Jan 26 2022 - Nvidia Stock Dips: Losing An Arm Will Not Kill It

Summary

• NVIDIA is reportedly walking away soon from its Arm deal.
• But, we don't think the deal is a make or break acquisition for NVIDIA.
• Moreover, the sell-off has brought NVIDIA stock back into its fair-value zone.
• We discuss why we think investors should buy on weakness.
• 

Justin Sullivan/Getty Images News

Investment Thesis

NVIDIA Corporation (NVDA) took center stage yesterday as Bloomberg reported an insider scoop over its troubled Arm acquisition that it could be called off soon. It reported that NVIDIA has informed its "partners that it doesn't expect the deal to close." The stock also tumbled in pre-market when the news was announced. However, NVIDIA continues to maintain its public confidence over the merits of its proposed acquisition as it emphasized (edited): "We continue to hold the views that this transaction provides an opportunity to accelerate Arm and boost competition and innovation."

However, we believe that we are not the only NVIDIA investors in town that have already discarded the notion that it will satisfy regulators across multiple jurisdictions. Moreover, Street analysts that we follow have also significantly discounted the potential completion of the deal since the US government filed its lawsuit to block it.

Nevertheless, with or without Arm in its pocket, we believe the NVIDIA growth story still has a clear runway ahead. We discuss why investors should consider the long game here with NVDA stock, and buy while it's weak.

Losing an Arm is No Big Deal

NVIDIA's close-knitted partnership with Arm as a licensing partner has worked tremendously well. But, NVDA wants a closer relationship. So, it saw the potential opportunities from owning Arm's IP. CFO Colette Kress emphasized:

Arm with NVIDIA is a great opportunity for the industry and customers. With NVIDIA's scale, capabilities and robust understating of data center computing, acceleration in AI, we can assist Arm in expanding their reach into data center, IoT and physics and advance Arm's IP for decades to come. The combination of our companies can enhance competition in the industry as we work together on further building the world of AI. (NVIDIA's FQ3'21 earnings call)

But, regulators have raised valid anti-competitive concerns given the criticality of Arm's architecture being used by NVIDIA's customers and competitors. While NVDA has contended that Arm architecture is nowhere near the dominance of x86, it has not gone down well with the US, UK, or even the Chinese regulatory authorities. Furthermore, even if it fortuitously passed the US regulators' review, the Chinese government was said to be inclined to shoot it down, according to Bloomberg. It added (edited): "The acquisition also faces resistance in China, where authorities are inclined to block the takeover if it wins approvals elsewhere. But they don't expect it to get that far."

CCS Insight also emphasized that investors shouldn't be surprised that NVIDIA intends to walk away from the deal. It added (edited):

The NVIDIA-Arm deal has faced intense scrutiny and pressure from the start and it’s no surprise the deal is in danger of collapse. Finding a way to appease regulators whilst maintaining the value and justifying the billion price tag has proven overwhelmingly challenging. (TechCrunch)

The company has to forego B it paid upfront (including a "break-up" fee of .25B). But if the revisions required by the regulators have "devalued" its acquisition to a significant extent, then NVIDIA should retain the status quo. It can continue churning out its custom CPUs, and maintain business as usual.

Bernstein also added that NVDA could walk away safely as it added: "While owning the asset could have been wonderful, we don't believe they had to have it either. The deal could have helped NVIDIA's push into data-center chips, but the company 'presumably can and will continue their stand-alone efforts."

Therefore, we believe that NVIDIA investors should consider that the deal is as good as dead. Even the Chinese government didn't think it needed to step in to block the deal. Sources also pointed out that SoftBank (OTCPK:SFTBY) is mulling over an IPO for Arm to encash its investment. Hence, we believe that it's time for us to move on. If there's a slim chance that it came back alive again, then consider it a bonus.

