Jacobs Is Charging Nvidia Prices for Engineering Services

# Jacobs Is Charging Nvidia Prices for Engineering Services

Let's get the headline out of the way first: Jacobs Solutions just reported a quarter that would make almost any industrial company blush, and then quietly watched its own stock get sold off anyway.

On August 4, Jacobs posted fiscal Q3 numbers that read like an AI-story success story wearing an industrial company's skin. Gross revenue jumped 34.5% year over year to roughly $4.1 billion. Adjusted net revenue—what actually counts—grew 8.3%, all organic, to $2.42 billion. Adjusted EPS came in at $1.84, ahead of the $1.83 street number. And the backlog? A record $28.9 billion, up 27% YoY, with a book-to-bill of 1.4x on a gross basis. Management has now raised full-year guidance for a third straight quarter, guiding to $7.20–7.30 adjusted EPS.

That's not a company that's flailing. That's a company that's running.

And yet the stock got crushed. It's down roughly 18.5% over the last three months, lagging its industrial peers, and it's trading around $136 on a market cap near $15.8 billion—the most recent print, after a slide from the mid-$150s down into the $136–145 range through September. So what is the market actually pricing, and is it right?

Here's my take: the market isn't pricing Jacobs poorly. It's pricing a story it's tired of.

Consider the bear case, because it's not weak. Gross revenue soared 34.5% while net revenue crawled at 8.3%. That gap is pass-through work—materials, sub-contractors, costs flowing straight through the income statement—and it's the oldest trick in the engineering-services playbook to make growth look enormous on the top line while the margins barely move. Adjusted EBITDA margin did expand 110 bps to 15.2%, so the company is earning its keep. But the street wants to see quality, and quality is a hard thing to feel when you're watching a pile of low-margin throughput masquerade as growth.

Then there's the valuation, and this is the part that should make a buyer nervous. Jacobs is going for a trailing P/E in the 40–50x range against a five-year median of roughly 27.6x and peer averages near 20x. That is not a modest premium for an industrial services firm. That is a full-blown AI-growth multiple, and multiples like that don't survive contact with a slow quarter. They don't survive when AI data-center demand starts looking cyclical. They don't survive when the tax rate spikes to 43.4% (versus 21.9% a year ago) on PA Consulting integration noise.

So where does that leave the tape?

The bulls have real ammunition. Jacobs is the #1-ranked firm by ENR in 18 categories, including data centers. AI infrastructure is now 11% of adjusted net revenue, and the company just landed a sole-source EPCM job for Hut 8's 1 GW "Beacon Point" data-center campus in Texas. It's deploying NVIDIA data-center digital twins, extending into hydrogen steel, locking up nuclear and wastewater work on multiple continents. The backlog is a rolling fortress.

The bears have the math. And in a market that has taught all of us a hard lesson about paying up for optimism, math tends to win at the checkout.

My read? Jacobs is a genuinely well-run business trading at a price that demands perfection on every subsequent print. The AI data-center optionality is real and repeatable. But you are not getting paid to be patient at 40-plus times earnings. The kind of quarter that would be a home run at $110—say, net revenue growth ticking up, margins continuing to expand, book-to-bill holding firm—becomes merely "expected" at $136, and any miss becomes a multiple-compression event, not a story failure.

That asymmetry is worth holding in your mind as we head into November. Jacobs reports Q4 FY2026 around November 19, with the conference call likely landing near November 17. That's the next real catalyst, and it's the moment the market decides whether this is a durable AI-adjacent compounder or an industrial company that got carried away with the tape.

Here's what I'll be watching when that call happens:

• Net revenue growth vs. gross. If net revenue growth accelerates past the 8.3% print, the pass-through worry loses its teeth. If it stalls, the premium isn't justified.
• Book-to-bill on the net basis. The TTM figure is 1.2x; the trailing quarter came in at 1.1x. That deceleration is the exact kind of slow-turning steering wheel that makes a multiple-compression trade.
• The tax rate. Management calls the 43.4% temporary. I'll believe it once the next filing shows it coming down, not up.
• The Shannon Miller follow-through. Miller's transition from President of Strategy, Growth & Digital—announced August 10, with SEC filings hinting at an advisory role rather than a clean exit—wasn't the whole story. Any real formalization or follow-up move is worth watching for signal on strategy continuity.

The honest bottom line: Jacobs handed the market a first-class quarter and got penalized for it, which is either a mistake or a warning sign depending on how you view the AI infrastructure buildout. My bias is that this is a high-quality franchise at a price that leaves almost no room for error. I'd be a buyer here only if you already own the thesis and can stomach the volatility—not to chase a multiple that's already at 40x.

The November earnings will tell you whether the market was right to be cold, or just early. Until then, this is a stock you watch, not one you chase.