Travel Reflections: Trust is Being Distributed. So is Intelligence.

 

The Weekender is my bi-weekly take on macro shifts and emerging themes. It’s not investment advice — or even our firm’s official view. I aim simply to inform, challenge, and maybe entertain. If you’d like this in your inbox every other Saturday morning via Northern Trust, subscribe to The Weekender.

I've just returned from two weeks in blissful Thailand for a family member's 80th birthday. It included plenty of beach time, jellyfish stings (vinegar, not urine, is the cure!), island hopping, jungle treks, snakes in the tent, an elephant sanctuary, a floating resort, lots (and lots) of temples, high-risk transport (tuk-tuks and zip lines), intriguing cuisine (crickets, ant eggs), and one very interesting — and educational — cabaret show, which my children and I are still recovering from.

When packing for the trip, I reminded myself of the adage "if you want new ideas, read old books," which I amended to "read any book," given the scarcity of my attention between four kids and an X account (and before they become extinct, with frontier labs now apparently buying millions of old books — and then often destroying them — for training data; a true story).

I took four, hoping 24 hours of daylight flying would allow me to finish at least two. They included 1873: The First Great Depression and the Making of the Modern World by Liaquat Ahamed, Salt by Mark Kurlansky, The Einstein Vendetta by Thomas Harding, and The Wide Wide Sea by Hampton Sides about Capt. James Cook’s last voyage. I started all four, finished the one on Einstein, most of the one on Cook, am at the point in Salt where Clarence Birdseye enters the story (read on), and am currently enthralled by 1873, given it portrays the global financial crisis of that year and so is rich in parallels — and distinctions — between the technology evolution then and now. Forgive me, for I'm now enriched in metaphor.

The Scarcity of Trust

But the thing that struck me most about Thailand — the thing that renewed my hope in humanity — was how trusting everyone was.

It was staggering.

The 20,000 baht in my shorts pocket, returned in full by the hotel cleaner. The waiter's insistence we check the bill to make sure they hadn't made a mistake — in our favor. The fact that many people never locked their doors, or when parking would leave the car in neutral, trusting others to push it out of the way if needed. I met one lady who returned to her car to find a man who'd sat there for several hours, waiting to apologize in person for bumping it. Karma, it seems, is flourishing.

Contrast that with the Western world, where secular religions are waning and where trust — at least as measured by the Edelman Trust Barometer — is at record lows. Per the report, most people hesitate to trust anyone who holds different values, views or information sources. Trust has moved from We to Me. Personal networks now outrank institutions in credibility, which explains the gravitational pull of social media. And nationalism. And echo chambers of every flavor. Rachel Botsman, in Who Can You Trust, goes further: It's not just that trust is eroding in its traditional, centralized forms — it's also changing shape. We may no longer trust our governments or media, but we trust a stranger to share a ride with (Uber), to share our room with (Airbnb), and to go on a date with (Tinder).

Trust is changing. It's being distributed.

So too, it seems, is intelligence.

AI is diffusing (but more on that in a second).

A Trusted Process

When trying to apply all this to markets — which, after all, is the point of these missives — I found myself asking what things, or people, I trust or value most when making investment decisions. What constants have stood the test of time, reminding us that things can change but people often don't. Things like being taught to seek companies that sell products consumers find difficult to substitute, abandon or forget — which explains why Altria (nicotine) is, according to Hendrik Bessembinder, the best-performing stock in history. Apple is another that produces a product people find hard to give up. Many of Meta and Alphabet's products are essentially free, an appealing quality and one that's seemingly hard to disrupt.

Invest in scarcity is another aphorism worth trusting — one the capital cyclists employ with (current) glee, especially if they own the sellers of scarcity and avoid the buyers of it. Returning to first principles — the fundamentals, the facts, not the feelings — means identifying the first-order movers, the foundational drivers of change, before thinking through the expressions of them. If you pick the rising tide, all boats float. And if I had to pick one tide that continues to rise, it would be demographics. Aging populations, in particular.

Here's why.

Growing Dependency, Greater Debasement

Exhibit 1 remains one of the key drivers of change. It's the rising tide, the first-order mover. The tide.

