A Holistic View Of Nvidia – Life After Crypto Tumble, Growth Prospects, Valuations, And Risks
Summary Nvidia’s share price has taken a 50% haircut since late 2018, we think the stock is worth a look at nearing a 2-year low. We discuss the four growth aspects for Nvidia, including the trend in AI, acquisition of Mellanox, future of gaming, and inventory management. We conduct […]
Click here to view original article: A Holistic View Of Nvidia – Life After Crypto Tumble, Growth Prospects, Valuations, And Risks
Nvidia's share price has taken a 50% haircut since late 2018, we think the stock is worth a look at nearing a 2-year low.
We discuss the four growth aspects for Nvidia, including the trend in AI, acquisition of Mellanox, future of gaming, and inventory management.
We conduct valuation using a DCF analysis and an earnings multiples analysis to determine the ideal entry price.
Finally, we look into the potential risks that investors need to keep in mind.
Research performed by Zhicheng (Jason) Zhang at DX2 Capital.
Nvidia (NVDA) needs no further introduction, the GPU giant was a Wall Street darling for a stretch of almost 4 years, from 2015 until late 2018. During that period, NVIDIA’s stock price soared by more than 1300%, from $20 (01/2015) to $280 (10/2018). The stock was unstoppable, and every quarterly earnings seemed to be a beat and raise. Its business was taking off, fueled by strong demand and the growth in gaming, AI computing, self-driving, and cryptocurrency.
Numbers then started to cool off in November 2018, when NVIDIA reported a Q3 revenue miss and Q4 guide down (expected revenue $3.3B down to $2.7B), followed by another guide down in January (from $2.7B to $2.2B). The stock then plummeted to a low of around $130 in January 2019 when the crypto bubble busted.
Since then, NVIDIA has reported two uninspiring but tolerable quarterly results. Investor sentiment has been cautious as the full-year outlook is unclear, and the ongoing US/China trade war obviously does no good to the stock price. A combination of negative sentiment, cautious outlook, and low expectation has taken Nvidia’s price near a 2-year low. Given the strong fundamentals of Nvidia’s business, we believe the longer-term prospect is solid, and the stock is worth investigating at the current price level.
We see four driving engines for Nvidia’s future recovery and growth:
1. As the dominant provider of GPU hardware, Nvidia will be able to ride the wave of artificial intelligence and enjoy growth in the next several years.
Artificial Intelligence and Machine Learning have played an important role in how business and society operate. According to a survey conducted by Cisco System in 2018, more and more companies have expressed that they are reliant on AI and ML.
The result of this is the continuous fast growth of spending on AI. It is a consensus estimate that the revenue of AI will continue to grow in the future. The growth rate of AI market for enterprise application has experienced a 134% and 93% increase in 2017 and 2018, while it is forecasted that the CAGR growth rate from 2016 to 2025 will be 64%.
With the high growth of AI, the hardware revenue driven by AI will also increase at an impressive rate. The market revenue for hardware will grow from $19.64 billion in 2018 to $234.6 billion in 2025 with a CAGR of 43%. GPU hardware will have a CAGR of 34% while other kinds of chips such as CPU, ASIC, FPGA, and SoC accelerator combined will have a CAGR of 48%
Due to the concerns of power consumption and efficiency, major players are spending more and more in exploring more specific task chips such as ASIC and FPGA, which requires significant up-front cost but are programmable and more aligned with the algorithm process. However, due to restrictions of adaptability in the change of process, it is expected that these chips are still majorly used as accelerators and GPU and CPU are still very essential for data center computation.
Source: Jon Peddie Research
Nvidia has established a lead in the GPU industry in the past years. The market weight for Nvidia GPU has been ranging from 60% to 80% where Intel (INTC) and AMD (AMD) make up for the rest. The following graph from Intento compares the GPU performance for each product family in each brand:
The GTX, RTX and Tesla families are from Nvidia, and Radeon is from AMD. Nvidia offers a more comprehensive and broader product spectrum for choices from high-performance to low-performance graphics cards on all ranges. The dominant position also comes from the software support of GPU in deep learning and algorithm training. While Nvidia’s product family supports almost all DL frameworks and none of AMD GPU are supported. This gives Nvidia’s GPU much superior compute performance for convolutional networks and making it the first choice for its customers.
