What does AI mean for how the stock market works?
That’s not the top question on your mind today, I know. You’re checking election returns. Me too. I wrote this yesterday.
In a week in New York, I’ll join a panel at IR Impact (hope to see you there!) on how AI is reshaping IR. For those of you unfamiliar, that stands not for “infrared” but Investor Relations, liaison to Wall Street at public companies.
Maybe we should rethink that definition of IR. Already, two-thirds of assets are managed by models. A growing number of those are powered by AI. Algorithms are run by AI. Order types at exchanges are managed by AI. Data are gleaned and compiled by AI.
There’s a lot less of “liaison” in the IR job today than there was 25 years ago. And a whole lot more strategy.
Take quarterly Board reports. As a central tendency, they’ll include strategic priorities and progress or setbacks, something on the macro environment, pro forma financial performance, probably sellside statistics, holder data, an operational update, likely a legal or governance update, something on cybersecurity and risk management and how AI is eating tokens prodigiously, and something about human resources.
Generalizing.
Unless you’re a ModernIR client there’s nothing on market structure, though the entire market is in a full-on reform convulsion. We’ve got tokens, 24-hour trading, AI running algorithms and data-gathering and order types, an utter rethink of Reg NMS pillars, machines dominating, Passives as sole net buyer.
Anything in there about that?
I would submit that it’s hard to run a public company without metering those factors because they are central forces behind shareholder value.
Passives don’t buy financial performance per se. It’s why they concentrate holdings in the S&P 500 and subsets of it. About 150 stocks are currently – depending on what happens any given moment – 75% of market cap. The S&P 500 is about 90%. Russell 1000, about 97%.
Owning SPX components is fundamental in the sense that it’s a quantitative screen for quality. Rather than assessing individual performance, investors take a pre-screened set.
You might wonder why, if you’re in the SPX, that you don’t perform like it. Because Passives buy different metrics than Actives do. They buy market quality, liquidity, stability. We meter those with our new Market Structure Scorecard, which assigns a score to each, versus peers, industry, sector, market-cap and so on. It’ll be official next week at IR Impact.
You can thus see your health versus something else’s. Or literally everything else. It’s part of our overall Proactive Passive Strategy. Appeal to the net buyers and keep them.
And speaking of net buyers, I’m going to posit a thesis. Managing risk is a Wall Street obsession. Paul Singer’s team at Elliott Management this week offered the notion that investors might swing from Fear of Missing Out, to Fear of Getting Obliterated. FOMO to FOGO.
I don’t know. I’ve said the AI bull trade ended in June. You don’t know it because it’s obscured by layers of derivatives, intermediation, leverage. I may be wrong.
But here’s the thesis: A large risk management system is failing.
Vast seas of money have some kind of now no doubt AI-powered overlay meant to assess information and recommend action. Maybe even independently take action. Aladdin from Blackrock is among those, covering some $25 trillion, though I know nothing about AI integration there.
We have model portfolios too. They’re quantitative. Sometimes they say “dump these stocks and buy those.” They’re models. They’re intended to adapt to evolving conditions.
Since October monthly options resets, we have seen a start-and-stop quality in some market patterns. Money seemed to be rushing to the Safety Trade, like Consumer Staples, Healthcare. Then it reversed, and then returned, in two days.
Most times, Patterns proceed in measured fashion. You can’t hide a squadron of bombers in a starling murmuration, tiptoe an elephant across a putting green, if you’re juking and jiving like a jitterbug.
Sudden destabilization of Patterns to me suggests something didn’t go as planned. Maybe it was AI? Wholly speculative.
What’s certain is the stock market viewed through a Demand/Supply lens is nowhere near mean-reversion. A market built on Passive flows will have a tendency to return to the middle. Stocks are trading roughly where they were June 1.
A long period of returning to the middle isn’t “consolidation,” a market finding footing. It’s the impact of packing the marketplace with ways to leverage returns. When stocks don’t consistently rise, leverage eats away at returns until it eats a hole someplace.
We’ll all find out today and in coming days how midterm elections are weighted in risk-management schemes.
Am I saying the market has mean-reverted to June? No, I mean that Demand and Supply susurrate or move in murmurations continuously in predictable patterns. Despite trading at June levels, Broad Sentiment, or Demand, is way over the red line even after big losses in the AI trade.
It’s not bottomed. It’s topped. It needs to come down from 6.4 on a 10-point scale, to about 4.8 to mean-revert. That’s a big move. If it accelerates, it’ll blow right through 4.8, headed down.