It helps explain why we have growing deficits, record money supply growth, and why currency debasement is likely to continue. It helps explain why hurdle rates for most investors are too low in real terms, and why interest rates remain stubbornly high despite market inflation expectations being staggeringly low (the one-year inflation breakeven sits below 2%). It helps explain why supply-constrained real assets should remain well bid, and why technological innovation generally, and AI especially, is now a sovereign strategic priority — one many Western governments want to see succeed, despite some participants suggesting it won't. Those skeptics are betting against large, vested interests with deep pockets and price-insensitive motivations.

On the top in Exhibit 1 is the labor force participation rate versus debt-to-gross domestic product, inverted. On the bottom is where labor force participation could be heading, based on extrapolating the birth rate — its key variable — from 14 years ago.

EXHIBIT 1: SHRINKING LABOR FORCE DRIVES DEBT

 

 

As the labor force shrinks, so does the number of taxpayers available to fund the growing number of retiring baby boomers. Absent higher taxes, new debt must be issued. Worryingly, labor force participation has fallen again — now at levels not seen since the mid-1970s (excluding Covid). If you don't remember that period, read John Plender, who does. That's the Way the Money Goes is one old book full of new ideas on what to do when inflation is a problem, why scarce and real assets are attractive when system trust declines, and why bonds were once called return-free risk.

Scarcity Is the Shield . . .

During periods such as these, it's important to think about both how you hedge against the corrosive effects of monetary debasement and how you cure it.

Then, invest in both.

The hedge, the shield, is supply-constrained real assets — those whose supply can't grow as fast as the money supply they're priced in. Gold is the obvious example, especially as system trust declines. I haven't discussed gold in a while, but I'm minded to after reviewing the exhibit above and seeing China add roughly 15 metric tons of gold in June, its largest monthly purchase since October 2023.

The important signal isn't the number, it's the behavior: Beijing accelerated purchases into a sharp drawdown in bullion, suggesting it's treating weakness as an allocation opportunity rather than a reason to pause.

Bitcoin is considered by some to belong in this category too (let's revisit post-Clarity Act), as do commodities — notably those exposed to secular demand shifts from AI, the energy transition and national security, like copper, silver and lithium — and stocks, assuming those stocks are experiencing supply withdrawal via buybacks.

De-Equitization and Supply-Constrained Equities

While the quantum of buybacks has declined in the U.S., it's off an incredibly high base. And as we've discussed previously, there's still authorization for around $1 trillion or so this year, which resumes in earnest once corporates emerge from their blackout period around mid-August. That other little market we've termed the Trade of the Decade — the U.K. — screens even better on this de-equitization metric.

According to Schroders, over half the market buys back a material percentage of its shares every year, while paying a decent dividend yield of about 3%, meaning the dividend compounds in reverse as the denominator falls. And that's only if companies have shares left to buy after record takeover activity — or before the government unleashes tax incentives to encourage British savers to buy British stocks (as we discussed in Recognising Change). Did I mention many of the constituents also provide claims on supply-constrained real assets, that there's a bit of an AI story going on too, and that the Financial Times Stock Exchange 100 Index has outperformed the U.S. year to date, with 11% total returns in U.S. dollars, hitting a record high this past week.

But I digress.

. . . And Productivity Is the Sword

While scarce assets are the shield, productivity likely is the cure — the sword — for high debt and deficits. Our greatest hope for productivity, or so it is according to Federal Reserve Chair Kevin Warsh and U.S. Treasury Secretary Scott Bessent, is technological innovation: AI, robotics and health care, including glucagon-like peptide-1 drugs.

The Congressional Budget Office's official baseline has debt-to-GDP rising to 156% by 2055 on roughly 1.3% productivity growth — which seems highly conservative and, as they themselves admit, is based more on recent history than the more distant past. If productivity were closer to 3%, however, the debt ratio could fall below 80% of GDP — a genuinely transformational fiscal outcome. If by distant past you mean the 1990s, when the second half saw productivity growth close to 3% thanks to the internet, then such an outcome may be possible once we're through the current impulse of techflation. After all, the internet was about communication; AI is about cognition. It innovates the innovators and could, some argue, become a force multiplier for productivity itself. Especially as it doesn’t need to wait for the infrastructure, the network or data layers to be built, which largely are — or soon will be. It can tap the powerful forces of compounding almost immediately.

