I spend hours this weekend crunching numbers for AI various AI related scenarios.
Long story short:
Yes, AI is the tail that wags the economy dog. But no, there is no 2007-like risk.
Yes, data center power is the key bottleneck. But no, that is accounted for.
Yes, 2nd tier AI plays (bundle ORCL into the 2nd tier) are prone to market stress. No, first tier hyperscalers are doing just fine. You can short one group and go long the other if you are really itching to short something.
Yes, there are some cracks in the system, especially in private equity. But no, again this is not going to be like RE bubble, at least not yet, and Fed officials called the member firms this weekend and told them to use the RIPO facilities when they need to.
So the overall message I got after a ton of number crunching was that it's going to be bumpy, but it is too premature to predict sky is falling and there is in fact a 93% chance that if you buy the dips, 6 months to a year later you will be happy. |