Another area of high growth in AI for Nvidia comes from autonomous driving. The projected market size for autonomous driving features is estimated to grow at a CAGR of 14% from 2016 to 2025.
Currently, Nvidia offers business solutions for hardware, software, and simulations for Autonomous Vehicles. Its DRIVE APX 2 platform offers level 2+ automated driving feature, and according to the company’s press release, it has made good progress by partnering up with major carmakers in the field. Audi, VW, VOLVO, Mercedes-Benz and TOYOTA have all announced partnerships with Nvidia with the first three plan on deploying DRIVE APX 2 for their cars AI platform. The company has also announced a partnership with Baidu, a leader in AI, to develop a cloud platform for autonomous vehicles.
The auto revenue disclosed by Nvidia experienced an increase in the middle of Q2-Q3 of 2019 but a slowdown in the last two quarters. The increase was a result of the company’s Xavier SoC that was announced in 2016 and later deployed in early 2018. Q2 2020’s revenue on Auto is going to be a slight increase as Nvidia is making steps in partnering up with carmakers and developing new techs. However, since there are no new game-changing products, numbers are unlikely to change dramatically.
While the auto revenue projected by Nvidia is developing at a slower pace, with the fast growth of the industry, it is likely that the financial results will pick up in the future as R&D becomes matured and commercialized. The company has announced NVIDIA DRIVE AutoPilot, a level 2+ driving system, commercially available in 2020, and road-testing S-class Mercedes on California roads in 2019. As Nvidia combines hardware and software together in the AV industry, it has increased its ability to profit from the expansion of the industry, and it certainly can harvest in the future as more AI-based technology becomes commercialized.
2. Nvidia’s CUDA AI platform and acquisition of MLNX enable it to provide a more comprehensive solution to customers.
Both GTX and Tesla product families offer very similar GPU power, but Tesla is much more expensive than the GTX family. So many universities and organizations choose to use GTX GPU because of budget concerns. So Nvidia announced that it will no longer support CUDA software on GTX GPUs, which then forces many organizations to buy much more expensive Tesla GPU. This is a demonstration of Nvidia’s monopoly market position.
When comparing to AMD products, one of the major advantages of Nvidia’s GPU is software support. By helping to develop and maintain an open-source community, Nvidia has boosted the development of multiple applications for data analytics and parallel computing structures for which its GPU can run flawlessly, but need certain work to be done when using AMD or Intel such as translate CUDA application instructions to OpenCL (a framework that used for AMD). For programmers using Nvidia’s GPU, they can easily find support in CUDA libraries if something goes wrong, but that is not the case with AMD. With its dominant market position, Nvidia is leading the development trend while AMD’s strategy is to offer products with better price-performance-power consumption ratio.
The acquisition of Mellanox, which is still pending for approval of regulatory issues, also demonstrates how Nvidia wants to boost its sales and profit by offering its customers more comprehensive solutions. Mellanox is a provider of InfiniBand, a computer networking technology. Their products together form an end-to-end integrated networking solution. The customers for Mellanox are cloud, web 2.0, server, storage, ML, and telco providers. Currently, the InfiniBand consists of a large share of network communications of data centers and supercomputers.
If Nvidia and Mellanox are combined, Nvidia will be able to offer datacenter customers high computational power hardware and software solutions, while Mellanox will be able to help customers with inter-server, internal web data transmission. There are two business synergies that can be realized through this deal:
The synergy of cost-cutting in the sales force. For now, sales and marketing compose about 10%-15% of Mellanox’s revenue with 460 sales and marketing personnel. Although it is unrealistic to save sales force for Nvidia, Mellanox certainly could be cutting cost by consolidating its sales team with Nvidia’s data center sales team.
