NXP Semiconductors: The Q2 Beat Is Real, the Discount Is Earned

NXP Semiconductors delivered the kind of quarter that normally reads as unalloyed good news, and the stock went down anyway. Revenue climbed 19% from a year earlier to $3.50 billion, adjusted earnings per share rose 33% to $3.61, and management guided the next quarter above what analysts had penciled in. Days after the July 28 report, the board declared another quarterly dividend of $1.014 a share. Yet the stock has shed about 16% in the month since earnings, to roughly $224 — about a third below its 52-week high of $340 and only about a fifth above its low of $183. When a company beats, pays, and still falls, either the market decided the good news was not good enough, or the good news was already in the price. For NXP the answer is a mix, and sorting it out is what decides whether this is a bargain or simply a cheap cyclical.

Start with the market's reaction, because it tells you what the price already carried. The revenue beat itself was thin — about $3.47 billion was expected — and the third-quarter guidance, while ahead of expectations at the midpoint, came with a wide range of $3.65 billion to $3.85 billion of revenue, a spread investors read as hedging rather than confidence. The stock had also already run: it jumped 25.6% in a single day after first-quarter results in late April, closed near $316 in late May, and was fading into the report. July then turned into one of the weakest months for semiconductor stocks in over a decade, as the market soured on the timeline for AI returns, and analysts who follow the name pointed to its exposure to the Chinese market as a reason for caution. NXP's trailing-twelve-month revenue of about $13.2 billion remains only modestly above the $12.6 billion it reported for all of 2024 — a reminder that this is a business climbing out of a downturn, not a freshly discovered growth story. That is the "cyclical recovery rather than momentum growth" label the market applied, and it accounts for most of the drop.

What the quarter actually proved runs through the operating numbers rather than the headline multiple. The critical detail is not the growth rate but its source. Management said company-specific growth drivers — the extra processing content going into each car and each machine — grew in the mid-20% year over year and made up roughly a third of second-quarter revenue, and that the growth reflected content rather than customers restocking inventories. The distinction is the durability test for a chipmaker: restocking recoveries die when warehouses refill, while content growth compounds. The segment table supports the reading. Automotive, still more than half of NXP's business at $1.94 billion of the $3.50 billion total, grew 12% year over year; industrial and IoT grew 38%; communications infrastructure grew 41%, helped by a data-center business management expects to pass $500 million in 2026 after about $200 million in 2025. The recovery is broadening beyond autos, precisely what a value investor wants to see from a name whose discount rests on auto concentration.

Margin and cash flow confirm the operating story is not an accounting artifact. Adjusted gross margin reached 58.0% and adjusted operating margin 35.1%, while even GAAP operating margin rose to 30.6% from 23.5% a year earlier, so the gain holds on the unadjusted statement as well. The quarter produced $860 million of operating cash flow and $791 million of non-GAAP free cash flow, 22.6% of revenue. Just as important for this particular stock, the balance sheet strengthened through the downturn rather than cracking: net financial leverage fell to 1.5x EBITDA from 1.8x a year earlier, gross leverage to 2.1x, and NXP repaid $750 million of notes in the quarter. NXP deliberately runs a levered balance sheet to boost returns on equity; the two-year down-cycle was the stress test of whether that leverage could be serviced, and the debt service held while the cycle turned.

That brings the valuation question into focus. At $224 NXP trades at about 19 times trailing earnings and roughly 12.8 times enterprise value to EBITDA, as of late August. Measured against its neighbors the gap looks startling: Texas Instruments trades near 26 times EV/EBITDA and Analog Devices near 26 times as well, both with price-to-earnings ratios close to 40. But that gap is not free money, and it is not new. It is structural, and investors can name each of its sources: automotive — a cyclical, China-linked market — is more than half of sales; a meaningful share of revenue comes from greater China, a demand variable and a tariff risk at once; the company carries the deliberate debt load its analog peers largely avoid; and as a Dutch company its dividends carry a 15% withholding tax for U.S. shareholders. Cheap relative to Texas Instruments is NXP's normal state, not the anomaly. The honest test is NXP against its own earnings. Taking adjusted EPS of $6.66 in the first half, the third-quarter guidance midpoint of $4.11, and a similar fourth quarter, full-year 2026 adjusted EPS lands near $15, which puts the stock at about 15 times this year's earnings while revenue grows 19%. That is a reasonable price for a business with mid-30s operating margins — but it embeds an assumption that the recovery continues.

The dividend is the part of the story the headline overstates. NXP pays $1.014 a share per quarter, about $4.06 a year, a 1.8% yield at the current price. It is well covered — the payout has run near a third of reported earnings, and in 2025 dividends absorbed roughly two-fifths of the company's non-GAAP free cash flow — and it has been unchanged over the past year, management choosing to protect the payout through the downturn rather than grow it. In the second quarter dividends and buybacks together took 45.5% of free cash flow; for all of 2025 the figure was 79.3%. None of this makes NXP an income stock: the yield is small, and the Dutch withholding tax trims it further on top for U.S. holders. What it does is support the value case in a different way. The returns are genuinely being routed to shareholders, and a dividend that stayed covered through a two-year chip downturn is evidence the balance sheet can absorb the next one — a durability datapoint, not a reason to hold the position for yield.

The condition, then, is the recovery itself. The evidence on its side is content-led growth rather than restocking, margins rising faster than revenue, falling leverage, and rising free cash flow. The evidence against it is inventory that still has not normalized — days of inventory on hand ran at 156, some 23 days above the five-year average — and a China-linked auto market whose durability analysts keep probing. After the de-rating, at mid-teens times forward earnings with a covered dividend, the stock has given back the froth of the spring. But a cheap cyclical is only a bargain while the cycle cooperates; if automotive demand rolls over the way it did in early 2025, this is exactly the discount the market is pricing. On three consecutive quarters of content-led, margin-expanding, cash-generating growth, the durable-recovery case earns NXP a place on a value investor's list, sized with the understanding that the gap to Texas Instruments and Analog Devices is earned rather than stolen. The day to step off is the one where content growth turns back into restocking, or the dividend's coverage starts to slip — not the day a beat gets sold.