"We believe their low frequency PAD is taking market share in China and is ramping faster than the company expected, but Qorvo is facing some yield issues and is in need of an outsourcing filter for this product. Results are largely in-line with our expectations, except for gross margin. We still like Qorvo as we believe the company is on track to take further market share. We continue to be negative on Skyworks as it faces increasing competition." Jun Zhung, Rosenblatt
"In our view, QRVO likely won the iP7 low-band PAD, driving top-line upside, but diluting margins. QRVO utilizes outsourced SAW filters on the product and is seeing increased component costs, impacting margins. Mgmt is looking to achieve "filter independence," which we see taking at least a couple of years. Mgmt was clear these issues are company-specific, not related to RF/analog ASP pressure. QRVO is also undergoing multiple manufacturing transitions impacting yields, which we believe will take 2-3 years to complete. TC-SAW/SAW filters and GaN wafers are transitioning from 4" to 6," BAW capacity is moving from 6" to 8" despite already low utilization, and assembly and test in China is being brought in-house." Rich Shafer, Oppenheimer
These two quotes ran in articles on Qorvo's (NASDAQ:QRVO) 10% dive on Aug 3 in response to the company's 1Q17 earnings call the previous evening. Two critical points are driving QRVO down today, and I believe they hold a message for Skyworks Solutions (NASDAQ:SWKS) investors as well.
• QRVO is investing in expanding in-house filter manufacturing capacity (what CEO calls "filter independence", or no reliance upon external filter supply) through capex on upscaling wafer size in existing fabs and converting an existing GaAs wafer fab to filter fabrication.
• QRVO appears to have taken out SWKS in the low-band LTE PAM in the iPhone 7 "...by diluting margins." In non-CFO-speak, they bombed price to win a socket. In consumer electronics, bombing price is a non-recoverable event; you don't get to raise prices later to recover GM, and you don't get a pass from the 2.5%/qtr price reductions demanded by OEMs. And lowering your price resets price expectations for you and your competitors.
In the early 2000s, it was a great time to be a GaAs wafer manufacturer. The GaAs HBT (heterojunction bipolar transistor) had replaced the silicon CMOS FET as the go-to transistor for mobile phone RF power amplifier manufacturing. 2G had taken cellular into the digital world, and 3G was rolling out with the promise of "data everywhere." Mobile phones were the center of the zeitgeist - everyone had to have one, and Scandanavian and Japanese consumers had to have at least 2. TQNT and RFMD traded in the $100 range on the promise of a greenfield market for mobile phone PAs that only GaAs wafer manufacturers could supply.
Then the transition "from 2G to 3G" evolved into "2G and 3G", increasing mobile phone bill of material cost and setting off a pricing war on 2G sockets. That multi-year price war impacted GaAs wafer fab gross margins and took stock prices down into the $5-15 range where they languished until 4G created a new PAM growth segment.
Today, annual growth in mobile phones has shrunk to single-digit rates. 5G is at least 5 years from significant network deployment. Filters dominate the 4G PAM bill of materials, and 5G will continue that trend, so GaAs manufacturers are expanding filter fabrication to meet projected demand. While GaAs PA die are less than 25% of the RF PAM bill of materials cost, designing and fabricating GaAs PA die requires far more expensive headcount and opex than designing and fabricating filters.
How does a manufacturer feed significantly increased factory capacity in a low-growth market? By taking share from competitors. And how have RF PA vendors taken share from competitors in the past? By bombing price to win sockets - because RF PAM functionality does not enable value-added feature opportunities.
And so we reach QRVO's 10% plunge in market value on the back of increasing their revenue forecast. They are increasing revenue by decreasing GM, through taking a low-value low-band 4G socket from SWKS. Low-band 4G PAMs use TC-SAW and SAW filters, not BAW filters, so little-to-no performance-based pricing opportunity.
This, of course, incrementally decreases SWKS's GaAs and filter manufacturing capacity utilization. Given this socket loss is at AAPL and not some Tier 3 Chinese phone pop-up vendor, the decrease in fab capacity utilization will be significant to SWKS GM. Not devastating - significant.
What does SWKS do in response? Underprice QRVO at Samsung and Xiaomi. What drives that? Well, Samsung and Xiaomi procurement will know to the fraction of a penny how QRVO priced the low-band 4G PAM to AAPL (iPhone is built by Taiwan's Foxconn at Chinese factories...).
Thus the next gross margin devastation in the GaAs business begins. QRVO bombing price begets SWKS bombing price, which begets QRVO bombing price - all while both companies work to increase filter manufacturing capacity. And what follows GM devastation on a high opex business like GaAs wafer fabrication-dominated SWKS and QRVO? Net income and valuation devastation.
And what happens to AVGO? Nothing. They don't mess about in low-band PAMs.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I tend to be negative in the near term on SWKS and QRVO, and positive on AVGO.