Synergy in new customers development: Mellanox has a large data center customer base with good quality (HP and Dell compose of more than 25% of the total revenue). This customer base will help Nvidia get easier touch with these companies for its own datacenter services. Thus, making customers development success rate higher, or even lower the churn rate. Also, by offering its customers more comprehensive solutions, Nvidia and Mellanox together will be able to charge a higher price.
Source: company press release
3. New technology and shifting of the trend will help Nvidia continue to secure its position in gaming and grab some growth opportunity.
The gaming sector has experienced a slow year. The attach rate (GPU sold as a consequence of PCs sold) for GPUs has been volatile for the past few years ranging from 130% to 140%, including both discrete and integrated GPUs, and overall PC market decreased 14.4% quarter over quarter and 4.2% year over year. These two data combined indicate that the overall shipment for GPUs as a result of PC sales is declining. The general trend for the gaming industry is that players are shifting from PCs to consoles, and mobile gaming is growing at a faster pace. The reason behind is that the performance gap between PCs with discrete GPUs, consoles, and mobiles is decreasing. The whole gaming sector revenue is projected to grow at 6% CAGR in the next 5 years.
Source: Jon Peddie Research
As consoles offer more affordable solutions to most gamers, and high-performance TVs are becoming cheaper, the shift is inevitable. Unlike the dominant position that Nvidia has in PCs, it faces intense competition in the consoles GPU market. Console gaming will likely experience faster growth in the future as sales of console games grew 15.2% vs only 3.2% for PCs for the year of 2018. In the E3 game show, both Sony (SNE) and Microsoft (MSFT) announced its next-generation consoles PS5 and project Scarlett. While Sony clearly announced that its consoles will have AMD Radeon Navi as GPU, Microsoft did not announce detailed GPU. It only said that it will support Ray-tracing technology, which currently only Nvidia can provide.
Source: Jon Peddie Research
There are two major reasons that Nvidia’s gaming segment could demonstrate strong growth in the next several years. One reason is that Nvidia currently offers ray tracing accelerating on its GPU, which is a rendering technology for better light, and AMD has no counter products to offer to the market. It is a huge topic in this E3 game show as it gives Nvidia the upper hand of the competition. AMD’s market share will be impacted by the lack of such technology in the near term since many gaming companies have launched plans for related game development based on this technology.
Ray tracing effects:
Recently Nvidia launched new products 2080 Super retails for $699, 2070 Super retails for $499 and 2060 Super for $399. The announcement came a month late of AMD’s announcement for offering AMD RX Radeon 5700 XT gaming card which retails for $449. RX Radeon 5700 XT is $50 cheaper than 2070 Super.
The product launches echo AMD and Nvidia’s strategies while AMD tries to counter Nvidia with better pricing, but Nvidia is focused on leading the market trend. The major difference between these two chips is computational power (see below Final Fantasy XV Benchmarks results), and 2070 Super offers ray tracing technology and DLSS (deep learning super sampling, an AI technology for boosting framerate). So, the consumers pay the extra $50 for the additional performance, ray tracing, and DLSS. This is the initial phase for major gaming factories to explore the technology, and as more games start to rely on this technology, it will boost Nvidia’s market position.
The second reason is Nvidia’s Tegra product, which is an SoC used on mobile devices such as HTC one cellphones, tablets, and most importantly, the Nintendo switch. It has experienced booming sales in the past three years. As of March 2019, Nintendo Switch has sold for 34.74 million units with 187.52 million units of software. The consoles are loved by the market due to several reasons such as portability, perfect product design, and most importantly, the games.
Switch uses the Tegra processor, and its strong sales also boosted revenue for Tegra - from $824MM in 2017 to $1,534MM in 2018 and $1,541MM USD in 2019. As Nintendo launches new game projects such as Zelda 2 and Pokemon, and the next generation of Switch (not officially confirmed, but it is expected this year with an upgrade of display and still use Tegra processor), the shipment of Tegra processor will also increase with the sales boost both this and next year. Another thing worth noting is that Tegra processor is also used in the autonomous driving business segment. So, as Nvidia partners up with more carmakers and commercializes more products, the revenue from Tegra processor could continue to increase.
