省流:8.48%_老虎社区_美港股上老虎

省流:8.48%

Nvidia: Valuation Seems Reasonable After Recent Rally

Nvidia has passed Apple and Microsoft to become the largest company in the world, and we evaluate if the company deserves such high of a valuation.We use a Reverse DCF model to obtain the market-impli

Nvidia: Valuation Seems Reasonable After Recent Rally

免责声明：上述内容仅代表发帖人个人观点，不构成本平台的任何投资建议。

推荐
最新

暂无评论

热议股票

关于我们·老虎社区守则·老虎社区账号管理规范·老虎社区服务协议·老虎社区隐私政策

公司名称：北京至简风宜信息技术有限公司

违法和不良信息投诉：010-5681-3562（工作时间9:30-18:30）

邮箱：service@laohu8.com

营业执照：91110105MA01A4U55R

ICP备：京ICP备18016422号

网安备：京公网安备11010502036490

关注我们

{"i18n":{"language":"zh_CN"},"data":{"magic":2,"id":319000919421120,"tweetId":"319000919421120","gmtCreate":1718894536330,"gmtModify":1718894537938,"author":{"id":3495002759494258,"authorId":3495002759494258,"authorIdStr":"3495002759494258","name":"京东商城","avatar":"https://static.tigerbbs.com/2b575974bc09e65f9560c90d814edf6f","vip":1,"userType":1,"introduction":"","boolIsFan":false,"boolIsHead":false,"crmLevel":3,"crmLevelSwitch":0,"currentWearingBadge":{"badgeId":"e50ce593bb40487ebfb542ca54f6a561-4","templateUuid":"e50ce593bb40487ebfb542ca54f6a561","name":"明星虎友","description":"加入老虎社区2000天","bigImgUrl":"https://static.tigerbbs.com/dddf24b906c7011de2617d4fb3f76987","smallImgUrl":"https://static.tigerbbs.com/53d58ad32c97254c6f74db8b97e6ec49","grayImgUrl":"https://static.tigerbbs.com/6304700d92ad91c7a33e2e92ec32ecc1","redirectLinkEnabled":0,"hasAllocated":1,"isWearing":1,"stampPosition":0,"hasStamp":0,"allocationCount":1,"allocatedDate":"2024.01.09","individualDisplayEnabled":0},"individualDisplayBadges":[],"fanSize":523,"starInvestorFlag":false},"themes":[],"images":[],"coverImages":[],"html":"<html><head></head><body>省流:8.48%</body></html>","htmlText":"<html><head></head><body>省流:8.48%</body></html>","text":"省流:8.48%","highlighted":1,"essential":1,"paper":1,"likeSize":0,"commentSize":0,"repostSize":0,"favoriteSize":0,"link":"https://laohu8.com/post/319000919421120","repostId":2444429849,"repostType":2,"repost":{"id":"2444429849","pubTimestamp":1718853852,"share":"https://www.laohu8.com/m/news/2444429849?lang=&edition=full","pubTime":"2024-06-20 11:24","market":"nz","language":"en","title":"Nvidia: Valuation Seems Reasonable After Recent Rally","url":"https://stock-news.laohu8.com/highlight/detail?id=2444429849","media":"seekingalpha","summary":"Nvidia has passed Apple and Microsoft to become the largest company in the world, and we evaluate if the company deserves such high of a valuation.We use a Reverse DCF model to obtain the market-impli","content":"<html><head></head><body><ul style=\"\"><li>Nvidia has passed Apple and Microsoft to become the largest company in the world, and we evaluate if the company deserves such high of a valuation.</li><li>We use a Reverse DCF model to obtain the market-implied assumptions of the company&#39;s growth rate, terminal value multiples, and profit margins and compare them against our own assumptions.</li><li>We believe the company&#39;s Blackwell architecture launch and expanded cloud partnerships with major providers contribute to its competitiveness and growth outlook.</li></ul><img src=\"https://community-static.tradeup.com/news/6144f6f01bec9d39bca22c82ac595a7e\" tg-width=\"750\" tg-height=\"500\"/>BING-JHEN HONGIn our previous analysis of NVIDIA Corporation (NASDAQ:NVDA) (NEOE:NVDA:CA), we highlighted the strong performance of Nvidia’s growth of 126% in CY2023, and we remained positive on Nvidia’s Data Center segment, supported by cloud partnerships and sovereign AI initiatives. For 2024, we projected a 58% overall revenue growth, led by the Data Center, with contributions from Gaming, Professional Visualization, and a recovery in PC GPU revenue, as well as 20.6% growth in the Automotive segment.Since our last analysis in March, we started reviewing Nvidia again as its share price had risen by 28% in May, with a further upside of 18.3% left based on our previous price target of $1,294.14 ($129.41 post-stock split). Therefore, we initially aimed to determine whether our expectations for Nvidia’s upside were justified. Since then, Nvidia has overtaken both Apple and Microsoft as the largest company globally by market cap, rising by a further 30% and meeting our price target. Firstly, we conducted a reverse DCF analysis determining the market implied revenue growth forecast, terminal value multiple, and profitability margins. Based on that, we compared the reverse DCF model with our revenue, terminal value and margin assumptions and determined whether our assumptions were appropriate.<h2 id=\"id_1356393531\">Sustainable Revenue Growth</h2>Firstly, we examined whether our revenue growth forecast is appropriate. Based on the reverse DCF model, holding all previous assumptions constant, we derived the market-implied revenue growth rate for Nvidia based on the current market price.<img src=\"https://community-static.tradeup.com/news/0ce46a7a6e946a7995df8c13ad343a0e\" tg-width=\"640\" tg-height=\"360\"/>Company Data, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Revenue Comparison ($ mln)</td><td style=\"text-align:left;\">2024F</td><td style=\"text-align:left;\">2025F</td><td style=\"text-align:left;\">2026F</td><td style=\"text-align:left;\">2027F</td><td style=\"text-align:left;\">2028F</td><td style=\"text-align:left;\">2029F</td><td style=\"text-align:left;\">2030F</td><td style=\"text-align:left;\">2031F</td><td style=\"text-align:left;\">2032F</td><td style=\"text-align:left;\">2033F</td><td style=\"text-align:left;\">Avr. (5-yr)</td><td style=\"text-align:left;\">Avr. (10-yr)</td></tr><tr><td style=\"text-align:left;\">Revenue (Reverse DCF)</td><td style=\"text-align:left;\">113,185</td><td style=\"text-align:left;\">163,293</td><td style=\"text-align:left;\">220,004</td><td style=\"text-align:left;\">273,487</td><td style=\"text-align:left;\">320,027</td><td style=\"text-align:left;\">358,148</td><td style=\"text-align:left;\">388,012</td><td style=\"text-align:left;\">410,660</td><td style=\"text-align:left;\">427,439</td><td style=\"text-align:left;\">439,664</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">Revenue Growth</td><td style=\"text-align:left;\">85.8%</td><td style=\"text-align:left;\">44.3%</td><td style=\"text-align:left;\">34.7%</td><td style=\"text-align:left;\">24.3%</td><td style=\"text-align:left;\">17.0%</td><td style=\"text-align:left;\">11.9%</td><td style=\"text-align:left;\">8.3%</td><td style=\"text-align:left;\">5.8%</td><td style=\"text-align:left;\">4.1%</td><td style=\"text-align:left;\">2.9%</td><td style=\"text-align:left;\">41.2%</td><td style=\"text-align:left;\">23.9%</td></tr><tr><td style=\"text-align:left;\">Revenue (Our Previous Forecast)</td><td style=\"text-align:left;\">95,720</td><td style=\"text-align:left;\">148,848</td><td style=\"text-align:left;\">225,273</td><td style=\"text-align:left;\">306,238</td><td style=\"text-align:left;\">383,284</td><td style=\"text-align:left;\">450,785</td><td style=\"text-align:left;\">506,357</td><td style=\"text-align:left;\">550,053</td><td style=\"text-align:left;\">583,280</td><td style=\"text-align:left;\">607,943</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">Revenue Growth</td><td style=\"text-align:left;\">57.1%</td><td style=\"text-align:left;\">55.5%</td><td style=\"text-align:left;\">51.3%</td><td style=\"text-align:left;\">35.9%</td><td style=\"text-align:left;\">25.2%</td><td style=\"text-align:left;\">17.6%</td><td style=\"text-align:left;\">12.3%</td><td style=\"text-align:left;\">8.6%</td><td style=\"text-align:left;\">6.0%</td><td style=\"text-align:left;\">4.2%</td><td style=\"text-align:left;\">45.0%</td><td style=\"text-align:left;\">27.4%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the reverse DCF model, we derived the revenue growth rate at a forward average of 41.2% in the next 5 years, compared to our average forward growth forecast of 44.8%. Furthermore, we derived a 10-year forward growth of 23.9% from the reverse DCF model, compared to 27.4% from our projections. Therefore, this indicates our revenue growth assumption is higher compared to the calculated growth rate from the reverse DCF model. In our previous analysis, we highlighted that we maintained a more stable long-term growth projection “derived independently based on individual segment projections and modeling” for the company “driven primarily due to its Data Center segment at a forward average of 60.9% underpinned by its AI leadership”.We examined the latest developments of the company which could support its growth outlook below, including its product launch of its Blackwell architecture, expanded partnerships, and competition.<h3 id=\"id_3707924856\">Competitiveness (Blackwell Launch)</h3>Nvidia announced the next-gen Blackwell architecture in March 2024, claiming that it enhances computing and generative AI by providing improvements in scale, performance, and efficiency, with customer data centers planned to be operational by Q4 2024. The Blackwell-architecture GPUs are “manufactured using a custom-built 4NP TSMC process” with 208 bln transistors. Previously, we had highlighted the company’s Blackwell architecture based on the company’s provided roadmap in 2024; thus, we believe this indicates its execution on its continued product launches for Data Center.