NVIDIA Stock Also Suffered the Brunt of the Recent Sell-off

NVIDIA & peers stock performance

Koyfin

NVIDIA stock & peers EV/NTM EBITDA valuations. Data source: S&P Capital IQ

Readers can quickly glean that both AMD stock and NVIDIA stock were battered significantly, dropping more than 20% since the year started. Broadcom stock followed closely after its post-earnings momentum spike was digested, losing 18% over the same period. QUALCOMM stock was also in the red with a -7.7% return. Only TSMC stock and Intel stock stayed above water. We don't think there should be any surprises here, given NVIDIA's premium valuation among its peers. It is trading at 46x NTM EBITDA, way above the rest of its peers, including AMD stock. Therefore, investors shouldn't be surprised that its stock also suffered the brunt of the recent selling. But does that mean investors should stay away from NVDA stock just because it's trading much higher than Intel stock, for example? We don't think there's a straightforward answer to this question. But, we have high conviction over NVDA's strong underlying thesis and management's ability to execute consistently with aplomb. The market is willing to ascribe such premium to NVDA stock over the last three years simply because its execution has been excellent. So, NVDA stock hasn't gotten its premium rating only recently. Its 3Y NTM EBITDA mean of 44.6x should give you a clear indication of the market's confidence. Moreover, it has outperformed most of its peers listed above by a wide margin.

NVIDIA stock & peers stock performance

Koyfin

Readers can quickly glean NVDA stock's outperformance in the chart above. We think it's straightforward. Would you choose NVDA stock that delivered 477% or one that offered a paltry 4.1% gain over the last three years? But, its 3Y EBITDA mean is 44.6x. So, NVDA stock has never been cheap. But, it has not stopped it from outperforming its peers. Furthermore, we believe that its opportunities across multiple secular themes and verticals are just getting started.

NVIDIA's A100 GPU Penetrated Just 10% of Data Centers

NVIDIA believes that all data center workloads need to be accelerated. But, its A100 Tensor Core GPU has penetrated only 10% of data center workloads. Given that we are gravitating towards more AI workloads, including ML and DL, the need for NVIDIA's A100 will become even more critical over time. Moreover, a recent presentation by the #1 smartphone SoC leader (by shipments) MediaTek (OTCPK:MDTKF) showed that AI business opportunities could increase to T by 2037, from T in 2020, representing a CAGR of 17%. MediaTek Chairman MK Tsai also articulated: "We often overestimate short-term benefits and underestimate long-term business opportunities." He believes that the semiconductor industry will pivot towards capturing opportunities based on capabilities in AI, including deep learning workloads, well into the future.

Therefore, all this requires even more power that NVIDIA has highlighted that the current data center CPU cannot manage. What makes its A100 so formidable is its ability for deep learning training and inference concurrently. Kress articulated (edited):

It really indicates that there is still a 90% opportunity going forward as people really see the importance of acceleration as Moore's Law nears the end for many of these applications. Our A100 architecture allows both hyperscalers and enterprises to address the needs they have. But, more importantly, A100 has the ability to address deep learning training and inferencing at the same time. Our work with customers, helping them with their inferencing solutions, helping them with their training models, as well as all of the different software that we have enabled for these industries really speak to the growth that we're going to see in terms of the industry going forward. (24th Annual Needham Growth Conference)

TSMC Chairman Mark Liu also shared at a recent forum, as DigiTimes reported (edited): "The new business opportunities from innovative applications will make semiconductor chips ubiquitous. As TSMC chairman Mark Liu noted at the forum, in the past, we followed Moore's Law in a tunnel of darkness, but now we can see the brightness and ascertain tremendous business opportunities ahead at the entrance."

TSMC reported a highly impressive FQ4 report card, which also saw its stock hit its all-time high recently. It guided for a robust FY22 topline growth of 29% YoY, ahead of its peers. TSMC has focused about 80% of its CapEx in the leading process. Therefore, it's crystal clear that high-performance computing (HPC) and 5G applications will continue their sustainable and robust growth moving forward. If that's not enough, then TSMC's substantially raised record-high CapEx of -45B for FY22 should inform investors about the growth in HPC and 5G moving forward. Notably, "70-80% of which will be utilized for the foundry's advanced technologies including 2nm, 3nm, 5nm, and 7nm process nodes."

