
CRUS · Technology
The market already discounts concentration risk, but the subtler danger is that content expansion is a double-edged gift — each new Cirrus socket inside an iPhone provides Apple's own silicon team a detailed specification for exactly what a first-party replacement would need to do.
$164.84
$145.00
Exceptional gross margins and compounding switching costs confirm a real moat, but a 94% customer concentration transforms every business quality into a contingent claim on one partner's strategic intentions — quality that exists at someone else's discretion isn't fully yours.
The fabless model is a cash conversion machine — minimal CapEx, a net cash balance sheet, and OCF running well ahead of net income across most years confirm this business funds itself through almost any cycle without external help.
Revenue grows with iPhone content density rather than unit volumes, and the PC and automotive optionality is real but embryonic; EPS compounds faster than revenue primarily through buybacks, which is honest but not the same as organic business acceleration.
The DCF base case lands nearly at today's price, multiples sit at the low end of the historical range, and the FCF yield is fair but not compelling — you're being offered a good business at a fair price, which is fine but not a margin of safety.
A single customer representing nearly all revenue isn't a risk factor on a checklist — it is the business model, and Apple's documented, multi-decade playbook of in-sourcing components it deems strategically critical makes this a binary outcome story with no meaningful second customer to cushion the landing.
Cirrus Logic is a top-quartile semiconductor business that has been priced like an average one, and the discount almost entirely traces back to one number. The gross margin stability across boom and bust cycles is a proof of pricing power that most chip companies cannot show. The fabless model throws off genuine free cash, management returns it without empire-building, and the mixed-signal engineering depth is the kind of capability that takes decades to replicate. The DCF neutral case and current multiples converge near the same number: fair value, not a bargain. The de-rating from historical averages reflects rational market skepticism about concentration, not mispricing. The forward trajectory has two real forces. Content density — smarter codecs, boosted amplifiers, haptic drivers — has been quietly growing revenue per device even as unit volumes plateau, and that dynamic has genuine runway. Against it, the PC and automotive opportunities are architecturally promising but years from meaningful revenue contribution. The SDCA program count tripling is a leading indicator worth tracking, not a current earnings driver. The specific risk that makes all other analysis conditional is Apple's silicon ambition. The company has a flawless, decades-long record of in-sourcing components once it builds the internal capability — processors, graphics, neural engines, cellular modems. Mixed-signal analog has been the exception because it demands a genuinely different engineering culture. But Apple has been systematically closing that gap, and the question is not whether they have the motivation but whether the complexity remains economically prohibitive. The day that calculus shifts, Cirrus Logic loses its largest customer with no meaningful fallback.