<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Category</td><td style=\"text-align:left;\">H100</td><td style=\"text-align:left;\">H200</td><td style=\"text-align:left;\">B100</td><td style=\"text-align:left;\">B200</td><td style=\"text-align:left;\">GB200 Superchip</td></tr><tr><td style=\"text-align:left;\">Price</td><td style=\"text-align:left;\">$24,000</td><td style=\"text-align:left;\">$24,000</td><td style=\"text-align:left;\">$30,000</td><td style=\"text-align:left;\">$30,000-$40,000</td><td style=\"text-align:left;\">$60,000-$70,000</td></tr><tr><td style=\"text-align:left;\">Increase %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">25.0%</td><td style=\"text-align:left;\">16.7%</td><td style=\"text-align:left;\">85.7%</td></tr><tr><td style=\"text-align:left;\">Memory Capacity (GB)</td><td style=\"text-align:left;\">80.00</td><td style=\"text-align:left;\">141.00</td><td style=\"text-align:left;\">192.00</td><td style=\"text-align:left;\">192.00</td><td style=\"text-align:left;\">384.00</td></tr><tr><td style=\"text-align:left;\">Memory Bandwidth (GB/s)</td><td style=\"text-align:left;\">3,352.00</td><td style=\"text-align:left;\">4,800.00</td><td style=\"text-align:left;\">8,000.00</td><td style=\"text-align:left;\">8,000.00</td><td style=\"text-align:left;\">16,000.00</td></tr><tr><td style=\"text-align:left;\">Improvement %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">43.2%</td><td style=\"text-align:left;\">66.7%</td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">100.0%</td></tr><tr><td style=\"text-align:left;\">TF32 (Teraflops)</td><td style=\"text-align:left;\">495.00</td><td style=\"text-align:left;\">495.00</td><td style=\"text-align:left;\">900.00</td><td style=\"text-align:left;\">1,100.00</td><td style=\"text-align:left;\">2,500.00</td></tr><tr><td style=\"text-align:left;\">Improvement %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">81.8%</td><td style=\"text-align:left;\">22.2%</td><td style=\"text-align:left;\">127.3%</td></tr><tr><td style=\"text-align:left;\">FP16 (Teraflops)</td><td style=\"text-align:left;\">989.00</td><td style=\"text-align:left;\">989.00</td><td style=\"text-align:left;\">1,750.00</td><td style=\"text-align:left;\">2,250.00</td><td style=\"text-align:left;\">5,000.00</td></tr><tr><td style=\"text-align:left;\">Improvement %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">76.9%</td><td style=\"text-align:left;\">28.6%</td><td style=\"text-align:left;\">122.2%</td></tr><tr><td style=\"text-align:left;\">FP8 (Teraflops)</td><td style=\"text-align:left;\">1,979.00</td><td style=\"text-align:left;\">1,979.00</td><td style=\"text-align:left;\">3,500.00</td><td style=\"text-align:left;\">4,500.00</td><td style=\"text-align:left;\">10,000.00</td></tr><tr><td style=\"text-align:left;\">Improvement %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">76.9%</td><td style=\"text-align:left;\">28.6%</td><td style=\"text-align:left;\">122.2%</td></tr><tr><td style=\"text-align:left;\">FP4 (Teraflops)</td><td style=\"text-align:left;\">1,979.00</td><td style=\"text-align:left;\">1,979.00</td><td style=\"text-align:left;\">7,000.00</td><td style=\"text-align:left;\">9,000.00</td><td style=\"text-align:left;\">20,000.00</td></tr><tr><td style=\"text-align:left;\">Improvement %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">0.0%</td><td style=\"text-align:left;\">253.7%</td><td style=\"text-align:left;\">28.6%</td><td style=\"text-align:left;\">122.2%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsAccording to the above table comparing the GPUs across Nvidia’s portfolio, the Blackwell chips (B100 and B200) show increasing capabilities including memory capacity, memory bandwidth, and improvements in floating-point performance metrics which play important roles in deep learning model training, as well as higher pricing compared to the Hopper chips (H100 and H200).Additionally, the GB200 Superchip which consists of two Blackwell GPUs and one Grace CPU, is reported to have a price range of $60,000-$70,000, along with doubled memory capacity and bandwidth in comparison to previous gen B200 and more than 100% of improvements in the above listed floating-point performance metrics for deep learning model training.Furthermore, we updated our comparison of the top data center GPU chips of Nvidia with its top competitors. Besides Nvidia, AMD (AMD) is expected to launch its new MI350 chip this year and Intel’s (INTC) new Gaudi 3 is expected to be released in Q2. Furthermore, Nvidia highlighted competition in China from Huawei which is expected to release its new and improved Ascend 910B chip this year as well following its Ascend 910 introduction in 2019.<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Data Center GPU Comparison Metrics</td><td style=\"text-align:left;\">AMD (MI300x)</td><td style=\"text-align:left;\">AMD (MI350)</td><td style=\"text-align:left;\">Nvidia (B100)</td><td style=\"text-align:left;\">Nvidia (GB200 Superchip)</td><td style=\"text-align:left;\">Intel (Gaudi 2)</td><td style=\"text-align:left;\">Intel (Gaudi 3)</td><td style=\"text-align:left;\">Huawei (Ascend 910B)</td></tr><tr><td style=\"text-align:left;\">Processor</td><td style=\"text-align:left;\">5nm (TSMC)</td><td style=\"text-align:left;\">4nm (TSMC)</td><td style=\"text-align:left;\">4nm (TSMC)</td><td style=\"text-align:left;\">4nm (TSMC)</td><td style=\"text-align:left;\">7nm (TSMC)</td><td style=\"text-align:left;\">5nm (TSMC)</td><td style=\"text-align:left;\">7nm</td></tr><tr><td style=\"text-align:left;\">Transistors (&#39;bln&#39;)</td><td style=\"text-align:left;\">153</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">208</td><td style=\"text-align:left;\">208</td><td style=\"text-align:left;\">&gt;100</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">-</td></tr><tr><td style=\"text-align:left;\">FP16 Peak (Teraflops)</td><td style=\"text-align:left;\">1,532</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">3,500</td><td style=\"text-align:left;\">10,000</td><td style=\"text-align:left;\">839</td><td style=\"text-align:left;\">1,835</td><td style=\"text-align:left;\">320</td></tr><tr><td style=\"text-align:left;\">INT 8 (Teraflops)</td><td style=\"text-align:left;\">3,064</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">7,000</td><td style=\"text-align:left;\">40,000</td><td style=\"text-align:left;\">1,628</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">640</td></tr><tr><td style=\"text-align:left;\">Memory Clock</td><td style=\"text-align:left;\">1.9GHz</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">8Gbps HBM3E</td><td style=\"text-align:left;\">8Gbps HBM3E</td><td style=\"text-align:left;\">1.56GHz</td><td style=\"text-align:left;\">3.6GHz</td><td style=\"text-align:left;\">-</td></tr><tr><td style=\"text-align:left;\">Memory Capacity</td><td style=\"text-align:left;\">192GB</td><td style=\"text-align:left;\">288GB</td><td style=\"text-align:left;\">192GB</td><td style=\"text-align:left;\">384GB</td><td style=\"text-align:left;\">96GB</td><td style=\"text-align:left;\">128GB</td><td style=\"text-align:left;\">64GB</td></tr><tr><td style=\"text-align:left;\">Memory Bandwidth</td><td style=\"text-align:left;\">5.2TB/sec</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">8TB/sec</td><td style=\"text-align:left;\">16TB/sec</td><td style=\"text-align:left;\">2.45TB/sec</td><td style=\"text-align:left;\">3.7TB/sec</td><td style=\"text-align:left;\">-</td></tr><tr><td style=\"text-align:left;\">Interconnect Bandwidth</td><td style=\"text-align:left;\">896 GB/sec</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">1800GB/s</td><td style=\"text-align:left;\">1800GB/s</td><td style=\"text-align:left;\">300GB/s</td><td style=\"text-align:left;\">600GB/s</td><td style=\"text-align:left;\">400G/s</td></tr><tr><td style=\"text-align:left;\">Max Power Consumption</td><td style=\"text-align:left;\">700W</td><td style=\"text-align:left;\">-</td><td style=\"text-align:left;\">700W</td><td style=\"text-align:left;\">1000W</td><td style=\"text-align:left;\">600W</td><td style=\"text-align:left;\">900W</td><td style=\"text-align:left;\">310W</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the table, Nvidia’s Blackwell chips show superior performance capabilities, especially GB200 Superchip which stands out with its 4nm processor, highest transistor counts at 208 bln, and exceptional FP16 and INT8 peak performance metrics, 10,000 teraflops and 40,000 teraflops, respectively. It also offers the largest memory capacity of 384GB and the highest memory bandwidth at 16TB/sec, although it comes with the highest power consumption of 1000W.In contrast, AMD has yet to provide details on its MI350 product with limited information about it besides its 4nm process and 288Gb memory capacity, which is lower than the Nvidia GB200. The previous AMD MI300x was built on a 5nm process, providing 1,532 teraflops of FP16 and 3,064 teraflops of INT8 performance with a memory capacity of 192GB and bandwidth of 5.2TB/sec, consuming 700W.Furthermore, Intel Gaudi 3 also uses 5nm technology, with 1,835 teraflops FP16 and 128GB memory but has a higher power consumption of 900W. Notably, the Huawei Ascend 910B, although using older 7nm technology, has the lowest power consumption of 310W.Overall, our forward average revenue growth rate projections (45%) are higher compared to the market implied growth rate through the reverse DCF model (41.2%), a difference of 3.8% which we expect due to the company’s enhanced competitiveness. For example, we believe the improvements in Nvidia’s data center GPU capabilities with Blackwell highlight the company’s execution of its product roadmap and commitment to future product launches, which could cement its competitiveness.<h3 id=\"id_1140507\">Partnerships</h3><table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Company</td><td style=\"text-align:left;\">Purpose of Adoption of Blackwell</td><td style=\"text-align:left;\">Cloud Market Share</td><td style=\"text-align:left;\">Avr. Revenue Growth (3-yr)</td><td style=\"text-align:left;\">Cloud Revenue ($ bln)</td></tr><tr><td style=\"text-align:left;\">AWS</td><td style=\"text-align:left;\">To securely build and run large AI models on AWS</td><td style=\"text-align:left;\">34%</td><td style=\"text-align:left;\">26.4%</td><td style=\"text-align:left;\">90.76</td></tr><tr><td style=\"text-align:left;\">Google Cloud</td><td style=\"text-align:left;\">To enable Google Cloud with advanced AI and data analytics capabilities</td><td style=\"text-align:left;\">12%</td><td style=\"text-align:left;\">37.7%</td><td style=\"text-align:left;\">33.09</td></tr><tr><td style=\"text-align:left;\">Microsoft</td><td style=\"text-align:left;\">To boost Microsoft Azure&#39;s capabilities regarding large-scale AI workloads handling</td><td style=\"text-align:left;\">26%</td><td style=\"text-align:left;\">38.8%</td><td style=\"text-align:left;\">70.20</td></tr><tr><td style=\"text-align:left;\">Oracle</td><td style=\"text-align:left;\">To enhance Oracle&#39;s AI model capabilities for faster and higher efficiency of data processing and AI training</td><td style=\"text-align:left;\">2%</td><td style=\"text-align:left;\">30.2%</td><td style=\"text-align:left;\">5.70</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsAlong with the launch of NVIDIA’s Grace Blackwell GPU platform, its cloud partners including AWS (AMZN), Google (GOOG), Microsoft (MSFT), and Oracle (ORCL) announced plans for adoption. Firstly, AWS will “offer the NVIDIA GB200 Grace Blackwell Superchip and B100 Tensor Core GPUs” and enhance the performance of its AI models with Nvidia’s Grace Blackwell GPU and DGX Cloud. Similarly, Google Cloud claims future adoption of the Grace Blackwell platform to its “AI Hypercomputer architecture in two configurations”, while also focusing on the development of trillion-parameter language models. Microsoft aims to improve its AI infrastructure with the integration of Grace Blackwell GB200 with Microsoft Azure. Oracle also plans on adopting Nvidia’s Grace Blackwell platform across its OCI Supercluster and OCI Compute, aiming to improve performance and efficiency for AI workloads such as LLM inference.Based on the table above, we believe these expanded partnerships with AWS, Google, Microsoft, and Oracle are significant for Nvidia as these major cloud providers purchase and use Nvidia’s chips. They account for 74% of the global cloud market share combined and have robust growth averages of between 26.4% to 38.8%. Furthermore, according to UBS (UBS), Microsoft is believed to be its largest customer with 19% of revenues last year.Overall, we believe the expanded partnerships of Nvidia with the major cloud providers market share combined bodes well for the sustainability of its growth outlook, which is reflected in our projections as we forecasted its forward 3-year growth rate of around 54% on average compared to the reverse DCF model growth which tapers down significantly from 86% to 35% in the next 3 years, as we expect the company’s growth to continue to be sustained by the adoption of upcoming products such as Blackwell.