Cree reports EPS tonight. I typically find “day of” EPS notes to be a confusing time to call out results / guidance, so I will spare everyone that confusion.
I simply want to share one important takeaway from our checks into the quarter, and also to explain our thesis view.
Single Most Important Factor (for us)
CREE and II-VI are in the process of doubling SiC substrate capacity into year-end, and into the 1H19 (CY). Both companies sell their products in the contract market, and both companies aim to reduce contract prices over time to enable the growth of the market. Spot prices have gone wild, and continue to move higher as tightness in the SiC market remains.
When we did our due diligence earlier this year, the main bear thesis we wanted to vet was the potential for cannibalization or commoditization of this technology thanks to new entrants scaling up capacity. We reviewed every new Chinese entrant we could find, and only one appeared threatening: Norstel-San'an. Norstel was a Swedish startup in the SiC market for many years, with good technology but limited resources. San'an is a giant Chinese LED manufacturer with cash and capacity, who somehow managed to acquire Norstel, despite global rules about the transfer of SiC technology to China (SiC can be used to make some fairly nasty weapons). In our diligence earlier this year we uncovered that San'an had placed an order for manufacturing capacity that, when ready, would equal ~2/3 of CREE installed capacity as of mid-2016. We had handicapped this risk given the lag from order to delivery to high quality output. But it was still a risk factor. Our recent updates indicate that the manufacturing will have to take place in Sweden, not China, and that only about 25% of the initial equipment order will be received in 2019, likely in 1H19 with output in 1H20. By 1H20 we estimate CREE will have nearly tripled their equivalent mid-2016 capacity, making the potential threat of a new entrant much less impactful, and the shift to manufacture in Sweden lowers the risk of technology proliferation in China.
What this means: CREE will likely retain dominance of the SiC market in the next 3-5 years, as far as we can tell at this point.
Thesis Update
There are 3 parts of CREE:
- Lighting – ugly, but improving, and with an accrued warranty reversal in the coming quarters that will drive an unexpected gross margin improvement in the category.
- LED – the worst, literally. However, the LED chips are fabricated on top of Wolfspeed SiC substrate, and the company has recently pivoted output away from the LED segment towards Wolfspeed. Which means that LED revenue/fundamentals will contract faster than anticipated by the market, but it also means that management has a choice in terms of what parts of the LED market they want to support; niche opportunities with good pricing and design lock-in will be supported, but reaching for victory in pure commodity applications will not. AND, as each $ of Wolfspeed revenue is incremental organic growth in a category that will change the world of power electronics, investors give a much higher multiple to WS rather than the commoditized LED segment.
- Wolfspeed – the market is in inning one of transition. The first applications are auto, solar, and fast charging. Application creation is growing and is mainly limited by available supply today. There are still major puzzle pieces in WS that CREE management needs to figure out by the 2020’s, but there is time, and we have confidence that G-LOWE has his finger on the button.
Please call or e-mail with any questions.
Ami Joseph
Managing Director
Yosef Vaitsblit
Analyst
