By K C Ma and Bilal Hashmi
Following Intel’s (INTC) lower second half guidance, a disappointing Nvidia's first quarter earnings result (ER) may be a foregone conclusion. Technically, there has been no good news for Nvidia since its fourth quarter earnings release. But there's evidence that the market looked beyond the immediate negativity (Point B Figure 1A) but saw a better second half 2019 (Point A in Figure 1A). Before the recent surprising new tariffs, the bull thesis is that organic gaming demand after crypto fades out will easily overcome the fading macro impact from China slowdown. First quarter consensus is mixed. The sales are expected at $2.2 billion which is 31% lower than in 2018. With 1Q gross margin seen at 59% and EPS of $0.81, the second half is projected to get better as 2Q sales are guided to $2.54 billion and EPS to $1.12 (Table 1). More importantly, Nvidia’s over 30X PE already has priced in a better 2019.
The (China) Gaming Outlook
The delay in the U.S./China trade talks and the resulting imposed tariffs obviously threw a monkey wrench to the sales of Tuning GPUs to the China market. Such sales were considered the first test of Nvidia’s ability to include lower-priced chips without ray tracing in its platform. The success of the test was supposed to clear the $1.6 billion overhanging inventory and drive the sequential growth after 1Q into 2H 2019. Gaming is expected to decline by 20%. Only Turing is significant enough to turn the growth direction around.
As gaming still remains the largest revenue contributor (43%), crypto mining demand has become a serious distraction to organic gaming demand in recent years. But even Nvidia has curbed its enthusiasm after the recent crypto collapse. Over time, it has been estimated that graphics card sales for crypto mining only account for about 10% of revenue at AMD and about 6% at Nvidia. In one of its recent earnings reports, Nvidia CEO Jensen Huang called crypto exposure a “small, but not zero, part of our business” after estimated revenue for chipsets used for mining fell by half to about $70 million in the third quarter from the second quarter.
It's only reasonable to expect the same crypto upside trend in 2017 may reverse itself in 2019. Particularly, the crypto effect will likely create the same seasonal trend for Nvidia in the first half 2019. A year ago, with first quarter stronger vs. 4Q18 and 2Q weaker than 1Q, sales got a $289 million boost from cryptocurrency SKUs, driving 1Q's 66% growth. Crypto impact was up from 4Q18's $200 million, or 7% of sales, to 9% of 1Q19 sales. As a result of the reversal, 2Q19 and 3Q19 sales fell to 3% of guidance. With the CEO’s pre-4Q warning, the remaining crypto demand should barely register in 2019 (Table 2). With or without the lingering crypto demand, a strong organic gaming demand always is needed to fuel strength but it seems unlikely to happen in 2H 2019 due to the unexpected process in trade talks. The net effect is that gaming sales are poised to drop about 20% sequentially in 1Q. An additional unintended complication of the gaming GPU pullback also may reflect that AMD GPUs have become a serious contender to Nvidia in 2019.
The Data Center: China Connection
Intel’s 1Q miss on Data Center Group sales may prove to be equally damaging to Nvidia’s 1Q data center sales. Intel cited the reason as China slow demand which can only be worsened for the rest of 2019 after new tariffs being imposed. Just a year ago, Nvidia’s data center triple-digit growth rates have been the “white knight” to save the day. As data center use of GPUs and widening PC gaming are long-tailed trends, sales growth rates have not been expected to dampen as much as expected in fiscal 2019. Yet, Nvidia’s data center has developed its own growth problem in recent quarters (Table 2). As the industry-wide cloud spending cut possibly from the trade-induced recession has filtered through the supply chain, data center sales growth slowed to 71% from 100%-plus over the prior seven quarters.
After 4Q, probably the only good news for Nvidia’s data center is Nvidia’s $6.9-billion acquisition of networking chipset and technology provider Mellanox (NASDAQ: MLNX). The acquisition has created synergies in high-performance computer and long-term partnerships. Nvidia will substantially increase its leadership in GPU-assisted HPC with its Tesla accelerator cards connected using the NVLINK technology within a server and Mellanox technology to connect between servers. Nvidia’s GPUs can be found in 127 of the systems on the Top500 supercomputer list and Mellanox’s interconnects in 265 of them. Nvidia also has common customer engagements with top server OEMs and with the U.S. and Chinese cloud giants running large hyper-scale data centers. There's an obvious synergy from cross selling on each company’s technology assets and customer bases.
Before the merger, Nvidia's Datacenter segment, with a 52% growth rate in Q1 2019, is the fastest growing and most profitable revenue segment. With an Intel-like DC gross margin 50%, Nvidia’s DC revenue share has increased rapidly from 10% to 30% since 2016 (Figure 2A). Aided by Mellanox’s new customer base, Nvidia becomes a serious contender for DC market share. Since 2016, Nvidia’s DC market share against Intel has risen from 3% to 10% and looks to increase to 12% by 2020 (Figure 2B).
Overhang Inventory Must Abate
Obviously, investors will watch Nvidia’s 1Q inventory like a hawk. The crypto-induced overhanging channel inventory has built up Nvidia’s own inventory. By the end of 2018, Nvidia’s inventory ratio has reached a record high of 71% of revenue (Figure 3). Fortunately, the crypto activity has fallen as fast as it has risen. However, after three difficult quarters, the channel inventory was expected to be cleaned up before 2H 2019 from the supposed recovery of China gaming demand. Although 1Q inventory should come down, the timetable for Nvidia’s overhang inventory to completely abate will be inevitably delayed due to the recent development of China trade talks. When and if that happens, the inventory ratio will drop back to a long-term average of 45% by 2021 (Figure 3).
Reasonable 1Q ER Expectation
For Nvidia’s 1Q ER, it's reasonable to expect that both gaming and data center will continue to see some softness in the next 2 quarters and the larger gaming’s recovery may be guided beyond 2H 2019 due to a lack of China trade resolution. Since cloud spending cut has not yet been reversed, it's gaming that carries the day. Data center recovery also will remain subdued by the end of fiscal 2020. Similarly to Intel, Nvidia 1Q data center sales growth will be significantly lower both sequentially and from a year ago. Although Mellanox’s acquisition allows Nvidia to execute a major data center market share play, there's less immediate revenue impact. Nvidia’s DC market share can easily increase to double digits by 2020 with an expected $1.4 billion contribution to the top line. The overhang channel inventory from the crypto activities should be significantly reduced but not over in 1Q, and the inventory ratio should return to normal range when and if there's any positive relief from a China slowdown and cloud-spending cuts.
In terms of ER price moves, as Nvidia is the last major tech firm to report, most information already is publicly available. Though, the question remains to what extent the current Nvidia price already reflects public information. In Figure 4, there may be some clues to answer that question. Since Intel’s 1Q ER, Nvidia’s price has moved more closely with Intel’s price than with AMD’s, suggesting both Intel and Nvidia already have responded to the common factors in their respective 1Q ER and the 2H 2019 outlook, e.g., data center sales growth and the renewed China impact. My guess is that, even if Nvidia reports a similar miss as Intel, there should be little downside left. In other words, Nvidia’s stock has more upside than downside after the 1Q ER.
Disclosure: I am/we are long NVDA. 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.