MACOM Is a Great Business, but the 35% Dip Isn't Cheap

MACOM is exactly the kind of growth stock investors love to catch on a dip: an AI-data-center chip supplier with a blowout quarter behind it, guidance well above Wall Street's numbers, and a share price roughly 35% below where it traded in May. The reflexive read is that the drop is a gift. The business earned its praise. But the price, even after the drop, still asks for years of perfection — which is why I wouldn't call this the low-risk entry the label implies.

Look at quality first, because nothing else matters if the growth is hollow. In its fiscal third quarter, ended July 3, MACOM posted revenue of $342.2 million, up 35.8% year over year. Adjusted operating margin jumped to 31.5% from 25.2% a year earlier — the operating leverage is real, and adjusted gross margin reached 59.7%. The demand signals back it up: a company-record book-to-bill ratio of 1.6, data-center revenue up roughly 40% sequentially to a record level, and a revenue guide of $415 million to $425 million for the current quarter that came in far above the roughly $364 million analysts had modeled. The company also holds net cash near $322 million. This is not a broken story: MACOM sells high-speed optical and gallium-nitride parts across data centers, defense, and telecom to more than 20 customers each pulling in over $10 million a year. The moat survives the exact stress that shook the price. Gate one passes cleanly.

The gate that fails is the one that matters for a contrarian entry. A drawdown only turns into a low-risk buying opportunity when it compresses a stock's multiple toward the market's while growth keeps running at a multiple of the market — that is the condition that shifts the burden of proof onto the bears. MACOM does not meet it. Even after falling from its May high of $418.90 to about $272, the stock still trades at roughly 86 times trailing earnings and about 73 times trailing EBITDA. That is on the order of four to five times the broad market's multiple — and essentially in line with, not cheaper than, its own AI-datacom neighbors: Marvell sits near 71 times EBITDA, Credo near 55 times, Astera near 168 times. All of them fell 3% to 5% on the same recent session, which tells you the selloff is a sector repricing rather than a MACOM-specific mistake. MACOM's dip took a hyper-expensive multiple down to a merely expensive one. The gap between price and any ordinary market multiple never opens, so the "buy the beaten-down winner" logic never actually fires.

That is what the current price is really asking for. At 86 times trailing earnings and roughly 18 times trailing sales, the market is paying full freight for the continuation of 30% to 50% annual growth several years into the future. That is a bet on the durability of the AI buildout, not a bet on a wounded business healing. The one bear fact this thesis has to answer is hyperscaler spending on optical connectivity: if it wobbles, the whole cohort reprices lower together, and no company-specific quality insulates you. Competition here is real — Marvell, Credo, Astera, and Coherent crowd the same optical and connectivity space. That is not the same as saying MACOM's edge is breached; it is to say the moat is an advantage inside a sector whose prices already assume the advantage lasts and none of whose names is cheap enough to represent a clean mispricing on its own.

So here is my honest read. This is a genuinely strong operator, and the stock is a clean study in how a big drawdown can coexist with no valuation cushion. The market is not wrong about MACOM's business — it's just not underpricing it. The setup I actually look for would need either the multiple to compress meaningfully toward the market while growth stays intact, or price action to confirm the selloff has run past the numbers. Neither has happened. I'd keep MACOM on the watch list and let the price earn a more aggressive call; right now the burden of proof sits with the bulls paying 86 times trailing earnings, not with the bears.