Another recent hot topic is Cloud gaming, the reason being the accessibility of faster internet.
Most cloud services require an internet speed of over 10-15 MBPs to run smoothly with minimum latency, the weight of fixed connection with speed over 10 MBPs has increased to 84.2% in 2017 from 69.4% in 2014.
Nearly all major players have devoted resources to launch their own cloud gaming projects. Some people assume that cloud gaming could be comparable to video streaming and it will shape the gaming industry. However, most companies are still exploring business models. Below is a table showing the services provided by product and their pricing.
Each company in this field has its own unique advantages. Google is strong in building infrastructure, and as a major player in cloud, it is the only company that offers 4k resolution on gaming. Also, by combining with its YouTube platform, Stadia could help gamers to share their playing time more easily or interact with other streamers seamlessly. It is possible that Stadia and YouTube together will realize a synergy.
Nvidia’s GeForce now offers multiple platform access and high GPU specs so that players can run the games at better graphics effects. It requires the gamers to own the games before play, so it is more like a cloud service instead of Netflix’s streaming business model. As Nvidia SHIELD offers a free content library, it is possible that cloud gaming could drive up the sales for SHIELD.
Sony PlayStation Now offers cloud gaming customers access to the contents that were exclusively available for PS2, 3, and 4 consoles owners. This is extremely attractive because PS gaming contents helped Sony to win the upper hand in the battle with Microsoft Xbox. And now these IPs could help Sony monetize on cloud gaming.
Microsoft and EA both launched their projects this year. Although they are a little behind Google, Nvidia, and Sony, their platform and content library will make their cloud gaming attractive to gamers.
Other startups are focusing on different models or areas in order to compete in this battle. For example, indie game (independent games such as Minecraft) players offer the cheapest price on the market, and they provide more than 100 indie games on their server. Shadow allows you to create your own virtual PC on their clouds with upgraded hardware to run the AAA games that have high requirements. PlayGiga chooses to serve business customers, help Telco and Publishers to put more games in their content library and expand their customer base.
While it is uncertain how the cloud gaming market will turn out in the end, currently, there are several issues faced by the companies. The subscription fees range from $10-$30 per month is not cheap. While a Sony PS4 console only cost from $250 -$400 dollars, it means that 10-20 months of subscription to the cloud service can buy you an equivalent console. Although the cost for PC desktops with the same level of hardware is more expensive, starting close to $1,000, many people will still choose to save up the money and assemble a desktop because subscription for higher internet speed also cost money. Another reason that many gamers choose local desktop is latency. For gamers playing FPS or action games, a 100-millisecond latency is still noticeable. However, this latency cannot be easily eliminated because of the multiple steps in connection from the cloud server to the local machine.
Cloud gaming will be a big opportunity for Nvidia because, no matter how successful GeForce Now will be, most cloud gaming provider will choose to use Nvidia’s GPU in their data centers, especially startup companies who lack the ability to design and manufacture their own chips.
It is likely that, in the future, as cloud services for gamers mature, the individual buying of GPUs will decrease, and cloud gaming providers will purchase more GPU to construct their data center. As long as the gaming demand is still increasing, it will drive the growth of Nvidia GPUs.
4. The crypto problem seemed to be in the past for Nvidia. The company’s focus on solving inventory issues is showing positive results, and management has commented multiple times that business is returning back on track.
Nvidia’s inventory management and turnover ratio were superb until the second half of 2018 when inventory spiked and inventory turnover more than halved those from the year prior. The decline in cryptocurrency evaporated the demand for graphics chips used by miners, and the resulting excess inventory caused a slowing of orders of the chips that are more typically used by PC gamers. The inventory issue translated into sales and earnings drop. Based on management’s comments, it should take Nvidia a few quarters to address the inventory problem (first half of 2019), after which is should be back to business (second half of 2019).
November 2018 - Jensen Huang: "our near-term results reflect excess channel inventory post the cryptocurrency boom, which will be corrected."
