Globalization Ends Not With a Bang, But Tariffs 2.0

​ Summarize

Source: Bloomberg

This time, the silence from markets and the media has been deafening.

July 28, 2026 at 8:54 PM EDT

Trump touts tariffs to GM workers.Photographer: Sarah Rice/Bloomberg

To get John Authers’ newsletter delivered directly to your inbox, sign up here.

Today’s Points:

Protectionism Fatigue

This is the way globalization ends, not with a bang but a whimper.

The current US administration has declared war on the global free trade system. Last year’s “Liberation Day” imposition of the highest tariffs in a century drove a dramatic selloff in both stock and bond markets that within a week forced a presidential rethink. In an only slightly less aggressive move, Washington announced new blanket tariffs last week of at least 10% on 60 of its closest trading partners on the pretext of combating forced labor

And this time, nobody seemed to notice. Having dominated the agenda last year, this episode of the trade war elicited minimal market reaction. The media response, captured by the Bloomberg News Trends function that counts daily stories from all sources that appear on the terminal, speaks volumes:

Tariffs? What Tariffs?

Media interest in the latest battle of the trade war is starkly reduced

Source: Bloomberg News Trends

Not only did this dog fail to bark; it never even woke up. This is true even though emerging evidence is that the full effect of last year’s tariffs is still to show up in economic data. Chris Watling of Longview Economics explains the non-reaction as follows:

The first time something comes up, everyone gets worked up about it. You sell now and work out what’s happening later. Now, we’ve all analyzed the data. It will add a bit to inflation, but people think that’s that. It’s contained. There’s less volatility about what he can announce now. There are more guardrails from the Supreme Court.

People have grown used to erratic announcements from President Donald Trump, and have factored them in among the many risks of doing business. The Supreme Court ruled in February that the sweeping discretion he awarded himself in last year’s Tariffs 1.0 didn’t pass constitutional muster, so he has to play by more predictable rules.

The Liberation Day levies sprang a surprise, as they were expected merely to match other countries’ tariffs on the US, which would have meant only minor increases. Instead, the president blindsided everyone with levies far out of proportion to anything America was being required to pay. Such an incident cannot be repeated with full force.

Critically, the bond market turned “yippy,” a keyfactor in the Trump climbdown. But now it’s keen for an increasingly indebted US government to get revenues from any source possible. For a few months, tariffs had provided a welcome extra income stream. Revenue actually turned negative in June after refunds began to flow in the wake of the the Supreme Court judgment:

Reversal of Fortune

Tariff revenue turned negative in June as companies demanded refunds

Source: Bloomberg

In current circumstances, it’s almost a relief that Uncle Sam is tapping some new funds again, even if the long-term economic consequences are probably awful. Northwestern Mutual’s Brent Schutte points to Federal Reserve research that many companies couldn’t pass on the extra tariff costs immediately because they were operating under longer-term contracts. That forced them to wait before they could hike prices. He added:

Others have adopted a gradual “trickle-up” strategy, increasing prices incrementally to avoid shocking customers while preserving flexibility if input costs continue to rise. Ongoing uncertainty surrounding tariff policy — including potential rate changes, exemptions, and retaliation from other countries — is also encouraging firms to spread price increases over time rather than implement one large adjustment.

Gregg Fisher, founder of Quent Capital, argues that the insidious effects of the new reality will be felt in the longer term — and that the latest tariffs, because they appear to be more legally durable than the previous round, might be less problematic:

Imagine you’re a business leader trying to decide whether to build a new factory or expand a supply chain. Each day, you wake up without knowing what tariffs or tax rates you’ll be dealing with, who will be making the decisions, or even which country you should be producing in. It is very hard to sign off on a 10‑year project in that environment. That is the “Great Hesitation” and almost by definition throws sand in the gears of global economic growth.

The result is a steady increase in the cost of doing business, reflected in higher inflationhigher costs of capital, and steadily rising bond yields. Just as globalization was a key factor in the multi-decade declining trend in bond yields that followed Paul Volcker’s war on inflation, so raising the drawbridges has now triggered a steady rising trend:

This hasn’t prompted a major market bang in the short term, but the chances are that we’ll be whimpering over the consequences for decades to come.

The Big Tech Kryptonite

Four tech behemoths – Microsoft Corp., Meta Platforms Inc., Amazon.com Inc. and Apple Inc. — are due to report to the market between Wednesday and Thursday. After last week’s reception for Alphabet Inc., the first hyperscaler to report, a lot is at stake. Alphabet delivered the kind of earnings that investors usually greet with enthusiasm, as its cloud computing revenuesurged by more than 80% and advertising also beat estimates. But as the AI build-out continues, investors are focusing far more on what it costs to generate that growth. Alphabet’s shares took a 7% dive, chiefly because it announced still further increases in capital expenditures.

That same backdrop will shape this week’s earnings, with the spotlight likely to fall less on revenue growth than on the returns generated from ever-rising AI investment. 

