AI chip testing is shaping up as the next bottleneck in the supply chain.
Nikkei reports an AI chip can take more than 10 minutes to test, versus under a minute for a phone processor, and many now require 100% testing across multiple stages.
That is driving a sharp revenue ramp for probing and test names like FormFactor, Micronics Japan, JEM, WinWay, and CHPT.
Trade idea that I published to my shower thoughts channel:
Korean Index volatility arbitrage and taking advantage of Black-Scholes models.
$EWY long options seem mispriced.
This is Blackrock's Korea Index, which is majority memory (Samsung Electronics, Sk Hynix).
The stock swings 2-5+% a day, and is up 136.25% 1Y, despite priced like a normal index IV.
Samsung is volatile. SK Hynix is volatile (eg. 65% - 80% est).
But the combination of the two through the index is priced way less than both low beta $GOOGL (37.33%) and $AMZN (39.12%) at ~32% IV.
I've been watching $EWY for a bit and it does look volatile.
As for pricing my guess is MMs priced in IV based on historical averages (5-10 years), where the Korean index was completely flat. And were expecting calls 2 years out to revert to the mean.
But this volatility should be the new norm as markets price in the new memory supercycle (eg. $TSM went from 30% IV to 46.2% IV).
Long calls should benefit from both Samsung + Sk Hynix carrying the index.
And the main benefit is vega expansion that you won't get from $KORU.
You also can't get this option MM pinning like individual US stocks since this is Korea's national index and long term.
TLDR: Individual components SK Hynix + Samsung are highly volatile.
They're basically half of the index, but options in index are priced with low volatility, perhaps due to historical 5-10 year data.
Long calls benefit from vega expansion that weren't priced in correctly as MM forward vol estimates are anchored too heavily on historical realized vol, which was low for $EWY over the past 5-10 years