Recall April of 2021. There was no such thing as ChatGPT. The term “large language model” was more appropriate for scholarly articles than casual conversations. Nvidia was undoubtedly a reputable chip manufacturer, but it wasn’t the most valuable company in human history. Quietly, Broadcom was making money. Analysts were genuinely unsure whether Palantir’s government-heavy revenue model could withstand a change in Washington’s priorities, and the company had just gone public, trading between intrigue and skepticism. When those three stocks were combined, no one thought, “This is the trade of the decade.”
And yet. Your $30,000 would be worth more than $260,000 today if you had invested $10,000 in each of those three businesses five years ago and just left them alone. Nvidia alone generated a return of almost 1,200%, turning that $10,000 into about $126,000. An additional $69,000 was added by Broadcom. Palantir would have turned that same $10,000 stake into about $65,000 even after declining 34.5% from its peak. Not because the math is difficult, but rather because it is so straightforward, it is nearly uncomfortable to look at. All you had to do was buy despite every headline telling you not to.
Nvidia, Broadcom & Palantir — Investment Return Snapshot (as of April 2026)
| Topic | AI Growth Stock Returns: 5-Year Portfolio Analysis |
| Nvidia (NVDA) — 5-Year Return | ~1,200% Top Performer $10,000 invested → ~$126,000 today |
| Nvidia Market Cap (Apr 2026) | ~$4.8 trillion — most valuable company in the world |
| Nvidia Current Stock Price | $200.68 (NASDAQ: NVDA) | Forward P/E: ~22 |
| Broadcom (AVGO) — 5-Year Return | ~590% | $10,000 invested → ~$69,000 today |
| Broadcom FY2025 Revenue | Just under $64 billion — nearly double its revenue from two years prior |
| Broadcom Market Cap | ~$1.6 trillion | Return on Equity (ROE): 47.5% |
| Palantir (PLTR) — 5-Year Return | ~560% | $10,000 invested → ~$65,000 today Valuation Risk |
| Palantir Q1 2026 Revenue | $1.41 billion — up 70% year-over-year; US commercial revenue up 137% YoY |
| Palantir Market Cap & Valuation | ~$316 billion | Price-to-Sales ratio: 68 (highest among large-cap tech) Expensive |
| Palantir Stock Drop from ATH | Down 34.5% from all-time highs (as of mid-April 2026) |
| Combined $30K Portfolio Value (5yr) | Over $260,000 today (Nvidia + Broadcom + Palantir) +767% blended |
| Best Mutual Fund Flow — Broadcom | $24.49 billion invested by top funds (largest single allocation reported) |
| Palantir Shares Outstanding Growth (5yr) | +28% — ongoing dilution risk for long-term holders Watch |
| Nvidia Revenue Growth (latest quarter) | 73% year-over-year — slowing but still historically elevated |
The most frequently reported story is that of Nvidia’s rise, and with good reason. Its graphics processing units, which were initially created to render video game environments, proved to be just what AI researchers needed to train massive models. The company’s own executives have publicly acknowledged that they were not prepared for the explosion in demand. In its most recent quarter, revenue growth reached 73%, which may seem remarkable, but keep in mind that previous quarters saw growth rates of 100% or higher.
The company is currently the most valuable publicly traded company in the world, with a market capitalization of approximately $4.8 trillion. Nvidia would need to reach a valuation of about $60 trillion in order to generate an additional 1,200% return from this base. That isn’t taking place. However, the stock isn’t particularly pricey either, with a forward price-to-earnings multiple of about 22. It’s just a much more common wager than it used to be, bearing the weight of extremely high expectations.
People may be surprised by Broadcom’s story because it is more subdued. While Nvidia dominated the news, Broadcom was collaborating closely with Amazon, Google, and Meta, the main cloud computing platforms, to help them create custom AI chips as a substitute for Nvidia’s escalatingly costly hardware. Its revenue for the fiscal year 2025 was slightly less than $64 billion, almost twice as much as it had a few years prior. Top mutual funds invested $24.49 billion in Broadcom stock in a single reporting period, the largest allocation shown on the screen, dwarfing even the $5.26 billion invested in Nvidia. It appears that the best institutional money in the market noticed before the majority of retail investors did. With a return on equity of 47.5%, Broadcom is demonstrating true operational efficiency rather than hype. Even though it doesn’t create the same level of excitement on social media, serious investors believe that Broadcom is the more resilient of the three AI infrastructure plays.
The story becomes truly complex at Palantir. There is no denying that the business is expanding. In its most recent quarter, revenue increased 70% year over year to $1.41 billion, while U.S. commercial revenue grew an incredible 137% during the same time frame. Its GAAP profit margin of 41% put an end to a long-running argument about the company’s viability. The backlog of contracts is increasing. The government connections, which were meticulously developed over more than 20 years through disputes, congressional hearings, and shifting administrations, are now producing $4.3 billion in total contract value, an increase of 138% from the previous year. Palantir is doing well by practically all operational metrics. Valuation is the issue. Its price-to-sales ratio of 68 is so much higher than that of any comparable large-cap tech company that it is hardly worth comparing. Around 36, Arm Holdings is the next closest. Purchasing Palantir now entails paying a remarkable premium for future growth, which must be not just good but outstanding for years to come.
It’s difficult to ignore the fact that the underlying question for all three of these stocks is the same: how long can the AI investment cycle maintain the growth rates assumed by current valuations? It has been truly fascinating to watch this develop over the past five years; this is the kind of market moment that is eventually covered in textbooks. However, the circumstances that led to those returns were peculiar. The market was still figuring out how to price it, the capital was inexpensive, and the AI narrative was novel. These three statements are no longer true. These businesses are outstanding. However, compared to the previous five years, the next five will call for a different kind of patience and likely a greater tolerance for the periods when the numbers move in the wrong direction.
