fair price with covered cash flow, a slow bleed not a break. The upside is more likely: margin expansion and doubled operating cash say the business is strengthening while revenue fell on budget timing and tax swing, not structural damage. The downside is capped by $38B of backlog and cash conversion improving.

The bet is still that the U.S. government keeps hiring Booz Allen to run its most technical work, the spy agencies, the Pentagon, the AI systems, and keeps paying about what it paid before. What breaks it is a deeper budget cut that pushes organic revenue down while margins give back their gains. The one pair of numbers that tells you first: funded backlog against quarterly operating margin. Hold both, and the 11x multiple is a mispriced trough, not a value trap.