1. I'm WAY outside my area of expertise on this, so I may be wrong about all of it.
2. Good point about April, although that was before we realized TACO.
3. I have a very strong hunch the market doesn't give a shit about the social and political entropy that's coming.
1. Trying to predict or explain the short-term market movers is a fool's errand. If it was predictable or explicable, someone would have put that predictability in their pocket and then it's not there any more for us to ponder.
2. It's especially a fool's errand when it comes to the United States at this point in history. That's because of the weird and counter-intuitive possibility that US assets will increase in market value even as their asset quality degrades. I've mentioned the multiple equilibrium problem before: let's assume that the US will default on the debt by December. Which way should the market move?
Well, the more natural intuition would be that the market will decline, because duh. And that's still a strong possibility. But it's also true that in panics, people flee to "safer assets" which, historically speaking, have been US Treasuries. Most trading is algorithmic based on historical and modeled dependencies. Those historical tendencies show that in times of distress, the right move is to sell your Vietnamese assets and buy US safe assets. This poses a huge dilemma for an investor, because the US collapse would cause a global panic. So as an investor, you have to think, "what will everyone else do in response to this news?" That's definitely what the trading algorithms are doing.
OK, so how do you price the US assets. If you think that most people are going to use historical correlations, then US assets would increase in value during the disruption and thus you should buy those assets, not sell. We're now deep in game theory. Every single person could think the assets are valueless, but still buy them on the expectation that others will find the assets valuable. This is what used to happen with corporate takeovers, where tender offers could be successful at prices that every single investor thought too low -- and it's why the front-loaded tender offer was made illegal by the Williams Act. On the other hand, it's possible that people will recognize that even US Treasury bonds are shit assets if there's been a default, in which case you'd expect the market to decline. So what to do if you expect a collapse? Ah, tricky question.
Keep in mind also that there has never been a brisk trade in US-Treasury CDS, even though you'd think that would be one of the best securities to insure in that way. Well, I shouldn't say never because maybe it's a market that has come to life when I'm not paying attention, but as of five or ten years ago it wasn't. Why? Because it was thought to be like nuclear war insurance; you're never going to get your payout because the event will wipe out the ability of your counterparty to perform. Well, maybe we're seeing the flip side of that. Why worry about a US economic collapse? If it happens, we're all dead (figuratively speaking) anyway.
3. Remember what happened in the pandemic. The earnings of big companies shot up during and after. It's because they survived while many small businesses went bust. This happened in 08 also. Well, the earnings of small businesses are not publicly traded. If the retail sector suffers a downturn and declines by 10% in value overall, retail stocks could still gain because the big firms could pull all the customers who used to shop at local small business.
So a key variable in the equation is "how much of the economy will be publicly traded" and if that number goes up, the publicly traded sector can get more valuable even if the overall economy declines.
4. So I don't think it's all about or even primarily about AI. I very much doubt there are many people trading big $$ on the potential prospect of huge AI deployment that throws people out of work. The optimists say it will happen within two years (which is still an eternity in today's hyper short term market), but really? And even if it does, will the early adopters benefit from it? It might take a while to get right. Often, when technology is introduced, there's an edge case that was unanticipated and ultimately dooms the whole effort. Long Term Capital Management had an "arbitrage" strategy that worked super-well for years. But it didn't account for political actors, and thus when the Russian Duma did something LTCM did not expect (I don't remember the trade exactly, but I think it had to do with currency controls), the careful arbitrage it had set up blew up in its face and it lost money on both sides of its trade. It had to be bailed out. Same will likely happen with AI.
gizmodo.com
