Vladyslav Koptiev
$NVDA (NVIDIA Corporation) ๐—ง๐—ต๐—ฒ ๐—–๐—ฟ๐—ผ๐˜„๐—ป ๐—๐—ฒ๐˜„๐—ฒ๐—น ๐—ผ๐—ณ ๐˜๐—ต๐—ฒ ๐—”๐—œ ๐—ฅ๐—ฒ๐˜ƒ๐—ผ๐—น๐˜‚๐˜๐—ถ๐—ผ๐—ป For years my portfolio leaned toward โ€œvalueโ€ names. Only in 2025 did I add Adobe, Google, and Nvidia. While I missed their early rallies, I donโ€™t regret itโ€”my portfolio still outperforms the S&P500. With a new NVDA position (outside Etoro), I want to look closer at this company, the undisputed leader of the AI era. ๐—ฆ๐—ฒ๐—บ๐—ถ๐—ฐ๐—ผ๐—ป๐—ฑ๐˜‚๐—ฐ๐˜๐—ผ๐—ฟ๐˜€: ๐—™๐—ฟ๐—ผ๐—บ ๐—š๐—ฟ๐—ผ๐˜„๐˜๐—ต ๐˜๐—ผ ๐— ๐—ฎ๐˜๐˜‚๐—ฟ๐—ถ๐˜๐˜† The sector surged in the 1980sโ€“90s with PCs and the dot-com boom, then slowed as chips commoditized. Since 2020, cloud and AI have reignited demand. Margins remain strong, though cycles persist. Leadership shifted: Japan faded, TSMC and Samsung rose, Intel stumbled, Nvidia thrived by targeting high-value niches. ๐—”๐—œ ๐—–๐—ต๐—ถ๐—ฝ๐˜€: ๐—ช๐—ต๐˜† ๐—š๐—ฃ๐—จ๐˜€ ๐— ๐—ฎ๐˜๐˜๐—ฒ๐—ฟ A GPU contains billions of transistors, enabling trillions of parallel calculations per second. This makes them perfect for gaming, visualization, and training AI modelsโ€”tasks CPUs cannot handle efficiently. Nvidia dominates this segment, holding ~85% of the AI data center market in 2024. ๐—™๐—ฟ๐—ผ๐—บ ๐—ฆ๐˜๐—ฟ๐˜‚๐—ด๐—ด๐—น๐—ฒ๐˜€ ๐˜๐—ผ ๐—Ÿ๐—ฒ๐—ฎ๐—ฑ๐—ฒ๐—ฟ๐˜€๐—ต๐—ถ๐—ฝ Founded in 1993, Nvidia nearly failed when its NV1 chip misaligned with Microsoftโ€™s DirectX. With only weeks of cash left, Jensen Huang bet everything on Riva 128 (1997), designed entirely in softwareโ€”an industry first. It worked, and in 1999 the GeForce 256 launched the modern GPU era. Since then, Nvidia has supplied Microsoftโ€™s Xbox, led in gaming, and capitalized on crypto and now AI. Drawdowns have been brutal (up to โ€“89%), but the company always rebounded stronger. ๐——๐—ฟ๐—ถ๐˜ƒ๐—ฒ๐—ฟ๐˜€ ๐—ผ๐—ณ ๐—ฅ๐—ฒ๐˜๐˜‚๐—ฟ๐—ป๐˜€ Over the past decade, Nvidiaโ€™s revenue grew >40% CAGR, EPS ~65%, ROIC consistently high. This mix of growth + capital discipline delivered ~80% annualized returns. Few companies combine scale, margins, and innovation so effectively. ๐—ฉ๐—ฎ๐—น๐˜‚๐—ฎ๐˜๐—ถ๐—ผ๐—ป ๐—ฆ๐—ป๐—ฎ๐—ฝ๐˜€๐—ต๐—ผ๐˜ My base case: Revenue CAGR: 25% (2025โ€“2030), 20% (2030โ€“2035). Market share declines from 85% to ~55%. Terminal growth: 3.5%. 2035 revenues: ~$ 900โ€“950B. Operating margin: 58%. WACC: 11%. Reinvestment ratio: Sales-to-capital 1.8. Exit multiple: 18x EBITDA (vs. sector 24x). Fair Value: $ 191/share. At current prices, NVDA trades ~12% below FV, implying ~4% annual alpha if closed in 3 years. ๐—ž๐—ฒ๐˜† ๐—ฅ๐—ถ๐˜€๐—ธ๐˜€ Taiwan exposure: Heavy reliance on TSMC. Nvidia is diversifying (Samsung, U.S. fabs), but full independence is unlikely. AI hype risk: Demand may normalize faster than expected. Customer concentration: Two hyperscalers = 39% of revenue. Competition: AMD, Google TPU, Amazon Trainium, Alibabaโ€™s 2025 chip all aim to reduce dependence on Nvidia. Regulation: Export bans, tariffs, and geopolitics could disrupt supply. Macroeconomy: Recession, FX, rates. Expectations treadmill: Harder to grow at scaleโ€”the law of diminishing returns. ๐—–๐—ผ๐—ป๐—ฐ๐—น๐˜‚๐˜€๐—ถ๐—ผ๐—ป AI is a transformative trend but not every AI stock is a winner. Nvidia, however, is the backbone of AI infrastructure. Risks remain, but leadership, profitability, and growth justify its premium. In my view, NVDA is a buy for long-term investors willing to accept volatility. Full article available on Patreon: www.patreon.com/posts/nvda-valuation-138609575?utm_medium=clipboard_copy&utm_source=copyLink&utm_campaign=postshare_creator&utm_content=join_link
4 replies
null
.