Meta Platforms, Inc. (FB) shared recently that it will be developing what it believes is the "fastest AI supercomputer in the world when it's fully built out in mid-2022," codenamed AI Research SuperCluster (RSC). It is critical to Meta's ultimate goal of creating what it believes is "the next major computing platform - the metaverse, where AI-driven applications and products will play an important role." Meta investors should know by now that the company has dedicated billions in additional CapEx for its metaverse ambitions over the next few years. The company recognizes that it needs to build its supercomputer to scale towards its vision. Given NVIDIA's leadership in AI data center GPUs, we believe there's simply no other better partner for Meta to continue on its journey. Meta was also effusive in its praise for NVIDIA as its AI accelerated computing partner. It added (edited):

 

All this infrastructure must be extremely reliable, as we estimate some experiments could run for weeks and require thousands of GPUs. The entire experience of using RSC has to be researcher-friendly so our teams can easily explore a wide range of AI models. A big part of achieving this was in working with a number of long-time partners, all of whom also helped design the first generation of our AI infrastructure in 2017. NVIDIA provided us with its AI computing technologies featuring cutting-edge systems, GPUs, and InfiniBand fabric, and software stack components like NCCL for the cluster.

- Meta Platforms

Think about it. Is there another company with a full-stack AI accelerated computing infrastructure to help companies like Meta power their metaverse ambitions? We believe that every company will need to have a metaverse strategy moving forward. But, without NVIDIA's AI infrastructure, the journey might get a lot harder, if not impossible. It isn't only NVIDIA CEO Jensen Huang who recognized that its accelerated computing dominance is a full-stack leadership. Even Meta emphasized it, as seen above.

Omniverse Could be a 0B Opportunity for NVIDIA

We shared in a previous article that NVIDIA's AI Enterprise is a B opportunity. We also worked out that NVIDIA's Omniverse engine could be a B opportunity over the next few years. As more companies and creators develop strategies for their virtual worlds, we believe that Omniverse will pick up momentum. NVIDIA is also encouraging the adoption of its Omniverse engine by making it free to creators using its RTX series graphics. However, the RTX upgrade cadence has reached only 25% of its installed base, so we are still early in the upgrade cycle. NVIDIA emphasized that a typical upgrade cycle peaks at about 50%. Moreover, NVIDIA highlighted just 100K of creators had downloaded its Omniverse engine. Given its estimates of 40M creators and designers who could be using Omniverse, we are still in the early innings of its massive opportunity.

UBS even highlighted that it thinks NVIDIA's Omniverse opportunity could be worth more than 0B over the next few years, given its transformative nature and channel checks. It added (edited):

We conclude that Omniverse is in some ways similar to CUDA as a platform to reduce friction and ultimately drive market expansion for NVDA's hardware solutions; for the virtual world, it is in some ways similar to the first browsers from the mid-1990's for a fledgling internet. Omniverse could potentially open the next ~0B+ in TAM for the company as the 3D/virtual work takes shape. (StreetInsider)

 

NVIDIA Stock is a Buy Now

NVIDIA stock EV/NTM EBITDA 3Y mean

TIKR

Readers can glean and observe that NVIDIA stock is back trading near its 3Y NTM EBITDA mean of 44.6x. NVDA stock has consistently found support near the 40x NTM EBITDA multiple over the past year. Moreover, our blended implied fair value estimates also point to a stock that seems fairly valued now (+/- 10% implied upside to fair value).

Therefore, we believe the market sell-off has created another opportunity for investors to add exposure to this fantastic AI computing leader.

As such, we reiterate our Buy rating on NVDA stock.