Nvidia Fell Out of the World's Top 10 Most Valuable Companies in 2022. By 2025 It Was #1.

Ten years ago, the world's most valuable companies were a mix of tech, oil, banking, and consumer staples. Today they're almost entirely tech — and the single biggest mover isn't even the company that's been #1 longest. It's Nvidia, which spent 2022 outside the top 10 entirely before becoming the most valuable company alive.

The last decade of the world's most valuable companies

↻ Replay2026

Apple

5T

Microsoft

4T

Alphabet

4T

Amazon

3T

Meta

2T

Nvidia

5T

Broadcom

2T

Berkshire Hathaway

1T

ExxonMobil

1T

JPMorgan Chase

1T

Wells Fargo

0T

Johnson & Johnson

1T

Walmart

1T

Tesla

1T

Eli Lilly

1T

UnitedHealth

0T

Visa

1T

Show the data

Topic	2016	2018	2020	2022	2024	2026
Apple	0.618	0.749	2.256	2.067	3.785	4.573
Microsoft	0.484	0.78	1.682	1.788	3.134	3.713
Alphabet	0.539	0.723	1.187	1.145	2.325	4.333
Amazon	0.356	0.734	1.634	0.857	2.307	2.961
Meta	0.332	0.374	0.778	0.32	1.478	1.508
Nvidia	0.058	0.081	0.323	0.364	3.289	5.425
Broadcom	0.075	0.104	0.178	0.234	1.087	2.035
Berkshire Hathaway	0.4	0.503	0.543	0.682	0.979	1.122
ExxonMobil	0.374	0.289	0.174	0.454	0.473	0.634
JPMorgan Chase	0.309	0.325	0.387	0.393	0.675	0.95
Wells Fargo	0.276	0.217	0.125	0.157	0.234	0.264
Johnson & Johnson	0.313	0.346	0.414	0.462	0.348	0.625
Walmart	0.212	0.271	0.408	0.382	0.726	0.89
Tesla	0.034	0.057	0.669	0.389	1.296	1.298
Eli Lilly	0.077	0.114	0.162	0.348	0.694	1.057
UnitedHealth	0.152	0.24	0.333	0.495	0.466	0.37
Visa	0.164	0.291	0.482	0.44	0.61	0.677

Nvidia's round trip: in, out, and then #1

Nvidia's story isn't a straight climb — it's a crash landing followed by liftoff. At the end of 2021, Nvidia cracked the world's top 10 for the first time, worth $732.9 billion at #6. Then came 2022's broad tech selloff: Nvidia's value fell 50%, to $364.2 billion, just short of that year's #10 cutoff ($393.3 billion, JPMorgan Chase) — knocked back out of the top 10 entirely, one year after finally arriving.

From that 2022 low, Nvidia's value multiplied roughly 15x in under four years — driven by the AI infrastructure buildout, where its GPUs became the scarce resource every major AI lab and cloud provider needed. No other company in this dataset moved from outside the top 10 to #1 in the same window; Broadcom, riding a similar AI-chip demand wave, made the same journey from outside the top 10 to #6 by 2026 ($2.04 trillion), but never overtook Nvidia.

The 2016 top 10 is half gone

Survived to 2026 top 106 of 10

Dropped out by 20264 of 10

Notice something about that dropped-out list: three of the four (ExxonMobil, Johnson & Johnson, JPMorgan Chase) actually grew in dollar terms over the decade — JPMorgan more than tripled, from $308.7B to $950.4B — and still fell out, because the #10 cutoff itself rose even faster, 3.8x versus JPMorgan's 3.1x. Only Wells Fargo actually shrank in absolute terms (down about 5%, to $263.8B) — the other three didn't lose the race by falling behind themselves, they lost it by not growing as fast as the bar in front of them. In 2016 you needed $276.4 billion to hold #10. In 2026 you need over a trillion.

