
SWKS · Technology
The market is treating Skyworks as a beaten-down cyclical approaching a rerating moment, but the more uncomfortable read is that the FCF yield that makes the stock look cheap is itself artificially inflated by investment starvation — and the core customer relationship it depends on is in secular, not cyclical, decline.
$58.70
$105.00
A genuine RF moat built over decades is visibly crumbling — ROIC collapsing from above 20% to single digits in four years is not a cycle, it's structural erosion, and the Qorvo merger is management's implicit admission that standalone Skyworks cannot solve the concentration problem organically at the speed required.
The cash generation is unambiguously real — OCF consistently bests net income, the balance sheet is net cash positive, and FCF margins in harvest mode are remarkable for a semiconductor company; the catch is that CapEx running at less than half of D&A is consuming the asset base rather than maintaining it, and that cannot continue indefinitely without consequences.
Broad markets hitting eight consecutive growth quarters and approaching half of projected revenue is a genuine green shoot, but the mobile core continues to shrink, blended content per phone is already flat year-over-year, and the Qorvo merger — the company's own answer to whether organic diversification was working — implicitly concedes it wasn't moving fast enough.
An EV/FCF below 11x on a business generating 33% FCF margins looks deeply cheap until you remember that FCF is artificially elevated by capital starvation — adjusted for normalized maintenance CapEx the discount narrows meaningfully, but even the pessimistic DCF scenario implies upside from current prices, which is unusual when the bear case is structural, not cyclical.
The risk profile is not diversifiable: 67% customer concentration in a single company that is mid-stream on internalizing the adjacent modem, manufactures in geopolitically contested supply chains, and possesses the engineering budget to solve RF front-end if sufficiently motivated — this is as concentrated a supplier dependency as exists anywhere in the semiconductor industry.
The investment case has a seductive internal logic: a world-class RF engineering organization trading at depressed multiples, with a pending combination that adds complementary power amplifier capabilities, targets meaningfully higher gross margins, and generates over half a billion in synergies. If broad markets diversification into automotive, Wi-Fi 7, and data center timing continues compounding at double digits, and the merger closes cleanly, there is a credible path where this business in 2028 looks nothing like the Apple-dependent supplier the market has discounted today. The FCF yield creates a real floor that limits downside in any scenario short of catastrophic customer loss. The trajectory depends on two things happening simultaneously: mobile content holding at current levels through the next device cycle, and broad markets accelerating fast enough to absorb future mobile erosion. The first is already fragile — management confirmed blended mobile content is flat year-over-year, meaning content-per-phone growth has stalled. The second is real but embryonic — eight quarters of broad markets growth is encouraging, not yet transformative. The Qorvo combination may be the most consequential strategic move in the company's history: building an RF platform large enough to remain technically indispensable even as the anchor customer internalizes adjacent components. The single biggest risk is not the business cycle — it is the anchor customer completing its in-house RF front-end development and removing Skyworks from the bill of materials in a future device generation. At current customer concentration levels, there is no organic replacement for that revenue stream at equivalent margins. The danger is not a gradual fade managed over years but a discrete socket loss arriving with a single architectural refresh announcement, and the window between now and the merger close is precisely the period of maximum structural vulnerability.