<h3 id=\"id_2518720661\">Current Development of CSPs With Their Own AI Chips</h3><img src=\"https://community-static.tradeup.com/news/955260cbea0325acfd8958c6dcb2006b\" tg-width=\"640\" tg-height=\"335\"/>TrendForceHowever, according to TrendForce, several cloud service providers such as Microsoft and Google have also been developing their respective AI chip technologies recently. For instance, Microsoft is currently in the process of developing their “in-house AI chip Azure Maia 100” and Cobalt 100, which we covered in our previous analysis. Along with Google’s recent development of its tensor processing unit (TPU) of v5e, which is particularly designed for training large AI models. Despite that, Google stated it will provide the optionality for customers to use Google TPUs and Nvidia’s Blackwell GPUs. Additionally, other major companies such as Tesla, Meta, and OpenAI are also pursuing their own respective development regarding AI accelerator chips.Therefore, as competition heats up against Nvidia from key competitors, we believe our projections fairly incorporated this as we tapered down its growth rate beyond 2026. Additionally, while the development of custom AI chips by its cloud customers could affect the company’s growth, we believe Nvidia has an advantage as these companies are not specialized in chip design. While these companies have higher total company absolute R&amp;D expenses than Microsoft, Amazon, and Google, they predominantly focus across the Software &amp; Services (Microsoft), Consumer Discretionary Distribution &amp; Retail (Amazon), and Media &amp; Entertainment industries (Google). Therefore, in comparison, Nvidia is more specialized as it focuses on Semiconductors, enhancing its competitiveness with improved capabilities over competition, including custom chips from cloud providers. For example, in our analysis of Microsoft, we highlighted Microsoft’s Maia 100 GPU trailing behind Nvidia’s H100 chip specifications as it uses older process technology (5nm) and has memory capacity and bandwidth. Furthermore, the Maia 100 would lag behind compared to Nvidia’s B100 chips with an even wider gap as the B100 has twice the number of transistors and 3x higher memory capacity and bandwidth than the Maia 100. Furthermore, we also highlighted the focus on cost as a rationale for custom chip development by cloud providers, such as Microsoft emphasizing Maia enabling cheaper models for customers and “to diversify and provide choices to customers” rather than replacing Nvidia’s high-performance GPUs as well as Amazon highlighting its low-cost AI inference enabled by Inferentia2.<h3 id=\"id_2945945714\">Forward Growth</h3>Overall, our revenue growth projections (44.8%) are higher compared to the calculated market implied revenue growth of the company (41.2%), but we believe it is appropriate given factors such as the enhanced competitiveness of Nvidia and execution on its product roadmap such as with the upcoming Blackwell launch, which we believe cements its competitive positioning against key competitors and custom chip developments by cloud providers, supporting our expectations for the company to continue outperforming the market growth as reflected by our revenue projections. Additionally, our projection which is more sustainable on a 3-year forward average compared to the market implied growth through the reverse DCF which we believe could be supported by the expanded partnerships of major cloud service providers with Nvidia.<h2 id=\"id_3227000819\">Determining Appropriate Terminal Value</h2>Furthermore, we conducted another reverse DCF analysis, adjusting our model holding all assumptions constant except for our terminal value which is derived based on the EV/EBITDA multiple.<img src=\"https://community-static.tradeup.com/news/9747b715cab9de5845d3b411e3e0c6d4\" tg-width=\"640\" tg-height=\"360\"/>Company Data, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Comparison</td><td style=\"text-align:left;\">Terminal Value</td><td style=\"text-align:left;\">EV/EBITDA</td></tr><tr><td style=\"text-align:left;\">Reverse DCF</td><td style=\"text-align:left;\">8,091,521</td><td style=\"text-align:left;\">19.17x</td></tr><tr><td style=\"text-align:left;\">Our DCF</td><td style=\"text-align:left;\">9,712,487</td><td style=\"text-align:left;\">27.78x</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the table, we calculated the Reverse DCF method a terminal value of $8,091 mln with an EV/EBITDA multiple of 19.17x. In comparison, our previous analysis, however, shows a higher terminal value of $9,712 mln and an EV/EBITDA multiple of 27.78x. Previously, we based our terminal value EV/EBITDA on the 5-year US-only chipmaker average of 27.78x.<h3 id=\"id_3270132057\">Industry Multiple</h3><img src=\"https://community-static.tradeup.com/news/87bc458d9fc8a868a53e4b313a36d68c\" tg-width=\"640\" tg-height=\"360\"/>Seeking Alpha, Investing.com, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">EV/EBITDA and CAGR Comparison</td><td style=\"text-align:left;\">EV/EBITDA (5-yr)</td><td style=\"text-align:left;\">5-year Revenue CAGR</td></tr><tr><td style=\"text-align:left;\">US Only</td><td style=\"text-align:left;\">27.78x</td><td style=\"text-align:left;\">11.70%</td></tr><tr><td style=\"text-align:left;\">European Only</td><td style=\"text-align:left;\">11.65x</td><td style=\"text-align:left;\">13.44%</td></tr><tr><td style=\"text-align:left;\">Asian Only</td><td style=\"text-align:left;\">8.26x</td><td style=\"text-align:left;\">6.99%</td></tr><tr><td style=\"text-align:left;\">Overall</td><td style=\"text-align:left;\">21.58x</td><td style=\"text-align:left;\">11.18%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsWe compiled the US-only chipmaker EV/EBITDA and compared it with the non-US chipmaker EV/EBITDA above. The overall chipmaker average is 21.58x, closer to our calculated reverse DCF model multiple of 19.17x. However, the US-only chipmakers have a much higher average of 27.78x compared to 9.96x for non-US chipmakers. Thus, this indicates a premium for US-only chipmakers, which we believe is more appropriate for Nvidia’s valuation.<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">EV/EBITDA and CAGR Comparison</td><td style=\"text-align:left;\">EV/EBITDA (5-yr)</td><td style=\"text-align:left;\">5-year Revenue CAGR</td></tr><tr><td style=\"text-align:left;\">Logic</td><td style=\"text-align:left;\">33.14x</td><td style=\"text-align:left;\">16.85%</td></tr><tr><td style=\"text-align:left;\">DAO</td><td style=\"text-align:left;\">19.89x</td><td style=\"text-align:left;\">10.98%</td></tr><tr><td style=\"text-align:left;\">Memory</td><td style=\"text-align:left;\">6.36x</td><td style=\"text-align:left;\">-3.66%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, we categorized companies within the industry based on their primary business segments. The logic segment demonstrates the highest multiple of 33.14x along with the highest CAGR of 16.85. On the other hand, the memory segment shows a relatively lower multiple at 6.36x with a negative CAGR of -3.66%. Hence, with the Logic segment’s higher EV/EBITDA, we believe this supports a higher multiple used for Nvidia’s valuation as a Logic chipmaker.<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">EV/EBITDA and CAGR Comparison</td><td style=\"text-align:left;\">EV/EBITDA (5-yr)</td><td style=\"text-align:left;\">5-year Revenue CAGR</td></tr><tr><td style=\"text-align:left;\">Higher Growth (20%+)</td><td style=\"text-align:left;\">45.80x</td><td style=\"text-align:left;\">31.22%</td></tr><tr><td style=\"text-align:left;\">Medium Growth (10% to 20%)</td><td style=\"text-align:left;\">21.94x</td><td style=\"text-align:left;\">11.67%</td></tr><tr><td style=\"text-align:left;\">Low Growth (0% to 10%)</td><td style=\"text-align:left;\">13.10x</td><td style=\"text-align:left;\">4.41%</td></tr><tr><td style=\"text-align:left;\">Negative Growth (&lt;0%)</td><td style=\"text-align:left;\">8.00x</td><td style=\"text-align:left;\">-6.12%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsAdditionally, we further compiled the above table by stratifying major companies within the industry by their respective growth rates. We observed that companies experiencing higher growth (20%+) hold the highest multiples at 45.80x along with a high CAGR of 31.22%. This contrasts with the companies possessing negative growth (&lt;0%), which are valued at a lower multiple of 8.00x with a CAGR of -6.12%. Thus, this further supports Nvidia’s higher multiple used as it aligns with the high growth category at a CAGR of 49.37%.<h3 id=\"id_2193681664\">Company Multiple</h3><img src=\"https://community-static.tradeup.com/news/8d03fd23857db8c07e637debc107c290\" tg-width=\"635\" tg-height=\"439\"/>YCharts<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Nvidia</td><td style=\"text-align:left;\">2019</td><td style=\"text-align:left;\">2020</td><td style=\"text-align:left;\">2021</td><td style=\"text-align:left;\">2022</td><td style=\"text-align:left;\">2023</td><td style=\"text-align:left;\">Average</td></tr><tr><td style=\"text-align:left;\">EV/EBITDA</td><td style=\"text-align:left;\">42.18x</td><td style=\"text-align:left;\">55.91x</td><td style=\"text-align:left;\">53.23x</td><td style=\"text-align:left;\">80.26x</td><td style=\"text-align:left;\">42.19x</td><td style=\"text-align:left;\">54.75x</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, we analyzed the company’s historical EV/EBITDA above. The company’s EV/EBITDA has been volatile with a significant spike in 2023 but has moderated since. Its 5-year median multiple shows a rising trend. Its 5-year average is 54.75x, thus using its 5-year average EV/EBITDA would result in a much higher terminal value and valuation of the company compared to our assumption based on the 5-year average US-only chipmakers.