In May, almost everything we published at VMF Research revolved around one uncomfortable question: What should investors do when the market has already discovered the obvious winners of a great theme?
In this case, the theme was AI.
The first stage of the trade has been increasingly well understood: semiconductors, memory, data centers, power, grid infrastructure, cooling systems, and the physical capacity required to sustain the AI buildout. These areas still matter enormously, but the market is no longer ignoring them. In many cases, it is already paying a very visible price for them.
So May became less about asking whether AI is real, and more about asking where the mispricing may still live.
That trail took us from the visible AI bottlenecks to the forgotten application layer; from broad China internet exposure to PDD as a direct stock-level expression; and finally to Alpha Tier’s “Tulip Test”, where we asked whether the first AI trade had become crowded enough to justify a different risk budget.
To make that progression easier to follow, we just published a new “Inside VMF Research” piece.
Think of it as a reading guide to our May work.
It connects the public Substack excerpts we released during the month (from “The Visible Side of the AI Boom” and “The Forgotten Application Layer” to “PDD: China Price, Temu Optionality” and “A Pattern Recognition Exercise”) and shows how they fit into one coherent investment process.
For free readers, it is the best place to understand the arc of our thinking.
For paid subscribers, May included the full implementation behind that arc: vehicle selection, position sizing, valuation work, portfolio changes, liquidity adjustments, and the monitoring framework we will use from here.
May was not a retreat from AI.
It was a rotation inside AI... away from what the market can now see clearly, and toward the areas where judgment still matters most.
marketview.vmfresearch.com/p…