For decades, investors looking for protection have turned to government bonds. Economist and fund manager Daniel Lacalle thinks that playbook deserves another look.

His preferred combination today is very different: U.S. technology stocks, gold and other hard assets.

The reason comes down to two powerful forces colliding at once. Governments are carrying enormous debt loads while artificial intelligence has the potential to dramatically lower costs and increase productivity.

Why Lacalle Is Betting on AI

Lacalle, chief economist at Spanish private bank Tressis and co-manager of the Adriza International Opportunities Fund, is unusually bullish on artificial intelligence.

“I’m not afraid of artificial intelligence. I’m afraid of human stupidity,” Lacalle told MarketWatch.

His argument goes beyond enthusiasm for AI stocks. Lacalle believes AI could become a major deflationary force by allowing businesses to produce more with fewer resources, streamline supply chains and complete analytical work dramatically faster.

That could matter enormously in an economy dealing with persistent inflation, high interest rates and massive government debt.

Lacalle sees AI-driven productivity as one of the few forces powerful enough to offset some of those pressures over the next decade.

For investors, that explains his preference for companies tied to innovation and long-term productivity growth, particularly U.S. technology companies.

The Bond Market Is Sending a Warning

The more controversial part of Lacalle’s argument concerns government bonds.

Long-term Treasury yields have risen sharply, with the 30-year yield recently reaching its highest level in roughly two decades. Investors are demanding substantially more compensation to lend money to the federal government for 20 or 30 years.

Lacalle sees that as a market response to years of heavy government borrowing, inflation and monetary expansion.

Higher yields create a simple problem for existing bondholders. When yields rise, bond prices fall. Investors who bought long-duration bonds when rates were much lower have already experienced significant losses.

The unusual part is that stocks have remained comparatively resilient. The S&P 500 total-return index was up about 13% for 2026 through September 29, even while parts of the bond market struggled.

That divergence challenges the assumption that government bonds will always provide dependable protection when financial conditions become more difficult.

Gold Is Filling Part of the Defensive Role

Lacalle believes investors looking for protection against inflation and fiscal instability should consider hard assets, particularly gold.

His reasoning follows the behavior of global central banks, which have increasingly viewed gold as a strategic reserve asset alongside traditional sovereign debt.

For an individual investor, Lacalle questions why the strategy should be dramatically different.

His preferred portfolio concept combines hard assets with equities exposed to innovation and economic growth rather than relying heavily on long-duration government debt.

That does not mean bonds have no value. With Treasury yields around levels unseen for many years, investors are receiving much more income than they did during the near-zero-rate era.

Other prominent market strategists have reached the opposite conclusion from Lacalle, arguing that yields above 5% now provide enough income to make Treasurys attractive again.

That disagreement may be one of the most important debates facing investors today.

AI Could Help Fight Inflation

The most interesting part of Lacalle’s outlook may have little to do with Nvidia, data centers or the next generation of AI software.

It is productivity.

Companies spend enormous amounts of money managing logistics, designing products, processing information, forecasting demand and coordinating supply chains. AI has the potential to reduce the amount of labor, capital and time required for many of those processes.

If companies can consistently produce more output at lower cost, the result would be deflationary pressure across parts of the economy.

That could eventually improve profit margins while helping restrain prices.

The key word is eventually.

The AI buildout itself is currently extremely capital intensive. Major technology companies are spending hundreds of billions of dollars on chips, data centers, electricity and infrastructure. Bank of America estimates major technology companies could issue roughly $330 billion in bonds during 2026 alone to help finance the expansion.

Investors therefore face an unusual situation: AI could become deflationary over the long run while creating enormous demand for capital in the short run.

The Risk Behind the AI Trade

Lacalle favors U.S. technology, but that does not eliminate valuation risk.

A relatively small group of AI-related companies continues to have an outsized influence on major indexes. Recent market data also show weakening breadth beneath the headline indexes, with fewer stocks participating strongly in the rally.

That makes earnings growth increasingly important.

If AI investment produces strong productivity gains and corporate profits, today’s spending boom may prove justified. If companies struggle to earn adequate returns on the hundreds of billions being invested, valuations could come under pressure.

Higher Treasury yields raise the stakes further. Rising interest rates increase the discount rate investors apply to future earnings, which can be particularly painful for highly valued growth stocks.

Investors therefore have to watch both sides of the equation: AI productivity gains and the cost of the capital required to achieve them.

A Different Kind of Defensive Portfolio

Lacalle’s strategy also challenges what investors traditionally consider defensive.

He is skeptical of businesses that resemble bonds because they offer limited growth and have difficulty passing higher costs to customers.

Instead, he favors companies with strong balance sheets, pricing power, innovation and exposure to structural growth.

Beyond technology, that can include certain healthcare companies focused on research and development and businesses positioned to benefit from expanding middle classes in developing economies.

His broader principle is straightforward: in an inflationary world, owning assets capable of growing cash flow may provide more protection than simply owning assets traditionally labeled “defensive.”

Three Signals Investors Should Watch

The debate between stocks, gold and bonds can be simplified into three indicators.

Long-term Treasury yields: Continued increases in 20- and 30-year yields would suggest investors still want greater compensation for inflation and fiscal risk.

AI profitability: Revenue growth alone will eventually be insufficient. Investors will increasingly demand evidence that enormous AI capital expenditures are producing measurable productivity gains and profits.

Gold versus bonds: If gold continues attracting capital while long-duration government debt struggles, it would strengthen the argument that investors are reconsidering what qualifies as a safe haven.

These three signals together provide a useful picture of whether Lacalle’s thesis is gaining or losing strength.

The Bigger Question

There is a paradox at the center of today’s markets.

AI companies are spending extraordinary amounts of money building infrastructure, yet the technology they are creating could ultimately reduce costs throughout the economy.

Meanwhile, government bonds offer their highest yields in years, yet investors are increasingly questioning the fiscal risks behind long-duration sovereign debt.

Lacalle believes the answer is to own the assets positioned to benefit from productivity and innovation while holding hard assets as protection against inflation and fiscal excess.

Whether that strategy wins will depend heavily on one question: Can AI generate enough real economic productivity to justify the enormous amount of capital being invested in it?

If the answer is yes, the biggest AI investment story may extend far beyond technology stocks.