<h3 id=\"id_2858467795\">Forward Multiple</h3><table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">EV/EBITDA Average</td><td style=\"text-align:left;\">Multiple</td></tr><tr><td style=\"text-align:left;\">US Only</td><td style=\"text-align:left;\">27.78x</td></tr><tr><td style=\"text-align:left;\">Logic Chipmaker</td><td style=\"text-align:left;\">33.14x</td></tr><tr><td style=\"text-align:left;\">High Growth</td><td style=\"text-align:left;\">45.80x</td></tr><tr><td style=\"text-align:left;\">Nvidia Historical</td><td style=\"text-align:left;\">54.75x</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsOverall, we believe our EV/EBITDA assumption used of 27.78x, based on the US-only average chipmakers, is more appropriate compared to our derived EV/EBITDA based on the reverse DCF model which is more in line with the overall semicon chipmaker average. Furthermore, comparing the average EV/EBITDA in the table summary above, the US-only average has the lowest compared to Logic Chipmaker, High Growth, and Nvidia&#39;s Historical average; thus we believe our assumption is relatively conservative.<h2 id=\"id_28603474\">Strong Profit Margins</h2>Finally, we conducted the reverse DCF analysis on its profitability margins, holding all other assumptions constant. We adjusted the reverse DCF model to obtain the EBIT and FCF margins based on the current market price.<img src=\"https://community-static.tradeup.com/news/fcd2847ba36389af6308a3cbdf5c1810\" tg-width=\"640\" tg-height=\"360\"/>Company Data, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Comparison</td><td style=\"text-align:left;\">EBIT Margin</td><td style=\"text-align:left;\">FCF Margin</td></tr><tr><td style=\"text-align:left;\">Reverse DCF</td><td style=\"text-align:left;\">41.6%</td><td style=\"text-align:left;\">24.8%</td></tr><tr><td style=\"text-align:left;\">Our Previous Analysis</td><td style=\"text-align:left;\">56.2%</td><td style=\"text-align:left;\">39.4%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsAs seen above, the reverse DCF model calculated an average 5-year forward EBIT margin of 41.6% compared to 56.2% for our assumptions based on our previous analysis. Additionally, in terms of FCF margins, the reverse DFC model implies an average 24.8% forward FCF margin, lower compared to our forecasts of 39.4%.<h3 id=\"id_4010505352\">Gross and EBIT Margins</h3><img src=\"https://community-static.tradeup.com/news/15df2ef56f27d9e3522347767415845f\" tg-width=\"640\" tg-height=\"360\"/>Company Data, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Earnings &amp; Margins</td><td style=\"text-align:left;\">2014</td><td style=\"text-align:left;\">2015</td><td style=\"text-align:left;\">2016</td><td style=\"text-align:left;\">2017</td><td style=\"text-align:left;\">2018</td><td style=\"text-align:left;\">2019</td><td style=\"text-align:left;\">2020</td><td style=\"text-align:left;\">2021</td><td style=\"text-align:left;\">2022</td><td style=\"text-align:left;\">2023</td><td style=\"text-align:left;\">TTM</td><td style=\"text-align:left;\">5-year Average</td></tr><tr><td style=\"text-align:left;\">Gross Margin</td><td style=\"text-align:left;\">55.51%</td><td style=\"text-align:left;\">56.11%</td><td style=\"text-align:left;\">58.80%</td><td style=\"text-align:left;\">59.93%</td><td style=\"text-align:left;\">61.21%</td><td style=\"text-align:left;\">61.99%</td><td style=\"text-align:left;\">63.31%</td><td style=\"text-align:left;\">64.93%</td><td style=\"text-align:left;\">56.93%</td><td style=\"text-align:left;\">72.72%</td><td style=\"text-align:left;\">75.29%</td><td style=\"text-align:left;\">66.6%</td></tr><tr><td style=\"text-align:left;\">EBIT Margin</td><td style=\"text-align:left;\">16.21%</td><td style=\"text-align:left;\">17.52%</td><td style=\"text-align:left;\">28.03%</td><td style=\"text-align:left;\">33.05%</td><td style=\"text-align:left;\">32.47%</td><td style=\"text-align:left;\">26.07%</td><td style=\"text-align:left;\">28.31%</td><td style=\"text-align:left;\">37.31%</td><td style=\"text-align:left;\">20.68%</td><td style=\"text-align:left;\">54.12%</td><td style=\"text-align:left;\">59.85%</td><td style=\"text-align:left;\">40.05%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the earnings and margins chart above, our derived average forward EBIT margin based on the reverse DCF model (41.6%) corresponds the closest to the company’s historical EBIT margin in 2021 (37.31%). In comparison, our projected EBIT margin (56.2%) is closer to the company’s historical margin in 2023 (54.12%).Therefore, we analyzed the specific differences between the company’s margins in 2021 and improvement in 2023. Firstly, in terms of gross margins, the company’s gross margins improved from 64.3% in 2021 to 72.72% in 2023 which indicates it benefited from economies of scale and contributed to its EBIT margin increase. Also, we analyzed in our previous analysis one main reason for the decline in margins in 2022 was due to inventory provisions which had “impacted its gross margins by 7.5%” because of “the slump in the GPU market as its shipments fell by 48%” that year, but expected it to be non-recurring as the GPU market improved beyond that.<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Nvidia</td><td style=\"text-align:left;\">2019</td><td style=\"text-align:left;\">2020</td><td style=\"text-align:left;\">2021</td><td style=\"text-align:left;\">2022</td><td style=\"text-align:left;\">2023</td><td style=\"text-align:left;\">Average</td></tr><tr><td style=\"text-align:left;\">Employee Headcount</td><td style=\"text-align:left;\">13,775</td><td style=\"text-align:left;\">18,975</td><td style=\"text-align:left;\">22,473</td><td style=\"text-align:left;\">26,196</td><td style=\"text-align:left;\">29,600</td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">Growth %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">37.7%</td><td style=\"text-align:left;\">18.4%</td><td style=\"text-align:left;\">16.6%</td><td style=\"text-align:left;\">13.0%</td><td style=\"text-align:left;\">21.4%</td></tr><tr><td style=\"text-align:left;\">SG&amp;A ($ mln)</td><td style=\"text-align:left;\">1,093</td><td style=\"text-align:left;\">1,912</td><td style=\"text-align:left;\">2,166</td><td style=\"text-align:left;\">2,440</td><td style=\"text-align:left;\">2,654</td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">Growth %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">74.9%</td><td style=\"text-align:left;\">13.3%</td><td style=\"text-align:left;\">12.7%</td><td style=\"text-align:left;\">8.8%</td><td style=\"text-align:left;\">27.4%</td></tr><tr><td style=\"text-align:left;\">SG&amp;A % of Revenue</td><td style=\"text-align:left;\">10.01%</td><td style=\"text-align:left;\">11.47%</td><td style=\"text-align:left;\">8.05%</td><td style=\"text-align:left;\">9.05%</td><td style=\"text-align:left;\">4.36%</td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">R&amp;D ($ mln)</td><td style=\"text-align:left;\">2,829</td><td style=\"text-align:left;\">3,924</td><td style=\"text-align:left;\">5,268</td><td style=\"text-align:left;\">7,339</td><td style=\"text-align:left;\">8,675</td><td style=\"text-align:left;\"></td></tr><tr><td style=\"text-align:left;\">Growth %</td><td style=\"text-align:left;\"></td><td style=\"text-align:left;\">38.7%</td><td style=\"text-align:left;\">34.3%</td><td style=\"text-align:left;\">39.3%</td><td style=\"text-align:left;\">18.2%</td><td style=\"text-align:left;\">32.6%</td></tr><tr><td style=\"text-align:left;\">R&amp;D % of Revenue</td><td style=\"text-align:left;\">25.91%</td><td style=\"text-align:left;\">23.53%</td><td style=\"text-align:left;\">19.57%</td><td style=\"text-align:left;\">27.21%</td><td style=\"text-align:left;\">14.24%</td><td style=\"text-align:left;\"></td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, another factor in the increased margins is lower growth in SG&amp;A and R&amp;D expenses. As stated by management, the increase in R&amp;D as well as SG&amp;A for fiscal year 2023 “was primarily driven by increased compensation and employee growth”. We compared the employee growth of Nvidia with its SG&amp;A and R&amp;D expense growth. As seen in the table, the company’s SG&amp;A expense growth has been lower than employee growth and revenue growth between 2021 to 2023, leading to lower SG&amp;A % of revenue from 8.05% to 4.36%. We believe this could be due to the company’s business model with an ecosystem of partners including a network of cloud providers, PC &amp; server makers, and distributors, allowing the company to depend on its established partner ecosystems to increase sales rather than requiring to scale its direct sales force.On the other hand, its R&amp;D expenses growth had been higher than employee growth over the period. However, its R&amp;D expenses growth had been lower than revenue growth, leading to a decline of R&amp;D % of Revenue from 19.57% in 2021 to 14.24% in 2023. We believe the reason for this is related to its business model, as the company is a chip designer. It designs main chip architectures such as Blackwell and focuses on developments to improve its chipset architectures; thus we believe its business model primarily requires better-skilled employees and talent which the company could attract by increasing its compensation, explaining the higher expense growth compared to employee growth.Overall, we believe the company could at least sustain its margins around its historical 2023 levels, by maintaining its level of economies of scale as its sales capacity is much higher compared to 2021. Additionally, we expect its operating expenses (R&amp;D and SG&amp;A) % of Revenue to be maintained due to its business model. We believe the company’s partner ecosystem is more established compared to 2021 as it has expanded partnerships with key cloud partners as well as management highlighted in its recent earnings briefing that “Blackwell will be available in over 100 OEM and ODM systems at launch, more than double the number of Hopper&#39;s launch”. Furthermore, we believe its increased compensation levels could indicate its ability to attract and retain better-skilled employees compared to 2021.