January 2019 – Jensen Huang: “Our business, as you guys know, was affected because we have excess inventory. That’s largely been a near-term business factor for us. We announced that, because of the post-crypto decline, we had excess inventory approximately equal to something that would have taken us another one to two quarters to sell. That affected our business and our earnings, the guidance for this last quarter. What’s going to happen is, we will sell through all of our inventory post-crypto in the channel between one to two quarters from last quarter. The excess inventory in the channel has been reported so excessively that I hope you guys don’t need to report it again.”
January 2019 – Jensen Huang: “It’s completely a crypto hangover issue. Remember, we basically shipped no new GPU in the market, to the channel, for one quarter. But the amount of excess inventory and market demand, channel velocity — you just have inventory divided by velocity, and that’s time. We said that it would take one to two quarters for all the channel inventory to sell out. 1080Ti has sold out. 1080 has sold out. 1070 has sold out. 1070Ti has sold out. In several more weeks, 1060s will sell out. Then we can go back to business.”
With a quick analysis on NVIDIA’s inventory, we saw that total inventory peaked during Q4 of last year ($1.575B), and have since reduced; inventory turnover, which bottomed in Q4 (1.47), also started to stabilize. This metric would be an important one to pay attention to, if inventory turnover could continue to trend up towards the second half of the year, it would be a clear demonstration that the inventory issue is under control, which should then translate to improvement in profitability.
Given NVIDIA’s earnings profile, we used two valuation approaches, 10-year DCF analysis, and earnings multiples analysis, which we believe are the most appropriate. Below, we discuss the results of both approaches and our findings.
- 2020 is going to be a down year for NVIDIA, but growth will re-emerge in 2021+, driven by growth in gaming and data center
- Long term EBITDA margin of 45%
- 15% tax rate
- Depreciation and amortization at 2.5% of revenue
- Capex to be at 4-5% of revenue in the short term, dropping to 3% in the long term
- WACC at 11% +/- 2%
- Terminal growth rate at 4% +/- 1%
- 609 million shares outstanding
Based on our assumptions, the DCF analysis gives us a price range of $122 - 293 per share, compared to the current price of $160. The model relies heavily on the assumptions that NVIDIA’s annual revenue surpassing $20B by 2023 and $30B by 2026, which is roughly a CAGR of 20% for the next 6 years. The wide range in price targets was the result of the sensitivity analysis in WACC and terminal growth rate. In an unfavorable scenario of a 12% discount rate and terminal growth rate of only 3% would put a $122 floor on NVIDIA, which is about 25% below the current price level, the favorable terms of 5% terminal growth rate and an 8% discount rate brings the estimated price up to $293, which implies an 80% increase from today’s price.
Although the price range is wide, this analysis gives a good understanding of the price range that NVIDIA could be trading at, based on the outcomes in revenue growth, EBITDA margin, and cost of capital. Given how volatile the stock has traded recently, a price below $130 would certainly be a nice entry point, provided that growth is still intact.
Forward Multiples Analysis
For the multiples analysis, we broke down each of NVIDIA’s business segments - gaming, professional visualization, data center, auto, OEM - and their projected growth for the next 3 years, and we state the reasonings behind our assumptions.
Gaming: Revenue for Q1 was $1.05B with a YoY decline of -38.77%. As the crypto bubble burst, the GPU demand changed drastically, and 2020 will be a year of recovery. The long-term GPU demand is still firm, so after its recovery, Nvidia’s gaming revenue should increase at a faster pace.
· The whole gaming sector revenue is projected to grow at 6% CAGR in the next 5 years.
· Tegra processor sales may get a boost from Nintendo switch’s sales increase from the launch of new games.
· Cloud gaming could help shift gamers from consoles to subscription gaming.
Professional Visualization: First-quarter revenue for PV was $266MM, a YoY increase of 5.98%. The estimated increase for PV revenue is 6% for FY2020.
· VR content market is expected to grow at CAGR around 69.8%, which can boost the PV revenue for the long-term.
· Ray-tracing technology may increase the demand for QUADRO platform.
Data Center: First quarter data center revenue was $634MM USD, a YOY decline of -9.56%. Datacenter customers are still digesting the huge CapEx from 2019, so DC revenuer for FY2020 is going to decrease at around 8% compared to the previous year.
