Equity Research · Realized Position
AMD
AMD · Semiconductors · Compute
Bought the AI second-source when the market doubted MI300; sold at $180 when it believed.
Quan Pham · M’Squared Capital Research · August 2026
What the market is missing
At $78 the market was pricing AMD as if hyperscalers would never want an alternative to a single accelerator vendor. I thought MI300 gave them a credible one, and that the optionality of a real second source was simply not in the price.
Why I owned it
The base case did not even need the accelerators to work. EPYC share gains in the data center funded the AI investment and underpinned the stock on their own; MI300 was the free option on top. I was buying skepticism and planning to sell belief.
What has to happen
It needed the MI300 ramp and cloud design wins to become visible, continued data-center CPU share gains, and the sell-side to keep upgrading the AI TAM.
What would prove me wrong
The standing risk was NVIDIA's software and ecosystem moat proving too deep for a second source to matter, plus ordinary semiconductor cyclicality and execution on the software stack.
Valuation & exit
I exited at $180 — a realized 130.8% — as sentiment across the whole complex turned euphoric and the risk-reward inverted. The entire trade was the sentiment swing from doubt to belief, and once it had swung I had no reason to stay.
Position
Long
Realized return
+130.8%
Entry → Exit
$78.00 → $180.00
Initiated
10 Apr 2025
Sector
Semiconductors
Status
Closed at target
Last updated
August 2026
This note is the opinion of M’Squared Capital and is provided for informational purposes only. It is not investment advice, nor an offer or solicitation to buy or sell any security. Price targets and entry and exit levels reflect the firm’s proprietary views and realized transactions; they are not forecasts of future results. The firm may hold, add to, reduce, or close any position at any time without notice. Past performance is not indicative of future results.