🔥Among the list of 'biggest AI winners,' half are already overvalued — the real opportunity might not be where you first look
The market is currently telling a very simple story:
“Whoever grows the fastest should rise the most”
But my own judgment is just the opposite —
The fastest-growing names are often also the ones whose future expectations are most easily priced in
If we break down this list, it can actually be divided into three layers of logic:
First layer: Core leaders that have already been “priced” by the market
$NVDA
$AVGO
This layer is characterized by:
The clearest logic
The most concentrated capital
The narrative is the most consistent
The problem is also the most obvious:
The market already knows they will grow
So the question is not 'whether or not they will grow,' but rather:
Can they still exceed expectations
The second layer: 'High elasticity components' in AI infrastructure
$MU
$SNDK
$SKH
The essence of this layer is:
Storage = the 'complementary bottleneck' of AI computing power
The current growth comes from:
HBM
Data center demand
The scale of AI training is expanding
But there's a risk that I've been closely monitoring:
This is a highly cyclical industry
Once supply starts to catch up,
Profit growth may not continue linearly
The third layer: the most easily overlooked, yet the most volatile part of the chain
$AAOI
$LITE
$CRDO
$ALAB
$SIMO
This layer is what I think is most worth unpacking:
They aren't 'telling AI stories',
but rather 'occupying a key position in the AI supply chain'
For example:
Light...