<h3 id=\"id_2296655564\">FCF Margins</h3><img src=\"https://community-static.tradeup.com/news/e81bbff1ef994c48084383e3ddc0bf29\" tg-width=\"640\" tg-height=\"360\"/>Company Data, Khaveen Investments<table style=\"border-collapse:collapse;\"><tbody><tr><td style=\"text-align:left;\">Earnings &amp; Margins</td><td style=\"text-align:left;\">2014</td><td style=\"text-align:left;\">2015</td><td style=\"text-align:left;\">2016</td><td style=\"text-align:left;\">2017</td><td style=\"text-align:left;\">2018</td><td style=\"text-align:left;\">2019</td><td style=\"text-align:left;\">2020</td><td style=\"text-align:left;\">2021</td><td style=\"text-align:left;\">2022</td><td style=\"text-align:left;\">2023</td><td style=\"text-align:left;\">TTM</td><td style=\"text-align:left;\">5-yr Avg</td></tr><tr><td style=\"text-align:left;\">Free Cash Flow Margin (Capex Only)</td><td style=\"text-align:left;\">17.05%</td><td style=\"text-align:left;\">21.86%</td><td style=\"text-align:left;\">21.70%</td><td style=\"text-align:left;\">29.88%</td><td style=\"text-align:left;\">26.30%</td><td style=\"text-align:left;\">38.22%</td><td style=\"text-align:left;\">28.75%</td><td style=\"text-align:left;\">30.82%</td><td style=\"text-align:left;\">14.10%</td><td style=\"text-align:left;\">43.56%</td><td style=\"text-align:left;\">48.49%</td><td style=\"text-align:left;\">33.1%</td></tr><tr><td style=\"text-align:left;\">Adjusted CapEx/Revenue</td><td style=\"text-align:left;\">2.6%</td><td style=\"text-align:left;\">1.7%</td><td style=\"text-align:left;\">2.5%</td><td style=\"text-align:left;\">6.1%</td><td style=\"text-align:left;\">5.1%</td><td style=\"text-align:left;\">4.5%</td><td style=\"text-align:left;\">6.8%</td><td style=\"text-align:left;\">3.6%</td><td style=\"text-align:left;\">6.8%</td><td style=\"text-align:left;\">1.8%</td><td style=\"text-align:left;\">1.5%</td><td style=\"text-align:left;\">4.1%</td></tr><tr><td style=\"text-align:left;\">Adjusted CapEx/Fixed Assets</td><td style=\"text-align:left;\">8.2%</td><td style=\"text-align:left;\">6.5%</td><td style=\"text-align:left;\">13.5%</td><td style=\"text-align:left;\">29.9%</td><td style=\"text-align:left;\">21.9%</td><td style=\"text-align:left;\">13.5%</td><td style=\"text-align:left;\">8.9%</td><td style=\"text-align:left;\">6.4%</td><td style=\"text-align:left;\">10.1%</td><td style=\"text-align:left;\">5.0%</td><td style=\"text-align:left;\">5.1%</td><td style=\"text-align:left;\">7.1%</td></tr></tbody></table>Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, the reverse DCF model average forward FCF margins (24.8%) are closer to the company’s historical FCF margins in 2020 (28.75%). In comparison, our projected forward average FCF margin (39.4%) is higher and more aligned with its 2023 FCF margin (43.56%). The improvement of the company’s FCF margin is mainly due to the rise in its EBIT margin as explained above from 2021 to 2023 (16.8%). Furthermore, another factor for the increase in FCF margins is due to its higher EBIT margins as explained above as well as lower capex intensity. From the table above, its capex % of revenue and fixed assets declined in 2023 and TTM to its lowest levels over the period which we believe is due to the lean operating business model of Nvidia as a fabless chipmaker with minimal capex requirements to support revenue growth, depending on foundry partners such as TSMC.<h3 id=\"id_2563638781\">Forward Margins</h3>Overall, the reverse DCF model margins are lower compared to our projected margins in terms of both EBIT and FCF margins. The reverse DCF EBIT average forward margins (41.6%) are in line with the company’s historical 2021 margin levels, while our projected margins are closer to its 2023 margins. We believe our margin assumptions are more appropriate as the company’s EBIT margin has improved between 2021 to 2023 due to improved economies of scale with higher gross margins as well as its lower operating expenses including SG&amp;A and R&amp;D due to its business model depending on its established partner ecosystem for sales and its R&amp;D activities as a chip designer. Furthermore, we expect its higher sales capacity and improved partner ecosystem and enhanced R&amp;D capabilities to sustain its current margin levels going forward; thus we believe our forecasted margins to be more appropriate. Additionally, we highlighted the improved EBIT margins supporting its FCF, which bodes well for its FCF margins.<h2 id=\"id_2066159317\">Risk: Growth Outlook</h2>Among the 3 factors analyzed, we believe our assumptions are appropriate compared to the reversed DCF model. However, we believe revenue growth could be a risk due to competitive factors as AMD is expected to release more details for its upcoming launch of its MI350X, and data center customers invest in their chip development which leads us to await further details on their performance capabilities to determine whether they could be formidable rivals to Nvidia. Based on our projections previously derived based on market growth and competitive factor for Nvidia, the competitive factor contributes 21% of growth for its Data Center segment. Further, Nvidia faces geopolitical risks such as in China with the imposition of sanctions on sales to Chinese chipmakers, which could affect its competitiveness in that market. China accounted for 17% of its total revenues in FY2024.<h2 id=\"id_1998547241\">Valuation</h2><img src=\"https://community-static.tradeup.com/news/4048dc1eeb437f7d3203c3adf2c80170\" tg-width=\"640\" tg-height=\"360\"/>Khaveen InvestmentsWe maintain the same assumption for our valuation from our previous analysis, as we have thoroughly justified the appropriateness in this analysis but with a lower discount rate of 13.5% (company’s WACC). This includes revenue growth (57.1%), EV/EBITDA based on US-only chipmakers (27.78x), average forward EBIT (56.2%), and FCF margin (39.4%) and derived an upside of 8.84%.<h2 id=\"id_3188397411\">Verdict</h2>We believe our 5-year forward average revenue growth projections for Nvidia are appropriate, albeit slightly higher than our market-implied growth assumption (44.8% vs 41.2%), driven by the execution of its product roadmap with new products including Blackwell supporting its competitiveness. Furthermore, we believe Nvidia&#39;s strong partnerships with major cloud providers further could contribute to sustained growth. In terms of the terminal value, we believe our EV/EBITDA assumption of 27.78x, based on the US-only average for chipmakers, is more appropriate than the reverse DCF model-derived value, which aligns with the overall semiconductor average. The US-only average is the lowest compared to Logic Chipmaker, High Growth, and Nvidia&#39;s historical averages, making our assumption relatively conservative. Finally, we believe our higher margin projections compared to the reverse DCF are more appropriate, reflecting Nvidia&#39;s improved economies of scale, higher gross margins, and lower operating expenses. Also, we expect its higher sales capacity, robust partner ecosystem, and advanced R&amp;D capabilities to sustain these margins and positively contribute to its robust FCFs. However, Nvidia’s share price has rocketed by 43% in the past month and 58% since our last coverage, meeting our price target while overtaking Apple and Microsoft as the most valuable company. Therefore, we believe its current valuation is appropriate, and we have updated our latest price target of $147.57 (14% higher than $129.71 previously mainly due to a lower discount rate) indicates limited upside in the near term of 8.84% based on our current conservative assumptions; thus we now downgrade it as a Hold.</body></html>","source":"seekingalpha","collect":0,"html":"<!DOCTYPE html>\n<html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<meta name=\"viewport\" content=\"width=device-width,initial-scale=1.0,minimum-scale=1.0,maximum-scale=1.0,user-scalable=no\"/>\n<meta name=\"format-detection\" content=\"telephone=no,email=no,address=no\" />\n<title>Nvidia: Valuation Seems Reasonable After Recent Rally</title>\n<style type=\"text/css\">\na,abbr,acronym,address,applet,article,aside,audio,b,big,blockquote,body,canvas,caption,center,cite,code,dd,del,details,dfn,div,dl,dt,\nem,embed,fieldset,figcaption,figure,footer,form,h1,h2,h3,h4,h5,h6,header,hgroup,html,i,iframe,img,ins,kbd,label,legend,li,mark,menu,nav,\nobject,ol,output,p,pre,q,ruby,s,samp,section,small,span,strike,strong,sub,summary,sup,table,tbody,td,tfoot,th,thead,time,tr,tt,u,ul,var,video{ font:inherit;margin:0;padding:0;vertical-align:baseline;border:0 }\nbody{ font-size:16px; line-height:1.5; color:#999; background:transparent; }\n.wrapper{ overflow:hidden;word-break:break-all;padding:10px; }\nh1,h2{ font-weight:normal; line-height:1.35; margin-bottom:.6em; }\nh3,h4,h5,h6{ line-height:1.35; margin-bottom:1em; }\nh1{ font-size:24px; }\nh2{ font-size:20px; }\nh3{ font-size:18px; }\nh4{ font-size:16px; }\nh5{ font-size:14px; }\nh6{ font-size:12px; }\np,ul,ol,blockquote,dl,table{ margin:1.2em 0; }\nul,ol{ margin-left:2em; }\nul{ list-style:disc; }\nol{ list-style:decimal; }\nli,li p{ margin:10px 0;}\nimg{ max-width:100%;display:block;margin:0 auto 1em; }\nblockquote{ color:#B5B2B1; border-left:3px solid #aaa; padding:1em; }\nstrong,b{font-weight:bold;}\nem,i{font-style:italic;}\ntable{ width:100%;border-collapse:collapse;border-spacing:1px;margin:1em 0;font-size:.9em; }\nth,td{ padding:5px;text-align:left;border:1px solid #aaa; }\nth{ font-weight:bold;background:#5d5d5d; }\n.symbol-link{font-weight:bold;}\n/* header{ border-bottom:1px solid #494756; } */\n.title{ margin:0 0 8px;line-height:1.3;color:#ddd; }\n.meta {color:#5e5c6d;font-size:13px;margin:0 0 .5em; }\na{text-decoration:none; color:#2a4b87;}\n.meta .head { display: inline-block; overflow: hidden}\n.head .h-thumb { width: 30px; height: 30px; margin: 0; padding: 0; border-radius: 50%; float: left;}\n.head .h-content { margin: 0; padding: 0 0 0 9px; float: left;}\n.head .h-name {font-size: 13px; color: #eee; margin: 0;}\n.head .h-time {font-size: 11px; color: #7E829C; margin: 0;line-height: 11px;}\n.small {font-size: 12.5px; display: inline-block; transform: scale(0.9); -webkit-transform: scale(0.9); transform-origin: left; -webkit-transform-origin: left;}\n.smaller {font-size: 12.5px; display: inline-block; transform: scale(0.8); -webkit-transform: scale(0.8); transform-origin: left; -webkit-transform-origin: left;}\n.bt-text {font-size: 12px;margin: 1.5em 0 0 0}\n.bt-text p {margin: 0}\n</style>\n</head>\n<body>\n<div class=\"wrapper\">\n<header>\n<h2 class=\"title\">\nNvidia: Valuation Seems Reasonable After Recent Rally\n</h2>\n\n<h4 class=\"meta\">\n\n\n2024-06-20 11:24 GMT+8&nbsp;&nbsp;&nbsp;<a href=https://seekingalpha.com/article/4699963-nvidia-stock-valuation-seems-reasonable-after-recent-rally-downgrade-hold>seekingalpha</a>\n\n\n</h4>\n\n</header>\n<article>\n<div>\nNvidia has passed Apple and Microsoft to become the largest company in the world, and we evaluate if the company deserves such high of a valuation.We use a Reverse DCF model to obtain the market-...