The longest-duration assets in the market should be the most sensitive to interest rates.
Instead they are the least.
Why? Either the marginal AI investor does not care about discounting cash flows but only about being blessed at the Ω-point by the coming Digital God, or it is the AI-driven datacenter buildout’s capital requirements and pretty much only the AI-driven datacenter buildout’s capital requirements that is driving rates up right now:
<https://www.ft.com/content/6d06a6b2-ea2f-4384-959d-682ce78a1693?syn-25a6b1a6=1> <https://www.ft.com/unhedged-podcast>Katie Martin: Stocks right now are not really what they seem. You look at the big indices, especially in the US, and the first thing that strikes you is how upbeat and stable they are. It’s quite a sight actually, given the long war in Iran and the endless horror show that is the government bond market. The thing is, all that stability comes from a tiny number of companies. The top 20 stocks in the US are making up for a pretty poor performance in the other 480 in the benchmark S&P 500 index. Three quarters of stocks in that index fell in September. [MUSIC PLAYING] We’re all used to markets being a bit skinny but it does feel like they’re overdoing the fat jabs….
Rob Armstrong: [The] stocks [that] are keeping the market as a whole up… [are] in order of value contribution: Apple, Meta, Nvidia, Microsoft, Advanced Micro Devices, Tesla, Micron, Intel…. That’s since the middle of August. So it’s just this slice of the market that is keeping the whole circus going….
The market is absorbing a rates shock and acting like one would expect…except for those AI names which are invulnerable…. Real estate and utilities…. are… bond substitutes because… [they] tend to pay a high yield…. Number of real estate companies in the S&P 500 that are up since the middle of August: zero. Number of utilities companies… one…. These two sectors are absolutely getting whacked….
Katie Martin: Does the rest of the market catch up with the AI stocks, or do the AI stocks sink back down to where the rest of the market is, or do we carry on with… a two-speed stock market… indefinitely?…
Rob Armstrong: The harder half of your question is…: Can the fun continue with the AI stocks?… Enthusiasm for these names seems undimmed. You don’t see much pressure around the edges…. No one seems to be hedging this trade….
Katie Martin: The drain that everything is circling around… is everything is contingent on… AI…. Te value of the dollar is contingent on it and the entire global stock market is contingent on this thing…. The whole corporate bond market is… a play on AI now…. Everything is computer and I don’t like it…. It’s weird. I think that’s the only conclusion… [that] I don’t know what it is exactly but it’s weird.
Rob Armstrong: I don’t think it’s that opaque actually. It is weird that the AI stocks don’t care. Like the honey badger, AI stocks don’t care.
Katie Martin: Honey badger don’t care….
Robert Armstrong: That’s a bit weird, but what is happening in the rest of the market is the kind of adjustment one would expect…. A strong economy but much higher rates, very rapid rise in rates, you would generally expect this kind of response….
**Katie Martin: **Listeners, choose your fighter, AI stocks or everything else…
- Brad DeLong here*: On the one hand, it really is weird.
“AI” stocks are the speculativist of speculative and the longest duration of long-duration assets. The proportional response of their prices to changes in interest rates should therefore be by far the highest. Interest rates go up. Their prices should go down, proportionately, the most. And yet it hasn’t happened.
Now you can say, with Katie Martin here, “honey badger don’t care”:
More mathematically, the valuation of stocks is supposed to be given by their fundamental expected present value: the expected value of the current cash-flow payout to shareholders divided by the difference between the required rate of return and the growth rate of the cash-flow payout:
But the valuation of AI stocks is given by something else: a fraction 𝝺 times the fundamental value plus a fraction (1-𝝺) times something else, call it Ω, the omega-point value at the near-singularity (for the economy at least) where ALL IS UNVEILED:
But, still, that (1-𝝺)Ω is something that is supposed to happen in the future, say T years hence, and so why isn’t it discounted by a factor that has more bite when the interest rate r rises?
What is it that makes it such that, when r rises, then the Ω rises in lockstep with it?
Maybe the marginal investor in the AI-component of the S&P500 right now really is Honey Badger. Maybe the effect of interest rate-induced discounts on present values is the last thing from their minds. Maybe they are fixated on the belief that if they invest in AI they will be rewarded, or at least not turned into paper clips, by the coming Digital God:
But there is a second, possibility, an alternative. The alternative is not rꜛ → Ωꜛ, but rather Ωꜛ→ rꜛ: that * the upward shock to the interest rate is driven by AI and by AI alone. *That neither inflation nor Federal Reserve communications and policy fumbles nor “geopolitical risks” (i.e., that the U.S. Republican Party backed and backs and that the American people voted for a catastrophic chaos-monkey moron) nor the satiation of demand for “safe” liquid collateralizable assets is playing (much of) a role in the current rise in interest rates, but only (or substantially) the investment requirements of the great datacenter buildout induced by the animal spirits shared by investors and by hyperscaler executives.
I am not sure what I think of this second possibility. “Honey Badger don’t care” seems preferred by Occam’s Razor.
##macro-outlook
##podcast-of-the-day
#katie-martin-rob-armstrong-there-is-only-one-trade
#katie-martin
#rob-armstrong
#there-is-only-one-trade
#ai-stocks #interest-rates-shock
#honey-badger
#unhedged
#financial-times
#stock-market
#market-concentratio
#two-speed-market
#present-value
#long-duration-assets
#ai-bubble
#datacenter-buildout
#hyperscalers
#animal-spirits
