_{Brought to you by:}

“Technology happens because it’s possible.”

— Sam Altman

Friday cryptic charts

Like Kremlinologists parsing inscrutable Soviet pronouncements for coded messages about the Cold War, the macro community spent the week decoding Nick Timiraos articles and Fed governor speeches for coded messages about monetary policy.

Chair Powell feels obligated to play this game of economic charades because the market treats the FOMC’s interest rate decisions as if they're as serious as nuclear war.

Not only does the fate of the world economy seem to hang in the balance on whether next week’s rate cut is 25 or 50bps, but it seems to hang as well on how well the Fed drops hints about it.

Getting the market’s expectations just right appears to be as important to the Fed as getting the rate decision right.

Spoiler alert: Neither of these things matter very much.

How could they?

Has any factory ever not been built at 5% fed funds that would have been built at 4.75% fed funds?

Probably not.

Even less likely is a factory not being built because the Fed failed to sufficiently signal a decision that was just a few days away.

Instead, the fate of the US economy will be decided by countless billions of decisions made by tens of millions of businesses and hundreds of millions of consumers — few of whom could tell you what a fed funds rate even is.

The market cares about fed funds, of course — but even there, probably not for long.

The Nasdaq rocketed 5% higher this week and that might be because the Fed was cryptically guiding us to expect a generous 50bps cut next week rather than the parsimonious 25bps one it had been expecting. 

Or, more likely, it might be because we learned this week that AI can think.

OpenAI’s new chatbot, o1, is trained to “think hard about problems before answering” them and the “the longer it thinks, the better it does on reasoning tests.”

Large-language models, which even AI researchers can only understand in terms of statistics and probabilities, have now magically learned to both reason and think things through, like humans.

They will rapidly get better at it: “o1 thinks for seconds,” an OpenAI researcher explained, “but we aim for future versions to think for hours, days, even weeks.” 

But they already think pretty good: Per the graphic at the top, o1 beats PhD-level scientists at PhD-level science questions. 

That makes its release this week “the beginning of a new paradigm,” according to OpenAI’s Sam Altman, because it demonstrates “AI that can do general-purpose, complex reasoning.”

Call me crazy, but that seems like a more significant development for humanity than whether the Fed goes for 25 or 50bps next week.

It’s probably a more significant development for markets too, because if anything is going to keep the US economy from a hard landing recession, it’s more likely to be AI than interest rates.

Oracle noted this week, for example, that it’s doubling capex because of booming demand for AI, which they expect to continue “for years” — and Nvidia said this week that the return on that capex is “instant” because new GPUs, 20x faster than the old ones, offer 10x savings on compute.

That’s a lot of investment and a lot of productivity and it will probably be happening for a long time.

Can booming AI keep the US economy out of recession?

Even o1 won’t be able to decode that riddle.

So let’s check the charts.

Just getting going?

Two-year Treasury yields… 

…hit a two-year low today.

Input prices are falling: 

Several big commodities — WTI, gasoline, nickel, natural gas, corn — are down 20% or more over the past year.

Most people’s inflation metric:

Gas at the pump is below $3 in much of the country. 

The other popular metric:

30-year mortgage rates are down to 6.20%, near a two-year low. Even better, as a percentage of the value of the property that backs them, US mortgage debt is near a 70-year low.

Unlikely to ever see another low:

The federal government’s annualized interest expense is now over $1 trillion.

They like us anyway:

Despite the endless supply of US-government debt on offer, foreign demand for US credit (as measured by corporate bonds, in this chart from Apollo) is near all-time highs.

History in the making:

When the Fed cuts rates next week, it will break its longest ever streak of leaving fed funds at a peak.

Too long?

If Truflation’s “real-time” measure of inflation is at least directionally correct, even a 50bps cut next week might be too little, too late.

Assuming it matters — which it probably doesn’t.

Have a great weekend, code-breakers.

— Byron Gilliam

Catch your favorite newsletter author (Byron, of course!) IRL at Permissionless as he draws insights out of the greatest investment minds in the liquid token market. 

The Lack of Bold Bets in Crypto

Jack and Mert discuss the state of SocialFi and what it means for crypto to succeed. Tune in for their thoughts on how proof-of-stake systems should work, how sandwich attacks can be reduced onchain and the exciting DePIN projects coming to market. 

Watch or listen to Lightspeed on YouTube, Spotify or Apple.

Blockworks Research is conducting a survey to gain insight into the institutional staking landscape. This data will help industry leaders adopt their strategies as the industry matures.

If you're an institutional staker, we want to hear from you (and if you’re new to Blockworks Research, get 20% off of our service while you’re at it!)