\n\n<a href=\"https://seekingalpha.com/article/4699963-nvidia-stock-valuation-seems-reasonable-after-recent-rally-downgrade-hold\">Web Link</a>\n\n</div>\n\n\n</article>\n</div>\n</body>\n</html>\n","type":0,"thumbnail":"","relate_stocks":{"LU0949170772.SGD":"Blackrock Global Equity Income A6 SGD-H","LU0289960550.SGD":"AB FCP I - GLOBAL EQUITY BLEND PORTFOLIO 'A' (SGD) ACC","LU1691799644.USD":"Amundi Funds Polen Capital Global Growth A2 (C) USD","AMD":"美国超微公司","LU1064131342.USD":"Fullerton Lux Funds - Global Absolute Alpha A Acc USD","GOOG":"谷歌","LU0985489474.SGD":"First Eagle Amundi International AHS-C SGD-H","LU0354030511.USD":"ALLSPRING U.S. LARGE CAP GROWTH \"I\" (USD) ACC","LU0149725797.USD":"汇丰美国股市经济规模基金","LU0738911758.USD":"Blackrock Global Equity Income A6 USD","GOOGL":"谷歌A","LU0878866978.SGD":"First Eagle Amundi International AHS-QD SGD-H","IE00BJTD4N35.SGD":"Neuberger Berman US Long Short Equity A1 Acc SGD-H","BK4548":"巴美列捷福持仓","LU2098885051.SGD":"JPMorgan Funds - Multi-Manager Alternatives A (acc) SGD","IE00B775SV38.USD":"NEUBERGER BERMAN US MULTICAP OPPORTUNITIES \"A\" (USD) ACC","LU2458330169.SGD":"FRANKLIN SHARIAH TECHNOLOGY \"A\" (SGD) ACC","IE00BFSS7M15.SGD":"Janus Henderson Balanced A Acc SGD-H","IE00B7SZLL34.SGD":"Legg Mason ClearBridge - Value A Acc SGD-H","BK4516":"特朗普概念","BK4554":"元宇宙及AR概念","BK4515":"5G概念","LU0109391861.USD":"富兰克林美国机遇基金A Acc","LU0068578508.USD":"First Eagle Amundi International Cl AU-C USD","INTC":"英特尔","LU1988902786.USD":"FULLERTON LUX FUNDS GLOBAL ABSOLUTE ALPHA \"I\" (USD) ACC","LU1506573853.SGD":"MANULIFE GF GLOBAL EQUITY \"AA\" (SGD) INC","LU0048573561.USD":"FIDELITY AMERICA \"A\" (USD) INC","NVDA":"英伟达","BK4122":"互联网与直销零售","BK4585":"ETF&股票定投概念","ORCL":"甲骨文","LU0056508442.USD":"贝莱德世界科技基金A2","BK4227":"交易和支付处理服务","LU1803068979.SGD":"FTIF - Franklin Technology A (acc) SGD-H1","BK4575":"芯片概念","BK4587":"ChatGPT概念","BK4525":"远程办公概念","IE00BBT3K403.USD":"LEGG MASON CLEARBRIDGE TACTICAL DIVIDEND INCOME \"A(USD) ACC","LU1804176565.USD":"EASTSPRING INV GLOBAL GROWTH EQUITY \"A\" (USD) ACC","LU0171293334.USD":"贝莱德英国基金A2","BK4077":"互动媒体与服务","LU0310800965.SGD":"FTIF - Templeton Global Balanced A Acc SGD","LU1489326972.SGD":"First Eagle Amundi International AHS-MD SGD-H","LU0061474705.USD":"THREADNEEDLE (LUX) GLOBAL DYNAMIC REAL RETURN \"AU\" (USD) ACC","UBS":"瑞银","MSFT":"微软","BK4574":"无人驾驶"},"source_url":"https://seekingalpha.com/article/4699963-nvidia-stock-valuation-seems-reasonable-after-recent-rally-downgrade-hold","is_english":true,"share_image_url":"https://static.laohu8.com/5a36db9d73b4222bc376d24ccc48c8a4","article_id":"2444429849","content_text":"Nvidia has passed Apple and Microsoft to become the largest company in the world, and we evaluate if the company deserves such high of a valuation.We use a Reverse DCF model to obtain the market-implied assumptions of the company's growth rate, terminal value multiples, and profit margins and compare them against our own assumptions.We believe the company's Blackwell architecture launch and expanded cloud partnerships with major providers contribute to its competitiveness and growth outlook.BING-JHEN HONGIn our previous analysis of NVIDIA Corporation (NASDAQ:NVDA) (NEOE:NVDA:CA), we highlighted the strong performance of Nvidia’s growth of 126% in CY2023, and we remained positive on Nvidia’s Data Center segment, supported by cloud partnerships and sovereign AI initiatives. For 2024, we projected a 58% overall revenue growth, led by the Data Center, with contributions from Gaming, Professional Visualization, and a recovery in PC GPU revenue, as well as 20.6% growth in the Automotive segment.Since our last analysis in March, we started reviewing Nvidia again as its share price had risen by 28% in May, with a further upside of 18.3% left based on our previous price target of $1,294.14 ($129.41 post-stock split). Therefore, we initially aimed to determine whether our expectations for Nvidia’s upside were justified. Since then, Nvidia has overtaken both Apple and Microsoft as the largest company globally by market cap, rising by a further 30% and meeting our price target. Firstly, we conducted a reverse DCF analysis determining the market implied revenue growth forecast, terminal value multiple, and profitability margins. Based on that, we compared the reverse DCF model with our revenue, terminal value and margin assumptions and determined whether our assumptions were appropriate.Sustainable Revenue GrowthFirstly, we examined whether our revenue growth forecast is appropriate. Based on the reverse DCF model, holding all previous assumptions constant, we derived the market-implied revenue growth rate for Nvidia based on the current market price.Company Data, Khaveen InvestmentsRevenue Comparison ($ mln)2024F2025F2026F2027F2028F2029F2030F2031F2032F2033FAvr. (5-yr)Avr. (10-yr)Revenue (Reverse DCF)113,185163,293220,004273,487320,027358,148388,012410,660427,439439,664Revenue Growth85.8%44.3%34.7%24.3%17.0%11.9%8.3%5.8%4.1%2.9%41.2%23.9%Revenue (Our Previous Forecast)95,720148,848225,273306,238383,284450,785506,357550,053583,280607,943Revenue Growth57.1%55.5%51.3%35.9%25.2%17.6%12.3%8.6%6.0%4.2%45.0%27.4%Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the reverse DCF model, we derived the revenue growth rate at a forward average of 41.2% in the next 5 years, compared to our average forward growth forecast of 44.8%. Furthermore, we derived a 10-year forward growth of 23.9% from the reverse DCF model, compared to 27.4% from our projections. Therefore, this indicates our revenue growth assumption is higher compared to the calculated growth rate from the reverse DCF model. In our previous analysis, we highlighted that we maintained a more stable long-term growth projection “derived independently based on individual segment projections and modeling” for the company “driven primarily due to its Data Center segment at a forward average of 60.9% underpinned by its AI leadership”.We examined the latest developments of the company which could support its growth outlook below, including its product launch of its Blackwell architecture, expanded partnerships, and competition.Competitiveness (Blackwell Launch)Nvidia announced the next-gen Blackwell architecture in March 2024, claiming that it enhances computing and generative AI by providing improvements in scale, performance, and efficiency, with customer data centers planned to be operational by Q4 2024. The Blackwell-architecture GPUs are “manufactured using a custom-built 4NP TSMC process” with 208 bln transistors. Previously, we had highlighted the company’s Blackwell architecture based on the company’s provided roadmap in 2024; thus, we believe this indicates its execution on its continued product launches for Data Center.CategoryH100H200B100B200GB200 SuperchipPrice$24,000$24,000$30,000$30,000-$40,000$60,000-$70,000Increase %0.0%25.0%16.7%85.7%Memory Capacity (GB)80.00141.00192.00192.00384.00Memory Bandwidth (GB/s)3,352.004,800.008,000.008,000.0016,000.00Improvement %43.2%66.7%0.0%100.0%TF32 (Teraflops)495.00495.00900.001,100.002,500.00Improvement %0.0%81.8%22.2%127.3%FP16 (Teraflops)989.00989.001,750.002,250.005,000.00Improvement %0.0%76.9%28.6%122.2%FP8 (Teraflops)1,979.001,979.003,500.004,500.0010,000.00Improvement %0.0%76.9%28.6%122.2%FP4 (Teraflops)1,979.001,979.007,000.009,000.0020,000.00Improvement %0.0%253.7%28.6%122.2%Click to enlargeSource: Company Data, Khaveen InvestmentsAccording to the above table comparing the GPUs across Nvidia’s portfolio, the Blackwell chips (B100 and B200) show increasing capabilities including memory capacity, memory bandwidth, and improvements in floating-point performance metrics which play important roles in deep learning model training, as well as higher pricing compared to the Hopper chips (H100 and H200).Additionally, the GB200 Superchip which consists of two Blackwell GPUs and one Grace CPU, is reported to have a price range of $60,000-$70,000, along with doubled memory capacity and bandwidth in comparison to previous gen B200 and more than 100% of improvements in the above listed floating-point performance metrics for deep learning model training.Furthermore, we updated our comparison of the top data center GPU chips of Nvidia with its top competitors. Besides Nvidia, AMD (AMD) is expected to launch its new MI350 chip this year and Intel’s (INTC) new Gaudi 3 is expected to be released in Q2. Furthermore, Nvidia highlighted competition in China from Huawei which is expected to release its new and improved Ascend 910B chip this year as well following its Ascend 910 introduction in 2019.Data Center GPU Comparison MetricsAMD (MI300x)AMD (MI350)Nvidia (B100)Nvidia (GB200 Superchip)Intel (Gaudi 2)Intel (Gaudi 3)Huawei (Ascend 910B)Processor5nm (TSMC)4nm (TSMC)4nm (TSMC)4nm (TSMC)7nm (TSMC)5nm (TSMC)7nmTransistors ('bln')153-208208>100--FP16 Peak (Teraflops)1,532-3,50010,0008391,835320INT 8 (Teraflops)3,064-7,00040,0001,628-640Memory Clock1.9GHz-8Gbps HBM3E8Gbps HBM3E1.56GHz3.6GHz-Memory Capacity192GB288GB192GB384GB96GB128GB64GBMemory Bandwidth5.2TB/sec-8TB/sec16TB/sec2.45TB/sec3.7TB/sec-Interconnect Bandwidth896 GB/sec-1800GB/s1800GB/s300GB/s600GB/s400G/sMax Power Consumption700W-700W1000W600W900W310WClick to enlargeSource: Company Data, Khaveen InvestmentsBased on the table, Nvidia’s Blackwell chips show superior performance capabilities, especially GB200 Superchip which stands out with its 4nm processor, highest transistor counts at 208 bln, and exceptional FP16 and INT8 peak performance metrics, 10,000 teraflops and 40,000 teraflops, respectively. It also offers the largest memory capacity of 384GB and the highest memory bandwidth at 16TB/sec, although it comes with the highest power consumption of 1000W.In contrast, AMD has yet to provide details on its MI350 product with limited information about it besides its 4nm process and 288Gb memory capacity, which is lower than the Nvidia