In what is already a tumultuous year for the tech giants, any disappointment could reinforce the narrative they are no longer the AI revolution’s clear-cut winners. Excluding Apple, the three companies reporting this week are down for the year. The Magnificent Seven’s underperformance is now so pronounced that Wall Streeters increasingly label them the “Lag Seven”:

It’s hard to say what prompted the apprehension over capex. However, Alphabet’s free cash flow turned negative for the first time since its IPO in 2004; that helps to explain why investors are growing less patient about funding an investment cycle whose payoff remains uncertain. This chart by John Osterweis of Osterweis Capital Management shows that hyperscalers’ free cash flow has turned negative, almost entirely as a result of higher capex: 

For all of AI’s promise, it’s not clear that the hyperscalers can monetize their investments, particularly once the costs of maintaining the new multi-billion-dollar infrastructure are taken into account. For now, the beneficiaries of this spending, in semiconductors and memory chips, are in the driver’s seat:

Even as the spending plans are scrutinized, it’s unlikely that the market has completely dismissed other measures of performance. Hyperscalers’ relative performance compared to chipmakers has recently climbed above its 50-day moving average after spending nearly eight months below it. A remarkable rerating may be coming to an end:

While it’s too early to declare the start of a sustained rally, the driving forces offer clues to how much further it may have to run. More than 80% of respondents in Bank of America’s latest Fund Manager Survey identified semiconductors as the market’s most crowded trade. Meanwhile, the Magnificent Seven have grown meaningfully cheaper; Microsoft now trades at about 19 times estimated earnings, well below its average multiple of 27 over the past decade. Meta trades at roughly 14 times, versus a 10-year average of 20:

Bloomberg Intelligence’s Jennie Li notes that if Big Tech’s results confirm that AI spending is still translating into revenue and earnings, the greatest opportunities may be where valuations have fallen the furthest without a comparable drop in the fundamentals. Companies that still look more fully valued, like Alphabet (trading at 17 times projected earnings even after last week), are unlikely to escape investors’ wrath.

Further, computing capacity (or “compute”) still lags demand. Hyperscalers can defend the view that their massive capex outlays are justified. Osterweis’ Nael Fakhry argues that they view data centers as critical to the future of their businesses, even if the spending is monumental. The five largest hyperscalers (Alphabet, Amazon, Microsoft, Meta, and Oracle Corp.) plan to invest roughly $740 billion in 2026, about 75% higher than in 2025.

On this basis, Morgan Stanley’s Stephen Byrd sees them raising available computing capacity to about 120 gigawatts by 2028, up from about 30GW last year. With compute in short supply, that could make this an attractive opportunity to buy, he suggests — just as a shortage of chips prompted a rally for semiconductor stocks this year. But even here there are risks; Moonshot’s emergence underscores China’s rapid advance in AI and raises the question of whether very capital-intensive business models are needed.

Richard Abbey

Survival Tips

Some more streaks that transcended luck and showed there must be something at work. Edwin Moses won 122 consecutive 400-meter hurdle races between 1977 and 1987, breaking the world record four times during that period. For a streak that proved to be part of arguably the biggest upset in the the history of soccer, Jamie Vardyscored in 11 consecutive games as Leicester City marched toward their gloriously improbable Premier League title in 2015. In squash, the Pakistan genius Jahangir Khan managed 555 consecutive wins, which appears to be the longest unbeaten run in any professional sport. 

Ted Williams reached base safely in 84 consecutive games during the 1949 season(since Moneyball, an achievement now recognized as more consequential than a streak of hits alone). That broke the record of 74 set by Joe DiMaggio in 1941 and later tied by Williams.The best anyone has done since then was 2004 Red Sox hero Orlando Cabrera’s 62-game streak. And the Atlanta Braves won 14 consecutive division titles from 1991 to 2005, which is astounding consistency — marred in memory by their failure to win more than one World Series in that time. 

On losing streaks, in Irish football, County Mayo won the All-Ireland football title in 1951, celebrated, and then “according to the myth, did not show enough respect on passing a funeral as they enjoyed their revelry from an open-top lorry.” A reader tells me that, according to legend, the presiding priest “put a curse on the county and said they would never win another title until all from that winning 1951 team had died.” Mayo played in 11 All-Ireland finals and lost them all until the last 1951 team member died in 2021. In its first final since then, Mayo defeated reigning champions Kerry last week to win their first title in 75 years. It’s still short of the Red Sox’ 86-year Curse of the Bambino, or the 108-year drought suffered by the Cubs in what became known as the Billy Goat’s curse, but nonetheless impressive. 

For a Madoff-ian streak that looks too good to be true, maybe we should look at the Harlem Globetrotters’ 8,829 consecutive wins from 1971 to 1995. Could it be that their competitors weren’t really trying? Surely not. Are there any more out there?

More From Bloomberg Opinion

  • A Troubling Glimpse into an AI Musktopia: Liam Denning
  • Spain Is Starting to Annoy the Rest of Europe: Mihir Sharma
  • Warsh’s Twist Could Be Higher Fed Rates and Lower Mortgage Rates: Robert Burgess

Want more Bloomberg Opinion? OPIN. Or you can subscribe to our daily newsletter.

Like Bloomberg’s Points of Return? Subscribe for unlimited access to trusted, data-based journalism in 120 countries around the world and gain expert analysis from exclusive daily newsletters like Markets Daily or Odd Lots.

Leave a comment