· The AI hardware industry is expected to increase at a pace faster than 60%, while GPU revenue for AI is expected to increase at 31.1%.
· Market dominant position and better software support will help Nvidia capture the wave of AI in the next several years.
· Developing a more comprehensive solution for its customers will increase the market share for Nvidia.
Auto: Revenue from Auto will increase by 10% for 2020 and 16%-18% for the year later as more applications and products being developed become commercialized.
· The projected market size is expected to increase at a CAGR of 14% from 2016 to 2025 while the unit shipment for AI-based vehicles will increase at about 20%.
· Partnering up with major car makers will help Nvidia to speed up its product development and launching, establishing a first-mover advantage.
· Providing software solution based on GPU will help Nvidia harvest higher profit than just selling GPUs to carmakers.
OEM & Others: Revenue from OEM has been on a steady decline for the past several quarters.
Total Revenue: For the first half of 2020, Nvidia is expected to achieve a revenue of around $4.7B ($2.2B for Q1, and $2.5B management projection revenue for Q2). We project that Nvidia is likely to achieve a revenue level of around $10.5 – 11B for the year. Based on that estimate, we project the 2021 revenue to be $12.67B and $14.9B in 2022, representing a YoY growth of 16.24% and 17.78%, respectively.
EPS: For the year of 2020, Nvidia’s net profit margin is expected to stay at 30%, lower than the previous two years of 31% and 35.34% because of demand slowdown and inventory pressure. But it is expected to resume to the 35% level in 2021.
The EPS range is expected to be $5.15 – 5.4 for this year and $6.85 – 7.46 for 2021.
With the EPS forecast and a range of PE ratios between 20x - 40x. (peer group is trading at a PE ranging from 15x – 35x). If projected at 30x PE using this year’s EPS, the projected price range is $154 – $161 per share, looking out 18 months, based on next year’s estimate and using a narrower 25-30x forward multiple, we believe the stock should be valued at $171-$223.
The PE multiple is a moving ratio because it incorporates investors expectation for the company’s future growth. As for now, Nvidia’s PE multiple is higher than its peers. However, as the growth prospect for Nvidia becomes clearer, it is likely to follow a more linear and slower long-term growth rate as AI, ML, and AV industries mature. At that time, the reasonable PE ratio may be lowered and trading at 20x – 25x range.
Using 2022’s projected EPS, and apply a 20x-25x multiple range, the projected price range is $168 – 210.
NVIDIA looks to acquire MLNX at price of $125/share, a 14.28% premium at the time when the deal was announced. The acquisition will be an all-cash deal at $6.9B. Considering that data center is about 28.55% of Nvidia’s total revenue, we think a 0% - 5% is the range for EPS synergy from the deal is realistic. The deal is subject to many restrictions and government approval, if for any reason that the deal didn’t go through, NVIDIA is obligated to pay a termination fee of 350 million USD.
Although Nvidia's businesses remain fundamentally sound and, in the long run, has great growth prospects, there are still many potential risks that cannot be ignored.
In the next 6-12 months, Nvidia needs to recover from the industry slowdown and adjust to the demand shift. The biggest risk that’s currently in front of Nvidia is the demand shift and market change. Even though most people believe that the strong demand for GPU chips will persist for a period, nobody knows what the demand for GPUs is going to be likely in 5 years. New development of technology and algorithm process can cause the structure of the industry to change. And once growth slows down, so will valuation.
Investors should also pay attention to management communication and disclosure. For a company with so many product families and services, failing to disclose business segment results will cause mistrust should the stock becomes volatile. As it has happened in the past year when the stock price goes down, many analysts challenged the transparency of Nvidia’s management. Below are the specific risks that investors should pay attention to.
· A significant shift in GPU demand could seriously impair Nvidia’s future sales number. Even though GPU currently plays a vital role in AI, ML, DL, and related fields; however, competition remains fierce with other more specific task chips such as TPU. If the process of algorithm changes, it may make some chips more suitable for certain tasks and decrease the demand for GPU in general. A change in GPU demand could also come from the below expectation development of certain industry such as AI, and SAAS.