GB200. The previous AMD MI300x was built on a 5nm process, providing 1,532 teraflops of FP16 and 3,064 teraflops of INT8 performance with a memory capacity of 192GB and bandwidth of 5.2TB/sec, consuming 700W.Furthermore, Intel Gaudi 3 also uses 5nm technology, with 1,835 teraflops FP16 and 128GB memory but has a higher power consumption of 900W. Notably, the Huawei Ascend 910B, although using older 7nm technology, has the lowest power consumption of 310W.Overall, our forward average revenue growth rate projections (45%) are higher compared to the market implied growth rate through the reverse DCF model (41.2%), a difference of 3.8% which we expect due to the company’s enhanced competitiveness. For example, we believe the improvements in Nvidia’s data center GPU capabilities with Blackwell highlight the company’s execution of its product roadmap and commitment to future product launches, which could cement its competitiveness.PartnershipsCompanyPurpose of Adoption of BlackwellCloud Market ShareAvr. Revenue Growth (3-yr)Cloud Revenue ($ bln)AWSTo securely build and run large AI models on AWS34%26.4%90.76Google CloudTo enable Google Cloud with advanced AI and data analytics capabilities12%37.7%33.09MicrosoftTo boost Microsoft Azure's capabilities regarding large-scale AI workloads handling26%38.8%70.20OracleTo enhance Oracle's AI model capabilities for faster and higher efficiency of data processing and AI training2%30.2%5.70Click to enlargeSource: Company Data, Khaveen InvestmentsAlong with the launch of NVIDIA’s Grace Blackwell GPU platform, its cloud partners including AWS (AMZN), Google (GOOG), Microsoft (MSFT), and Oracle (ORCL) announced plans for adoption. Firstly, AWS will “offer the NVIDIA GB200 Grace Blackwell Superchip and B100 Tensor Core GPUs” and enhance the performance of its AI models with Nvidia’s Grace Blackwell GPU and DGX Cloud. Similarly, Google Cloud claims future adoption of the Grace Blackwell platform to its “AI Hypercomputer architecture in two configurations”, while also focusing on the development of trillion-parameter language models. Microsoft aims to improve its AI infrastructure with the integration of Grace Blackwell GB200 with Microsoft Azure. Oracle also plans on adopting Nvidia’s Grace Blackwell platform across its OCI Supercluster and OCI Compute, aiming to improve performance and efficiency for AI workloads such as LLM inference.Based on the table above, we believe these expanded partnerships with AWS, Google, Microsoft, and Oracle are significant for Nvidia as these major cloud providers purchase and use Nvidia’s chips. They account for 74% of the global cloud market share combined and have robust growth averages of between 26.4% to 38.8%. Furthermore, according to UBS (UBS), Microsoft is believed to be its largest customer with 19% of revenues last year.Overall, we believe the expanded partnerships of Nvidia with the major cloud providers market share combined bodes well for the sustainability of its growth outlook, which is reflected in our projections as we forecasted its forward 3-year growth rate of around 54% on average compared to the reverse DCF model growth which tapers down significantly from 86% to 35% in the next 3 years, as we expect the company’s growth to continue to be sustained by the adoption of upcoming products such as Blackwell.Current Development of CSPs With Their Own AI ChipsTrendForceHowever, according to TrendForce, several cloud service providers such as Microsoft and Google have also been developing their respective AI chip technologies recently. For instance, Microsoft is currently in the process of developing their “in-house AI chip Azure Maia 100” and Cobalt 100, which we covered in our previous analysis. Along with Google’s recent development of its tensor processing unit (TPU) of v5e, which is particularly designed for training large AI models. Despite that, Google stated it will provide the optionality for customers to use Google TPUs and Nvidia’s Blackwell GPUs. Additionally, other major companies such as Tesla, Meta, and OpenAI are also pursuing their own respective development regarding AI accelerator chips.Therefore, as competition heats up against Nvidia from key competitors, we believe our projections fairly incorporated this as we tapered down its growth rate beyond 2026. Additionally, while the development of custom AI chips by its cloud customers could affect the company’s growth, we believe Nvidia has an advantage as these companies are not specialized in chip design. While these companies have higher total company absolute R&D expenses than Microsoft, Amazon, and Google, they predominantly focus across the Software & Services (Microsoft), Consumer Discretionary Distribution & Retail (Amazon), and Media & Entertainment industries (Google). Therefore, in comparison, Nvidia is more specialized as it focuses on Semiconductors, enhancing its competitiveness with improved capabilities over competition, including custom chips from cloud providers. For example, in our analysis of Microsoft, we highlighted Microsoft’s Maia 100 GPU trailing behind Nvidia’s H100 chip specifications as it uses older process technology (5nm) and has memory capacity and bandwidth. Furthermore, the Maia 100 would lag behind compared to Nvidia’s B100 chips with an even wider gap as the B100 has twice the number of transistors and 3x higher memory capacity and bandwidth than the Maia 100. Furthermore, we also highlighted the focus on cost as a rationale for custom chip development by cloud providers, such as Microsoft emphasizing Maia enabling cheaper models for customers and “to diversify and provide choices to customers” rather than replacing Nvidia’s high-performance GPUs as well as Amazon highlighting its low-cost AI inference enabled by Inferentia2.Forward GrowthOverall, our revenue growth projections (44.8%) are higher compared to the calculated market implied revenue growth of the company (41.2%), but we believe it is appropriate given factors such as the enhanced competitiveness of Nvidia and execution on its product roadmap such as with the upcoming Blackwell launch, which we believe cements its competitive positioning against key competitors and custom chip developments by cloud providers, supporting our expectations for the company to continue outperforming the market growth as reflected by our revenue projections. Additionally, our projection which is more sustainable on a 3-year forward average compared to the market implied growth through the reverse DCF which we believe could be supported by the expanded partnerships of major cloud service providers with Nvidia.Determining Appropriate Terminal ValueFurthermore, we conducted another reverse DCF analysis, adjusting our model holding all assumptions constant except for our terminal value which is derived based on the EV/EBITDA multiple.Company Data, Khaveen InvestmentsComparisonTerminal ValueEV/EBITDAReverse DCF8,091,52119.17xOur DCF9,712,48727.78xClick to enlargeSource: Company Data, Khaveen InvestmentsBased on the table, we calculated the Reverse DCF method a terminal value of $8,091 mln with an EV/EBITDA multiple of 19.17x. In comparison, our previous analysis, however, shows a higher terminal value of $9,712 mln and an EV/EBITDA multiple of 27.78x. Previously, we based our terminal value EV/EBITDA on the 5-year US-only chipmaker average of 27.78x.Industry MultipleSeeking Alpha, Investing.com, Khaveen InvestmentsEV/EBITDA and CAGR ComparisonEV/EBITDA (5-yr)5-year Revenue CAGRUS Only27.78x11.70%European Only11.65x13.44%Asian Only8.26x6.99%Overall21.58x11.18%Click to enlargeSource: Company Data, Khaveen InvestmentsWe compiled the US-only chipmaker EV/EBITDA and compared it with the non-US chipmaker EV/EBITDA above. The overall chipmaker average is 21.58x, closer to our calculated reverse DCF model multiple of 19.17x. However, the US-only chipmakers have a much higher average of 27.78x compared to 9.96x for non-US chipmakers. Thus, this indicates a premium for US-only chipmakers, which we believe is more appropriate for Nvidia’s valuation.EV/EBITDA and CAGR ComparisonEV/EBITDA (5-yr)5-year Revenue CAGRLogic33.14x16.85%DAO19.89x10.98%Memory6.36x-3.66%Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, we categorized companies within the industry based on their primary business segments. The logic segment demonstrates the highest multiple of 33.14x along with the highest CAGR of 16.85. On the other hand, the memory segment shows a relatively lower multiple at 6.36x with a negative CAGR of -3.66%. Hence, with the Logic segment’s higher EV/EBITDA, we believe this supports a higher multiple used for Nvidia’s valuation as a Logic chipmaker.EV/EBITDA and CAGR ComparisonEV/EBITDA (5-yr)5-year Revenue CAGRHigher Growth (20%+)45.80x31.22%Medium Growth (10% to 20%)21.94x11.67%Low Growth (0% to 10%)13.10x4.41%Negative Growth (<0%)8.00x-6.12%Click to enlargeSource: Company Data, Khaveen InvestmentsAdditionally, we further compiled the above table by stratifying major companies within the industry by their respective growth rates. We observed that companies experiencing higher growth (20%+) hold the highest multiples at 45.80x along with a high CAGR of 31.22%. This contrasts with the companies possessing negative growth (<0%), which are valued at a lower multiple of 8.00x with a CAGR of -6.12%. Thus, this further supports Nvidia’s higher multiple used as it aligns with the high growth category at a CAGR of 49.37%.Company MultipleYChartsNvidia20192020202120222023AverageEV/EBITDA42.18x55.91x53.23x80.26x42.19x54.75xClick to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, we analyzed the company’s historical EV/EBITDA above. The company’s EV/EBITDA has been volatile with a significant spike in 2023 but has moderated since. Its 5-year median multiple shows a rising trend. Its 5-year average is 54.75x, thus using its 5-year average EV/EBITDA would result in a much higher terminal value and valuation of the company compared to our assumption based on the 5-year average US-only chipmakers.Forward MultipleEV/EBITDA AverageMultipleUS Only27.78xLogic Chipmaker33.14xHigh Growth45.80xNvidia Historical54.75xClick to enlargeSource: Company Data, Khaveen InvestmentsOverall, we believe our EV/EBITDA assumption used of 27.78x, based on the US-only average chipmakers, is more appropriate compared to our derived EV/EBITDA based on the reverse DCF model which is more in line with the overall semicon chipmaker average. Furthermore, comparing the average EV/EBITDA in the table summary above, the US-only average has the lowest compared to Logic Chipmaker, High