Image source: NVIDIA.
Generally speaking, I'm not a huge fan of Citron Research, the often short-oriented commentary site run by Andrew Left. I consider Citron one of those market prognosticators that makes constant calls, and on occasion, some of them turn out to be fantastic, well-researched picks. And the street credit earned from a subset of correct calls overshadows the incorrect calls based on poor theses that no one pays attention to. Regardless, Citron certainly does have the power to move the market in the short term, which incidentally helps its positions and becomes a self-fulfilling prophecy -- at least for a day or two. (I've always strongly suspected that Citron front-runs its own picks.)
That being said, I don't necessarily disagree with Citron's latest note, which targets high-flying graphics specialist NVIDIA (NVDA +1.78%). Following an incredible year of gains, Citron took the company down a notch yesterday with a single tweet, with shares falling 7%.
The $4.3 billion tweet
NVIDIA's drop wiped out $4.3 billion in market value, so yeah, Citron can move the markets. In the tweet, Citron pointed to six risks that it believes NVIDIA investors are discounting, which combined with the soaring valuation this year could be a recipe for a pullback.
Citron readers know we have long been fans of $NVDA,but now the mkt is disregarding headwinds. In 2017 we will see $NVDA head back to $90 pic.twitter.com/n4U7f6eV7G
— Citron Research (@CitronResearch) December 28, 2016
I don't disagree -- just a couple weeks ago I expressed some concern over stretched valuation multiples, even while acknowledging that NVIDIA has executed extremely well on many of its growth initiatives. Still let's dig through Citron's concerns.
Growth and market share: It's true that a lot of NVIDIA's top-line growth has been driven by market share gains, mostly from primary GPU rival AMD (AMD +1.19%). AMD clawed back a little bit of market share in Q2, as NVIDIA had been gaining ground for a string of quarters. But AMD gave that back in Q3 following NVIDIA's launch of Pascal-based products. Remember that the PC market hasn't been a growth market in years, nor has the PC gamer niche within it, and the core GPU business accounts for over 80% of sales. The best ways to grow revenue under those circumstances are to steal market share or increase prices; NVIDIA is doing both. NVIDIA is growing into new addressable markets, just not by much as those businesses are still very small compared to the GPU segment. I don't see this as a "risk." If anything, I see it as evidence of strong execution.
Data center competition: NVIDIA has made some progress in the data center, which it hopes will be a growth avenue going forward. There is potential, but Citron is right that competition will only intensify further. This is a relatively small part of the business currently, so while the downside risk may be somewhat limited if competition starts to eat into NVIDIA's results, the issue is valuation. If investors are pricing in data center growth that fails to materialize, shares could be punished from multiple contraction. At the very least, competition could erode gross margin, and the data center business is quite profitable relative to NVIDIA's overall corporate average (56%).
Intellectual property and licensing revenue: NVIDIA and Intel (INTC +3.21%) signed their patent cross-licensing deal back in January 2011, settling years of litigation between the two companies. It's true that Intel's royalty payments will end soon and that revenue will be missed, but investors have known about this for half a decade. This should all already be priced in by now, and you can't argue that NVIDIA's valuation is predicated on royalty revenue. NVIDIA collected $66 million in licensing revenue from Intel last quarter, or just 3% of sales.
Gross margin sustainability and competition: As mentioned above, concerns around gross margin are completely valid, given the expected intensification of competition in many of the growth areas that NVIDIA is enjoying right about now. Intel's vertical integration with silicon fabrication does give it cost and strategic advantages, which it will surely leverage.
It wouldn't be particularly surprising if NVIDIA does pull back to $90 in 2017. Stocks that run up as fast as NVIDIA has in 2016 often do pull back if the market got ahead of itself, which is entirely possible given the significant multiple expansion that NVIDIA experienced this year. Some of Citron's concerns are worth acknowledging, but others are probably already priced in.
