Over the past year, a lot of people have experienced the same thing. A position they bought as a small addition to their portfolio has grown so much that it now decides their entire result. It's a nice problem to have, but I think it's harder to solve than the opposite one. With a losing position, I can at least tell myself the thesis and either it holds or it doesn't. With a winner that tripled in a year, I'm not wondering whether the company is good. I'm wondering whether I want a single stock to decide my year. I've dealt with this three times in the last six months, and each time I decided differently.
I sold my memory companies entirely. The profit came in a very short time, and mainly I didn't like how those stocks behave. Micron $MU is up from 114 to 938 dollars over the last twelve months, but at the same time it's about 25% below its annual high. SanDisk $SNDK is down by more than a third from its high. This year, both names have had several days with drops of six to ten percent, and Micron's beta is 2.2, meaning double the swings against the market.
What's strange is that based on valuation, it doesn't look overheated. Micron trades at a forward P/E of around 6.5, SanDisk around 7. But with a cyclical company, a low multiple at the peak of the earnings cycle is more of a warning than a discount. It all comes down to a single question: how long will the cycle last? Based on the capex the companies are projecting, it will last a while longer. The five largest American cloud and AI infrastructure providers plan to invest 660 to 690 billion dollars this year, roughly double last year, and Amazon alone is planning 200 billion. I probably sold before the cycle ends, and I consider it the price of peace of mind.
I kept ASML $ASML and only took some profit off the table. The position had grown too large for me, and I didn't want one company to decide my portfolio. But getting rid of it entirely didn't make sense to me, because ASML doesn't rely solely on machine sales. Service and upgrades of already installed machines brought in 2.76 billion euros out of total 9.3 billion in the second quarter, almost a third of revenue, and this money keeps coming even in years when customers don't order new machines.
With the big tech four, meaning $AMZN , $GOOG, $META and $MSFT , on the other hand, I've only been lightly adding the whole time and haven't sold anything.
What about you? Have you already taken some profits from AI positions, or are you letting them ride?
