2026 first half ends, did you beat the index?

Today is the last trading day of the first half, so let's do a routine review.
If you asked anyone at the beginning of the year which AI stocks to buy in the first half of '26 to make money, the answer would most likely be names like NVIDIA, Microsoft, and Google.
After all, the past three years have been led by the Magnificent Seven, so buying them seemed like a sure bet.
But looking at the data at the end of the half, the S&P 500 is up 8.5%, and the NASDAQ is up 11.1%. That sounds good.
Now look at the mid-cap index: the S&P 400 is up 14%, leaving the large-cap index in the dust. The real outperformers in the first half of this year weren't the trillion-dollar giants, but much smaller mid-cap companies.
For most ordinary people, they instinctively think AI means buying the biggest and strongest—NVIDIA. But the biggest isn't necessarily the one that rises the most.
Doubling NVIDIA's market cap now would require an astronomical amount of capital; even if the entire global capital market pulled together, it might not be enough.
But those companies in its supply chain with market caps of just tens of billions? Doubling only needs one or two quarters of earnings beats.
Both are feasting on AI's dividends, but the elasticity of an elephant dancing vs. a rabbit hopping is on completely different scales. That's why I recommended building positions in Marvell and DRAM a long time ago.
It's not that I think NVIDIA is bad—it's that I know NVIDIA's growth will cascade down to the smaller companies in its supply chain, and those companies have much greater elasticity than NVIDIA itself.
Every time NVIDIA sells a system, Marvell gets another DSP order—yet Marvell's market cap is just a fraction of NVIDIA's.