The Threat — and Opportunity — from Diffusion

The bigger question, perhaps, is whether China could run the solar, battery and electric vehicle playbook again — this time targeting the U.S. AI stack across memory, inference and robotics. What if the economic value lies less in the frontier model and more in the diffusion layer? If history is any guide, the biggest winners from a general-purpose technology are not always the firms that invent it, but those that diffuse it most effectively — the lowest-cost scale provider — the infrastructure that enables it, such as energy and compute, and those who build applications on top of it.

For example, Nikolaus Otto invented the combustion engine, but Henry Ford industrialized it. Thomas Edison was rich by conventional standards, but not as wealthy as Clarence Birdseye (Frozen Foods) or presumably Elisha Graves Otis (or Otis Elevators) which made skyscrapers possible. Killer apps like the Fordson Tractor, alongside Clarence Birdseye's frozen foods, transformed American industry in ways unforeseeable at the time. Farmworkers accounted for around 40% of all workers in 1910. Largely thanks to these applications, they account for less than 3% today.

And rather than precipitating mass unemployment or civil unrest, that shift ushered in the Golden Age in America, les Trente Glorieuses in France and the Wirtschaftswunder in Germany. If AI follows other technology revolutions and unit costs converge toward marginal cost, what happens to adoption, to the infrastructure layer, to the application layer? Who are the Fords, the Birdseyes, or the Otises of this revolution likely to be?

And is there a limit to the demand for intelligence?

No one knows, but given the rate of change, we may soon find out.

From Provider to Participant

If AI models proliferate and intelligence tokens commoditize, prices may increasingly be set by the marginal producer. Chinese models are emerging as a potentially important force in that process, offering increasingly capable output at dramatically lower cost. Oxygen is also being given to the open-source movement — currently a hot topic not just in Silicon Valley, but in Washington, D.C., too (my money, for what it’s worth, is on open source). Meaning this free-rider problem poses a challenge to frontier labs like Anthropic, which — needing to justify lofty valuations — may look to pivot from AI provider to AI participant (or perhaps both).

One strategic response might be to leverage workflow telemetry gathered from servicing enterprises to move further up the value chain in selected categories, potentially creating tension with customers over time (if this sounds like Amazon, trust your instincts). There's already evidence this is happening: In financial services, Anthropic launched 10 agents in May, announced a $1.5 billion joint venture with Blackstone, Goldman Sachs and Hellman & Friedman to expand enterprise deployment, and has moved into biotech drug discovery.

It's too early to say who the winners and losers will be, but it's probably worth trusting the lessons of history: The value may not be in the diffusion layer mostly captured by scale, but by the enablers (energy/compute) and those applications built on top of it. Yes, some of these will be created by the frontier labs integrating vertically, but many more will be created by those yet to emerge, which could change the world in ways yet unforeseen. See above: farmworkers.

It’s going to be some ride (a bit like my tuk tuk).

I trust you can hold on, which means avoid leverage at all costs or you could end up like this chap, Leopold Aschenbrenner.

Is the Greatest Threat also the Greatest Opportunity?

Which raises one final question.

If open-source AI represents the greatest threat to U.S. AI dominance, and China is helping accelerate that diffusion, could China also become an increasingly important hedge against the concentration embedded in many portfolios today? Should investors be lifting their (under) allocations to Chinese equities, if only to find returns orthogonal to their greatest concentration risk? AI . . .

Well, I trust you know my views on this already.

But we'll pick up that thread again next time. Until then, have a great week.

GP

 

Glossary

1. Currency debasement: The reduction in the purchasing power of money..
2. Financial Times Stock Exchange 100 Index (FTSE 100): A capitalization-weighted index of 100 of the largest companies listed on the London Stock Exchange.
3. Labor force participation rate: The percentage of the U.S. population that is either working or actively looking for work.
4. Hurdle rate: The minimum required or target rate of return that investors expect to receive on an investment.
5. Inflation breakeven: What market participants expect inflation to be in a given time period, based on the difference between Treasury yields and inflation-linked bond yields..
6. Orthogonal: Orthogonal means two vectors are independent and have zero correlation. In an investment context, orthogonal suggests that returns between two assets have little correlation.