TTM Technologies: The Board-Shop Discount Is Closed — Now the Cash Has to Show Up

TTM Technologies: The Board-Shop Discount Is Closed — Now the Cash Has to Show Up

The market just did something that looks backwards, and the explanation is the most useful part of the story. TTM Technologies, a manufacturer of printed circuit boards and mission electronics, reported the best quarter in its history in early August: revenue crossed $1 billion for the first time, non-GAAP earnings jumped 71% from a year earlier, and management raised its full-year forecast. Since then the shares have fallen roughly a quarter in a month and now sit about half below their June peak. Record results, halved price.

The two facts are not contradictory. The drop is what happens when a price has already finished paying for a story — and the story, in market shorthand, was always "priced like a board shop, booked like a defense prime." The market has now acted on that story in both directions. What is left is a company that has to convert its accounting earnings into actual cash, with $1.1 billion in newly borrowed money riding on the outcome.

The "booked like a defense prime" half is real.

TTM builds the physical circuit boards every electronic product sits on, plus radio-frequency components and, increasingly, whole mission systems. A commodity board shop is a cyclical price-taker. TTM spent years trying not to be that: it bought Telephonics, a defense surveillance-and-communications house, for $330 million in 2022 to move up the value chain into military programs.

By the June quarter the pivot shows up in the order book. Aerospace and defense was 37% of sales at $382.8 million, up 14%. The defense program backlog was above $1.7 billion, up from $1.5 billion a year earlier, and the 90-day backlog reached $901 million, up 81%. Total book-to-bill — orders booked against orders shipped — was 1.49, and in aerospace and defense it was 1.3: new work is arriving faster than it ships. Behind the backlog sits a qualified proposal pipeline management puts above $7 billion, which deserves distrust until it becomes ordered work.

There is a second engine, and it is bigger. Data center and networking was 40% of second-quarter sales, up 91% year over year, because AI buildouts demand a specific class of advanced board that TTM makes in new factories it is still ramping. Management's shorthand: artificial intelligence and defense together are about 80% of net sales.

That is a real change, with one honest caveat. A Lockheed-technology backlog covers several years of revenue; TTM's $1.7 billion covers roughly a year of its aerospace-and-defense sales. TTM is a defense-prime supplier, not a prime. But a program book that keeps refilling and a data-center leg growing 90% is a genuinely different demand profile from the commodity board business — and it is the reason the stock was re-rated in the first place.

The market already paid for that.

Watch what the re-rating did. Over the past year TTM went from a 52-week low near $43 to an intraday peak of $223.83 in late June, then halfway back to about $107 today. At the June peak the stock traded at roughly 45x even the full-year EPS target management has since raised to "approaching $5.00". At $107 it is about 21x that target, and about 17x the roughly $6.34 consensus for the following year — a near-50% year-over-year EPS jump built into the consensus. The collapse is not the market deciding the business is broken; it is a multiple reset from a price that had run ahead of even the raised numbers, alongside the roughly $17–20 million of insider selling reported in the 90 days before the report. Aggregate signals like AInvest's still label the stock a Buy — a measure of where sentiment sits, not of what the math supports.

And the old comps confirm the cheap label is gone. TTM already trades at roughly 3.3x trailing sales and 24x trailing EBITDA, versus a legacy manufacturing peer like Sanmina at 0.8x sales and Flex at 1.4x. Its trailing P/E of about 48 sits above defense primes including Lockheed (~21x), Northrop Grumman (~17x), and General Dynamics (~23x) on trailing GAAP earnings. What separates TTM from those slow-growing primes is the growth: non-GAAP EPS of $2.45 last year is on track to roughly double this year, and management guides to 15–20% organic revenue growth through 2028.

The number the market is really waiting on is cash.

Here the record meets the accountant. The "record" earnings are non-GAAP: quarterly GAAP EPS was $0.77 against the $0.99 non-GAAP number, and trailing GAAP EPS is only about $2.2, the difference dominated by amortization of purchase intangibles from past deals. The cash statement is harsher still. Over the trailing twelve months, capital spending of about $338 million ran ahead of operating cash flow of about $323 million, leaving free cash flow roughly flat to slightly negative while TTM builds an ultra-high-density plant in Syracuse, ramps a new factory in Penang, and adds AI-board capacity. The June quarter itself finally flipped positive — $96.4 million of operating cash flow against $50.4 million of capex — but the board then added another $45 million to 2026 spending.

Then came the biggest deal in company history. On August 17 TTM agreed to buy Epiq Design Solutions, a software-defined-radio and space-compute business with defense and signals-intelligence customers, for $1.1 billion in cash — about 10% of the company's market cap, and more than three times what it paid for Telephonics. Banks have committed the financing. Net leverage goes from 0.9x to about 2.3x at close, with a plan back to 1.5–1.7x within 12 to 18 months; management values the deal at 17.4x expected adjusted 2027 EBITDA counting synergies, and expects it to add to non-GAAP EPS only in 2028. Because the purchase is all cash, most of that $1.1 billion becomes intangible assets to amortize — widening the gap between the GAAP books and the non-GAAP ones that show the record. Negative trailing free cash flow, a tripling of net leverage, and a wider accounting gap: that is why a stock can fall on a great quarter.

One more expectation to price in. Simple arithmetic from the guide: "approaching $5.00" for the year, with $0.75 and $0.99 behind it and $1.24 the midpoint of third-quarter guidance, implies fourth-quarter EPS around $2.00. The full-year target depends on a fourth quarter roughly double the current record, which means the AI-board ramp landing exactly on schedule. The market's 21x multiple is really a bet on that quarter.

What would prove the case

The checks fall out of the numbers. The book must keep refilling — book-to-bill above 1.0, the 90-day backlog growing, since that is the contracted evidence that demand is not just a pipeline story. Free cash flow has to turn structurally positive once the building program peaks; a non-GAAP EPS number that never converts to cash is an estimate, not earnings. And leverage has to come back inside roughly 2x with Epiq delivering its promised 2028 accretion — a bet that size cannot be waved through on management's confidence.

Two further variables sit outside the company's control and inside the thesis: hyperscaler capital spending, which drives the data-center board ramp (management guides ~$600 million of second-half revenue from the advanced N+M boards alone), and continued defense award flow.

The compressed case

The halving was a multiple reset, not a broken business — the backlog, the book-to-bill, and the twice-raised guidance say the demand is contracted and funded, and a supported scale-up's heavy investment can be temporary when the demand is real. But the asymmetry that made TTM interesting is gone; nobody is being offered a defense prime's order book at a board shop's price anymore. At roughly 17x next year's non-GAAP EPS against a growth rate that usually commands more, the stock is a reasonable growth price that no longer builds in much room for disappointment, and the entire margin of safety rests on one conversion: record non-GAAP earnings turning into cash while $1.1 billion of new debt does its round trip. That conversion is checkable quarter by quarter. Track it, not the chart, and this becomes an execution story you can verify — not a collapse you are guessing about.