Growth, and Nvidia's Historical average; thus we believe our assumption is relatively conservative.Strong Profit MarginsFinally, we conducted the reverse DCF analysis on its profitability margins, holding all other assumptions constant. We adjusted the reverse DCF model to obtain the EBIT and FCF margins based on the current market price.Company Data, Khaveen InvestmentsComparisonEBIT MarginFCF MarginReverse DCF41.6%24.8%Our Previous Analysis56.2%39.4%Click to enlargeSource: Company Data, Khaveen InvestmentsAs seen above, the reverse DCF model calculated an average 5-year forward EBIT margin of 41.6% compared to 56.2% for our assumptions based on our previous analysis. Additionally, in terms of FCF margins, the reverse DFC model implies an average 24.8% forward FCF margin, lower compared to our forecasts of 39.4%.Gross and EBIT MarginsCompany Data, Khaveen InvestmentsEarnings & Margins2014201520162017201820192020202120222023TTM5-year AverageGross Margin55.51%56.11%58.80%59.93%61.21%61.99%63.31%64.93%56.93%72.72%75.29%66.6%EBIT Margin16.21%17.52%28.03%33.05%32.47%26.07%28.31%37.31%20.68%54.12%59.85%40.05%Click to enlargeSource: Company Data, Khaveen InvestmentsBased on the earnings and margins chart above, our derived average forward EBIT margin based on the reverse DCF model (41.6%) corresponds the closest to the company’s historical EBIT margin in 2021 (37.31%). In comparison, our projected EBIT margin (56.2%) is closer to the company’s historical margin in 2023 (54.12%).Therefore, we analyzed the specific differences between the company’s margins in 2021 and improvement in 2023. Firstly, in terms of gross margins, the company’s gross margins improved from 64.3% in 2021 to 72.72% in 2023 which indicates it benefited from economies of scale and contributed to its EBIT margin increase. Also, we analyzed in our previous analysis one main reason for the decline in margins in 2022 was due to inventory provisions which had “impacted its gross margins by 7.5%” because of “the slump in the GPU market as its shipments fell by 48%” that year, but expected it to be non-recurring as the GPU market improved beyond that.Nvidia20192020202120222023AverageEmployee Headcount13,77518,97522,47326,19629,600Growth %37.7%18.4%16.6%13.0%21.4%SG&A ($ mln)1,0931,9122,1662,4402,654Growth %74.9%13.3%12.7%8.8%27.4%SG&A % of Revenue10.01%11.47%8.05%9.05%4.36%R&D ($ mln)2,8293,9245,2687,3398,675Growth %38.7%34.3%39.3%18.2%32.6%R&D % of Revenue25.91%23.53%19.57%27.21%14.24%Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, another factor in the increased margins is lower growth in SG&A and R&D expenses. As stated by management, the increase in R&D as well as SG&A for fiscal year 2023 “was primarily driven by increased compensation and employee growth”. We compared the employee growth of Nvidia with its SG&A and R&D expense growth. As seen in the table, the company’s SG&A expense growth has been lower than employee growth and revenue growth between 2021 to 2023, leading to lower SG&A % of revenue from 8.05% to 4.36%. We believe this could be due to the company’s business model with an ecosystem of partners including a network of cloud providers, PC & server makers, and distributors, allowing the company to depend on its established partner ecosystems to increase sales rather than requiring to scale its direct sales force.On the other hand, its R&D expenses growth had been higher than employee growth over the period. However, its R&D expenses growth had been lower than revenue growth, leading to a decline of R&D % of Revenue from 19.57% in 2021 to 14.24% in 2023. We believe the reason for this is related to its business model, as the company is a chip designer. It designs main chip architectures such as Blackwell and focuses on developments to improve its chipset architectures; thus we believe its business model primarily requires better-skilled employees and talent which the company could attract by increasing its compensation, explaining the higher expense growth compared to employee growth.Overall, we believe the company could at least sustain its margins around its historical 2023 levels, by maintaining its level of economies of scale as its sales capacity is much higher compared to 2021. Additionally, we expect its operating expenses (R&D and SG&A) % of Revenue to be maintained due to its business model. We believe the company’s partner ecosystem is more established compared to 2021 as it has expanded partnerships with key cloud partners as well as management highlighted in its recent earnings briefing that “Blackwell will be available in over 100 OEM and ODM systems at launch, more than double the number of Hopper's launch”. Furthermore, we believe its increased compensation levels could indicate its ability to attract and retain better-skilled employees compared to 2021.FCF MarginsCompany Data, Khaveen InvestmentsEarnings & Margins2014201520162017201820192020202120222023TTM5-yr AvgFree Cash Flow Margin (Capex Only)17.05%21.86%21.70%29.88%26.30%38.22%28.75%30.82%14.10%43.56%48.49%33.1%Adjusted CapEx/Revenue2.6%1.7%2.5%6.1%5.1%4.5%6.8%3.6%6.8%1.8%1.5%4.1%Adjusted CapEx/Fixed Assets8.2%6.5%13.5%29.9%21.9%13.5%8.9%6.4%10.1%5.0%5.1%7.1%Click to enlargeSource: Company Data, Khaveen InvestmentsFurthermore, the reverse DCF model average forward FCF margins (24.8%) are closer to the company’s historical FCF margins in 2020 (28.75%). In comparison, our projected forward average FCF margin (39.4%) is higher and more aligned with its 2023 FCF margin (43.56%). The improvement of the company’s FCF margin is mainly due to the rise in its EBIT margin as explained above from 2021 to 2023 (16.8%). Furthermore, another factor for the increase in FCF margins is due to its higher EBIT margins as explained above as well as lower capex intensity. From the table above, its capex % of revenue and fixed assets declined in 2023 and TTM to its lowest levels over the period which we believe is due to the lean operating business model of Nvidia as a fabless chipmaker with minimal capex requirements to support revenue growth, depending on foundry partners such as TSMC.Forward MarginsOverall, the reverse DCF model margins are lower compared to our projected margins in terms of both EBIT and FCF margins. The reverse DCF EBIT average forward margins (41.6%) are in line with the company’s historical 2021 margin levels, while our projected margins are closer to its 2023 margins. We believe our margin assumptions are more appropriate as the company’s EBIT margin has improved between 2021 to 2023 due to improved economies of scale with higher gross margins as well as its lower operating expenses including SG&A and R&D due to its business model depending on its established partner ecosystem for sales and its R&D activities as a chip designer. Furthermore, we expect its higher sales capacity and improved partner ecosystem and enhanced R&D capabilities to sustain its current margin levels going forward; thus we believe our forecasted margins to be more appropriate. Additionally, we highlighted the improved EBIT margins supporting its FCF, which bodes well for its FCF margins.Risk: Growth OutlookAmong the 3 factors analyzed, we believe our assumptions are appropriate compared to the reversed DCF model. However, we believe revenue growth could be a risk due to competitive factors as AMD is expected to release more details for its upcoming launch of its MI350X, and data center customers invest in their chip development which leads us to await further details on their performance capabilities to determine whether they could be formidable rivals to Nvidia. Based on our projections previously derived based on market growth and competitive factor for Nvidia, the competitive factor contributes 21% of growth for its Data Center segment. Further, Nvidia faces geopolitical risks such as in China with the imposition of sanctions on sales to Chinese chipmakers, which could affect its competitiveness in that market. China accounted for 17% of its total revenues in FY2024.ValuationKhaveen InvestmentsWe maintain the same assumption for our valuation from our previous analysis, as we have thoroughly justified the appropriateness in this analysis but with a lower discount rate of 13.5% (company’s WACC). This includes revenue growth (57.1%), EV/EBITDA based on US-only chipmakers (27.78x), average forward EBIT (56.2%), and FCF margin (39.4%) and derived an upside of 8.84%.VerdictWe believe our 5-year forward average revenue growth projections for Nvidia are appropriate, albeit slightly higher than our market-implied growth assumption (44.8% vs 41.2%), driven by the execution of its product roadmap with new products including Blackwell supporting its competitiveness. Furthermore, we believe Nvidia's strong partnerships with major cloud providers further could contribute to sustained growth. In terms of the terminal value, we believe our EV/EBITDA assumption of 27.78x, based on the US-only average for chipmakers, is more appropriate than the reverse DCF model-derived value, which aligns with the overall semiconductor average. The US-only average is the lowest compared to Logic Chipmaker, High Growth, and Nvidia's historical averages, making our assumption relatively conservative. Finally, we believe our higher margin projections compared to the reverse DCF are more appropriate, reflecting Nvidia's improved economies of scale, higher gross margins, and lower operating expenses. Also, we expect its higher sales capacity, robust partner ecosystem, and advanced R&D capabilities to sustain these margins and positively contribute to its robust FCFs. However, Nvidia’s share price has rocketed by 43% in the past month and 58% since our last coverage, meeting our price target while overtaking Apple and Microsoft as the most valuable company. Therefore, we believe its current valuation is appropriate, and we have updated our latest price target of $147.57 (14% higher than $129.71 previously mainly due to a lower discount rate) indicates limited upside in the near term of 8.84% based on our current conservative assumptions; thus we now downgrade it as a Hold.","news_type":1},"isVote":1,"tweetType":1,"viewCount":268,"commentLimit":10,"likeStatus":false,"favoriteStatus":false,"reportStatus":false,"symbols":[],"verified":2,"subType":0,"readableState":1,"langContent":"CN","currentLanguage":"CN","warmUpFlag":false,"orderFlag":false,"shareable":true,"causeOfNotShareable":"","featuresForAnalytics":[],"commentAndTweetFlag":false,"upFlag":false,"length":10,"xxTargetLangEnum":"ZH_CN"},"commentList":[],"hasMoreComment":false,"orderType":2}