· The roadmap of development for the next generation of consoles and gaming could impact Nvidia’s future sales number. If Nvidia fails to supply for next-generation consoles, it could miss a huge growth opportunity and impact its market position. It could also be a serious impact for Nvidia if Nintendo decided to launch the next generation of switch without using Tegra processor. And if the gaming market continues to shift toward mobile gaming, Nvidia’s sales will also be impacted as it does not have a significant presence in mobile gaming.
· Even though Nvidia invest 20% of revenue into R&D every year, the development of certain technologies may face challenges or slower than anticipated.
· If the trade tension between China and the US gets worse, it could negatively impact Nvidia’s revenue as nearly a quarter of sales comes from China. Even though it is unlikely the government will block the sales of Gaming and general GPU to the Chinese customers, it is still likely that certain products for data center customers and certain software be blocked. It could cause Nvidia to lose one of the large markets, which is also growing at a fast pace. More seriously it could jeopardize Nvidia’s dominant position in the chip industry as China is investing heavily in semiconductor and there are many startup chip makers in China craving for Nvidia’s market share (there is a long road to walk for starters to catch up on technology).
· The highly anticipated Autonomous Vehicle business still faces some uncertainty. The industry is facing heavy regulation as several accidents have attracted public attention, and this could seriously slow down the development of AV. Also, even though Nvidia has partnered up with major carmakers, it is still unclear how these partnerships will turn out. If carmakers decide to heavily invest in their own AI platform and decrease the reliance on Nvidia, they are likely to drop out from the partnership, as Tesla did. Also, it is unclear how Nvidia’s AV revenue will be structured as it offers solutions from AI platform, hardware, to simulations, so it is hard to predict its growth rate and profitability.
· A failed acquisition of MLNX would mean for Nvidia to pay a $350MM termination fee, which could negatively impact the financial result. The deal is still under regulatory approval, and the uncertainty is not neglectable.
· Nvidia plans its production and estimate demand form its distribution channels. During the crypto bubble, the management wrongly estimated the demand of the market and, as a result, inventory has piled up at a worrying level. Now Nvidia is working hard to resolve the issue; however in the future, if the management misread the market again, it could negatively impact the operational efficiency of the company and lower the net profit expectation.
· The non-transparency of Nvidia’s management to communicate operating data, such as average unit price or unit shipment for each product family, with investors could lead to inaccuracy in forecasts or, even worse, breed trust issue between the street and Nvidia, impacting Nvidia’s valuation.
· Currently, Nvidia is trading at a relatively expensive ratio because of its dominant position in GPU, fast-growing forecast and profitability. However, as the macro trend for the economy turns, the market may not want to pay the premium in the valuation anymore, and this could seriously impact Nvidia’s share price.
In our eyes, Nvidia took a big punch in the face during the crypto bust, and its stock price suffered meaningfully. Post its share price correction, fundamentally speaking, Nvidia is just as solid as it was before. Although potential risks still exist, the growing prospect for Nvidia is bright, driven by the projected demand increase in AI, gaming, and autonomous vehicles.
Based on the earnings results from the first two quarters of this year, Nvidia seemed to be on track for its turnaround and, although management did not provide any guidance for the second half of the year, if Nvidia delivers a strong Q3 and/or Q4 performance, the stock will certainly be testing the $200+ range. However, if the turnaround lasts longer than anticipated, or if growth slows, another selloff would be warranted. The pending acquisition of MLNX and the outcome of the trade war will also be impactful and should be monitored closely.
With the stock currently trading at $160 – $170 level, while the DCF model gives us a price range of $122 - $293 and the multiples analysis gives a base case between $168 - $215 based on 2021 and 2022 multiples. So currently the buying position is not justified with enough margin of safety. If the fundamentals do not change for Nvidia and price falls to ~$130 (15-20% drop), it will certainly make the stock much more attractive and worth an entry.
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in NVDA over the next 72 hours. 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.