Today, Besi shareholders didn't get the earnings report they were waiting for, but were taken out by a single sentence instead:

HBM5 may also not be in a hurry to use hybrid bonding.

He was still calculating how many hybrid bonding machines would be needed for 20-layer HBM last night. When he opened the valuation model this morning, he silently pushed the 2029 profits back by two years.

The company hasn't lost orders, and AI hasn't stopped developing.

It's just that everyone suddenly realized: the new technology that was thought to be "absolutely necessary" might be able to wait a bit longer.

This is the cruelest part of tech stocks.

A technology doesn't need to be obsolete; it only needs to be realized two years later to be enough to bring down the stock price that had already discounted a decade's worth of expectations.

So the real risk for Besi has never been whether hybrid bonding has a future.

It's how much the market has already paid for this future.

$Bloom Energy(BE.US) $Taiwan Semiconductor(TSM.US) $Applied Materials(AMAT.US) 

Besi's recent decline is precisely because the market is worried about Samsung and SK Hynix delaying the adoption of hybrid bonding; this conflicts with Besi's previous expectation that high-layer HBM would drive the adoption of this technology.