Friday charts: Crazy train investing

This is a segment from The Breakdown newsletter. To read more editions, subscribe.

Stock markets made new highs again this week, but the mood is hardly celebratory — possibly because it feels like investing is about to get much harder.

How do you invest for the long term when 1) AI might change everything and 2) the US government is $38 trillion in debt (and counting)?

This week, Jim Cramer said the ballooning national debt made him “worried for his kids” — so much so that he took the unthinkable step of recommending bitcoin as a hedge against monetary debasement.

Quite the reversal for someone who said as recently as 2022 that he “would not touch crypto in a million years.”

(Note: 2022 was three years ago.)

I get it, though.

Watching yesterday’s surreal exchange between President Trump and Fed Chair Powell, both in hard hats, made me despair for the long-term health of my investment portfolio. 

What investments can you rest easy with when the world’s only risk-free asset — US Treasurys — no longer seem free of risk?

You could buy gold, of course, and that has been working lately: The original inflation hedge is up 40% over the past year.

But an inflation hedge is all gold usually is: Over the last 2,000 years, its real, inflation-adjusted return is approximately 0%.

That will maintain your purchasing power, which is nice, but it won’t help you retire.

Also, you don’t want to be hiding in gold when they start mining it from asteroids or creating it in fusion reactors.

Is bitcoin a better hiding place then?

That, too, has been working: The newest inflation hedge is up 80% over the past year, cementing its reputation as digital gold.

But it’s hard to have a 50-year view on such a new asset.

On that kind of timeline, people as smart as Cliff Asness and Matt Levine can think that “most of the probability” for bitcoin is “zero.”

Finding investments where most of the probability is not zero is not as easy as it used to be.

Land has been a reliable safe haven over the centuries — they’re not making more of it, as they say.

But in an age of demographic collapse, we might need less of it.

It’s estimated that by the year 2038, one-third of the homes in Japan will be abandoned for lack of anyone to live in them.

I’m not sure land will hold its value when the entire developed world looks like Japan a decade or two from now.

Still, if we’re headed toward monetary debasement in this decade, a portfolio of gold, bitcoin and land is probably a good bet.

But what if this decade turns out to be surprisingly good?

If the real story over the next 10 years or so turns out to be an AI-driven boom in productivity, a portfolio of gold, bitcoin and land might doom you to decades of underperformance (and declining purchasing power).

That is what the researchers at Epoch AI expect: They estimate that “even when AIs have only automated 30% of all economically useful tasks, economic growth rates may exceed 20%.”

20%!!!

It’s hard to fathom a world of 20% GDP growth — the US government might even manage to balance its budget in that scenario. 

But even that unthinkable outcome may be understating the upside.

The Economist noted this week that growth theorists who attempt to model the progress of technology “have long posited that if ideas beget more ideas with sufficient velocity, growth should increase without limit.” (Emphasis added.)

Sam Altman seems to think that’s where we’re headed: He believes AIs will be capable of begetting ideas as soon as next year, and that “2027 may see the arrival of robots that can do tasks in the real world.”

That would put markets on a crazy train to who knows where.

If investors sense that labor will soon be automated out of existence, there will be a mad scramble to secure the largest possible share of the post-labor economy.

“What should you do if you think an explosion in economic growth is coming?” The Economist asks. “The advice that leaps out from the models is simple: own capital, the returns to which are going to skyrocket.”

But not just any capital.

20% growth would likely mean 20% interest rates too, which would destroy the value of your gold, bitcoin, land and any stocks not directly involved in the AI productivity boom.

You’ll have to choose carefully, and maybe sooner than you’d expect.

It’s exhausting to think about. And maybe we shouldn’t.

This week, Anthropic publicized its findings that AI models suffer from “inverse scaling” — which is to say an AI model that’s given more time to think tends to produce worse results.

AIs think pretty much like humans do, so we too probably suffer from inverse scaling.

If so, now may be the time to follow Ozzy’s advice — in investing and everything else — and try our best to live in the moment.

Let’s check the charts.

AI automation and GDP growth:

Pick your level of AI automation (the X-axis) and this chart from Epoch AI will tell you what level of GDP growth to expect. The TL;DR is that “the fraction of labor tasks automated by AI doesn’t need to be very high to yield very large growth impacts.” Per the chart, if only 20% of tasks are automated, they expect GDP to grow at an unfathomable 10% per year.

Revenge of the dorks:

The Reddit gang is back and this time they’re betting on Dunkin Donuts, Opendoor, Rocket Companies and Kohl’s. Why those four? I doubt even an all-knowing AI could make sense of it. But if you think AI is about to take all the jobs, rolling the dice one last time in memestocks may be less irrational than it seems.

There’s a lot of dice rolling right now:

Data from Goldman Sachs shows that, as a percentage of overall activity, trading volumes in the most speculative stocks have rarely been higher than now. As an indicator, that is generally bullish before it’s bearish.

Tariffs have not slowed China’s export boom:

The more obvious reason for equities being at all-time highs is that the Liberation Day tariffs were not nearly as disruptive to the global economy as generally feared. China’s exports, for example, continue to boom. That show of strength may be why President Trump is reportedly more willing to make concessions to get a deal done.

China is flooding the world with cars:

I’m glad we still have an auto industry in the US, but if I could buy a high quality EV for $8,000, I would.

Humans remain in demand:

The Atlanta Fed’s wage tracker shows US wages growing at 4.2% — still well in excess of inflation (however measured).

Robots are not taking the jobs yet:

Against the popular narrative, the FT’s data analysts find that “US jobs at high risk from generative AI have not been more likely to shed young workers since ChatGPT launched.” (So far.)

False alarm:

Polymarket bettors believe that the odds of the US going into recession this year are down to just 18%.

Comfortably familiar: 

The US economy is on pace to grow 2.4% in Q3. 

Could we soon be growing at 10x that?

It’s thinkable — but a long way from predictable.

Have a great weekend, productive readers.

Get the news in your inbox. Explore Blockworks newsletters: