- Nvidia continues to exhibit momentum on strengthening gaming trends.
- Furthermore, there are some additional catalysts to OEM channel revenue.
- However, getting aggressive purely on historic price performance is inadvisable.
Investing into NVIDIA (NSDQ:NVDA) at this point may sound a little risky. With the company performing better than buy-side expectations in the current fiscal year with almost a doubling of the stock price, it’s hard to imagine how this outperformance will continue. Surely, there must be some sort of upper ceiling to the stock returns, right?
What's causing investors to buy NVIDIA stock?
After looking at the recent quarterly report and continued momentum following the report, I think it’s safe to take some money off the table, but it’s still worth maintaining a position in the stock given the on-going revenue ramp of Pascal GPUs and the lack of inventory in the retail channel for the company’s current flagship GPU (the GTX 1080). I’ve been recommending the stock for quite a while now, and I believe investor perception has shifted given the low unit penetration into enthusiast PCs and the on-going sales ramp of the Pascal family.
That being the case, I believe the long-term growth drivers are still intact, and while refresh driven demand isn’t always perceived favorably by investors it’s worth noting that the stock will continue to outperform as I anticipate enthusiasts to continue refreshing their older hardware with Pascal series graphics.
Furthermore, it’s been a while since we’ve seen broad based recovery in the desktop space, but a great leading indicator is enthusiast demand, and with enthusiasts refreshing GPUs it’s only a matter of time before the OEM retail channel sees an uptick in demand as the current desktop and Notebook installed base has aged five to six years. Assuming the OEM channel starts to exhibit momentum, it will likely happen in FY’17, which gives NVDA another pillar by which investors can rally around, as the shipment of discrete units to enthusiasts was the primary catalyst in the current fiscal year.
The most interesting data point in the current quarter was the reported revenues from the gaming segment (q/q growth of 12%) as opposed to OEM and Datacenter. While there are promising applications for CUDU within the datacenter, the impact isn’t too substantial. But I do anticipate the segment to outperform the unit shipment growth of Intel despite on-going competition for custom silicon, i.e. FPGAs and Co-Processor units.
Does the GTX 1070 and 1080 provide enough upside?
What was interesting about the most recent quarter was the release of the GTX 1070 and 1080. The release of the 1080 was on May 27th, whereas the 1070 was released on June 10th, so the impact was limited to a single month. While gross margin figures were above consensus, this was due to the higher ASPs driven by the founder’s edition models. It’s likely that margin comps will eventually stabilize at a lower level given the launch of various 3rd-party cards to the Pascal line-up (lower ASPs from third-party cards when compared to founder’s edition).
However, the unit shipment trends will quickly accelerate going into the next quarter driven by the full-quarter sell-in of 1080s and 1070s, wafer supply stabilization, and diminished competition at the high-end of the market as AMD strictly targeted the entry-level enthusiast category.
So, even with the launch of the GTX 1060, I’m anticipating that Nvidia will lose some share in the mid-end of the market, but it won’t translate negatively on sales/margins as the flagship launch of the Titan and on-going production ramp of high-end units should translate into healthy revenue and profitability for the duration of the current fiscal year.
Here’s what Jen-Hsun Huang (CEO of Nvidia stated on the call):
Well, our guidance is our best estimate, and we'll know how everything turns out next quarter when we talk again. But at some high level, I would agree with you that as we move further and further, and more and more into our platform approach of business, where our platform is specialized and rich with software, that increasingly the value of the product that we bring has extraordinary enterprise value, that the benefits of using it is not just measured in frames per second, but real TCO for companies and real cost savings as they reduce the number of server clusters, and real increases and real boosts in their productivity.
Jen more or less dodged the question by the analyst from BMO Capital, as he wanted more direct commentary on the flat gross margin outlook. But, essentially the flat margins are driven by the drop-off in pricing as more third-party 1070 and 1080 units enter into the retail channel. Since the ASPs of the Pascal generation is significantly higher than the pricing of Maxwell series cards from the prior cycle launch, I believe the company is indicating that the margins will be great on an y/y basis, but further margin expansion remains unlikely as pricing declines will reduce gross margin expansion. The founder's edition cards are priced at a slight premium and are soon exiting the retail channel given limited availability.
The logic seems fairly consistent, and it’s why I’m not going to model improvement to gross margins and net profit margins. Sure, the yields from the new 16nm fab from TSMC will likely improve, but it’s not enough to drive substantial q/q gross margin gains.
Furthermore, analysts are hesitant to re-rate the stock higher.
Rick Schafer maintained his “perform” rating, which is equivalent to a hold rating. The analyst is one of the most reputable within the semiconductor space, and while he doesn’t offer a price target, it’s worth noting that NVDA’s price has tripled in the span of 12-months.
So, adding to your positions sound too risky. Reducing some exposure sounds logical, however maintaining a reasonable allocation still makes sense given the sustained earnings/ sales ramp of data center, autonomous vehicles, and PC OEM channel recovery. The GPU channel will have a one-time effect on earnings/revenue, but sustained growth in a major cyclical launch to the tune of 30% sustained gaming unit growth is unlikely beyond the next quarter.
Here’s what Schafer from Oppenheimer Co. mentioned in his note:
NVDA reported F2Q (Jul.) sales/EPS of $1.43B/$0.40, easily topping consensus $1.35B/$0.37, primarily on core gaming strength. Management's F3Q (Oct.) revenue outlook of $1.68B (+17% Q/Q) was even more impressive, besting the Street's $1.45B. NVDA trades at 27x (ex-cash) our new CY17E EPS of $2.00, a premium to peers’ 21x. We applaud mgmt. execution as the company shifts focus toward strategic platforms and away from PC/OEM. At these levels we view risk-reward as balanced, however and we remain on the sidelines for now.
I honestly believe management guidance is still conservative going into the next quarter, as it wasn’t going to take much for the company to top the consensus estimate range in Q2'17 either. As such, analysts are revising estimates higher, but anticipating sustained out performance for the duration of the year seems unlikely even if the company were to blow out earnings in the next quarter.
NVIDIA stock's recent momentum isn't sustainable even if there is an earnings beat, and anticipating another doubling in valuation would be dangerously unrealistic. However, shorting the stock is also inadvisable given the broad based multiple expansion of stocks in recent weeks, and the sustained premium Nvidia will likely trade at – given numerous levers for growth.
Furthermore, I will update my financial model and provide these added insights over the next couple weeks. Despite impressive earnings results, and likely earnings beat next quarter, I’m no longer reiterating my high conviction buy recommendation and am now lowering NVDA to a buy rating.