Tech ate the list

201658.2%

201865.7%

202075.0%

202266.7%

202485.3%

202687.6%

Five of 2016's top 10 companies were tech (Apple, Alphabet, Microsoft, Amazon, and Meta); by 2026, tech holds seven of the ten seats (those five plus Nvidia and Broadcom). But the share of value concentrated in tech nearly doubled, from 58.2% to 87.6% of the combined top-10 total — even the dip in 2022, when the broader tech selloff briefly gave the "old guard" (ExxonMobil, JPMorgan Chase, UnitedHealth, Johnson & Johnson) a relatively larger slice, recovered and kept climbing. By 2026, the ten most valuable companies on Earth are worth $28.02 trillion combined, and $24.55 trillion of that — seven dollars in eight — sits in tech.

What this shows, and what it doesn't

What holds up: the world's most valuable companies became dramatically more tech-concentrated over the last decade, real churn happened even among giants (4 of 2016's top 10 are gone), and Nvidia's 2022-to-2026 run — falling out of the top 10, then becoming #1 — is a genuinely unusual trajectory with no close parallel in this dataset.

What limits it:

• This tracks a fixed candidate list of 17 companies, chosen because independently published market-cap history shows each one held a global top-10 seat at some point since 2016 — it is not a live, unbounded search across every public company on Earth for every checkpoint. A company briefly cracking the top 10 in a year we didn't check (we sampled 2016, 2018, 2020, 2022, 2024, 2026) wouldn't be caught by this method.
• Meta/Facebook's historical share-count data has a gap: our live source (Yahoo Finance) only carries Meta's share count back to mid-2022. For 2016, 2018, and 2020, we used independently published market-cap figures instead of computing them ourselves — the same self-computed method we used for every other company and checkpoint, cross-checked against those published figures, matched to within a fraction of a percent at every point we could verify, which is why we trust the substitution.
• This measures publicly traded companies with a USD market cap on major exchanges — it excludes Saudi Aramco, which briefly held the #1 spot around its 2019 IPO on the Saudi exchange, and other non-US-exchange-listed giants, simply because they fall outside our data source's coverage.
• Market cap is a snapshot, not a measure of profit, revenue, or economic footprint — it reflects what investors are willing to pay for future earnings, which for AI-era companies like Nvidia has moved faster than any underlying business fundamental could on its own.

Sources

Methodology

For each of 17 candidate companies — chosen from independently published top-10 market-cap history as ones that held a global top-10 seat at some point since 2016 — we pulled Yahoo Finance's historical share-count series and daily close-price history back to 2016. Market cap at each of 6 checkpoints (year-end 2016/2018/2020/2022/2024, plus a live pull for 2026) is share count × close price at or nearest before that date.

One correction mattered: Yahoo's close price is split-adjusted (restated to the current share-count basis) even without requesting price adjustment, but the historical share-count series is a true point-in-time raw count that jumps at actual stock splits. Multiplying them directly undercounts every date before a company's most recent split by exactly the split ratio — we caught this because Apple's 2016-2019 figures came out roughly 4x too low against published history, matching Apple's August 2020 4-for-1 split exactly. The fix: pull each company's split events (date and ratio) and scale every historical share-count reading up by the product of every split that happened after that date, so it matches the price series' already-adjusted convention. We validated the corrected figures against independently published market-cap history (linked above) at every checkpoint we could cross-check — Apple's 2022 figure ($2.067T) and Microsoft's 2018 figure ($0.780T) matched the published record almost exactly.

Alphabet uses the GOOGL ticker alone, not GOOGL+GOOG combined — Yahoo's historical share-count series already reports Alphabet's total combined share count under either ticker (confirmed identical), so summing both would double-count. Berkshire Hathaway uses the BRK-A ticker, whose own market-cap figure already reflects the full company in BRK-A-equivalent share terms.

Want to run your own cut? Crawlora's Yahoo Finance endpoints cover the historical price and share-count data used here, and pricing has the free tier.