Welcome back. As this is my last edition before the new year, it seems a fitting moment to look back on 2025 and revisit a few counter-consensus calls I’ve made over the past 12 months.

From US President Donald Trump’s sweeping protectionist agenda to persistent fears of a market bubble, the newsflow has been relentlessly downbeat.

But vibes don’t always match reality. This year has surprised to the upside. For all the gloom that took hold in the first quarter and following the April 2 “liberation day” announcements, the global economy, trade and markets have outperformed consensus expectations.

Reflecting on Free Lunch on Sunday’s contrarian takes (relevant links throughout), I wanted to highlight three reasons why many economists underestimated 2025.

1) Translating Trump

The US president’s tariff agenda was at the centre of most global macro judgments.

The speed with which his administration imposed duties in its first few months caught many off guard. By April 2, more observers began to believe Trump would follow through on his trade threats.

But in the April 6 edition, I argued, against the grain, that the president’s initial “reciprocal tariff” regime was unlikely to endure, given the economic and political constraints facing his administration. Days later, Trump postponed it after ructions in US equity and bond markets.

In the following months, the president’s policy threats were increasingly modelled as opening gambits and often reversed in what my colleague Robert Armstrong called the “Trump always chickens out” (Taco) trade.

As the year has progressed, the US effective tariff rate has fallen owing to a mix of negotiations and carve-outs designed to ease burdens on key industries and mitigate cost-of-living pressures. In turn, the outlooks for global growth and markets have improved.

2) Resilience

But tariff delays and dilution alone aren’t enough to explain this year’s solid economic out-turns. After all, the average effective US tariff rate is still around seven times higher than last year.

A key factor, I think, was analysts’ underestimation of the resilience and adaptability of economies, businesses and markets.

In the January 19 edition, I was bullish on China’s ability to outfox US protectionism based, in part, on the country’s dominance of supply chains, ability to export to other markets and innovative strengths. (This wasn’t the dominant narrative at the start of the year!) DeepSeek stunned with its low-cost artificial intelligence model just a week later.

Indeed, China’s strength in 2025 has supported global economic activity. This week the IMF upgraded its annual growth forecast for the country to 5 per cent. That’s an entire percentage point higher than its late-April projection.

Global policymakers’ efforts to counter the economic hit from Trump’s “America First” measures with spending, reforms and new trade deals have also propped up confidence and activity. I highlighted this possibility in my analyses on why macro narratives around European equity markets (January 5 edition), Vietnam (February 23 edition), India (September 21 edition) and Germany (February 9 edition) were too downbeat. (Former German chancellor Olaf Scholz cited the latter edition in a televised debate as part of his failed re-election bid.)

That said, I was guilty of undervaluing the resilience of US businesses. This contributed to my overly negative calls on the American economy and stocks (July 20 edition).

Irene Tunkel, chief US equity strategist at BCA Research, says a mix of mitigation strategies meant the projected impact of tariffs on large companies’ margins has, so far, proved too pessimistic.

“Companies reduced their reliance on imports by building inventories in advance, sourcing domestically and importing from lower-tariff countries,” she says. “Cost-cutting and efficiency improvements also allowed firms to preserve margins.”

Moreover, the frontloading of imports ahead of tariffs meant global trade was stronger than expected this year.

3) Shock absorbers

Even accounting for corporate resilience, the strength of the US economy in the face of the remaining tariffs and uncertainty has still surprised me. A few under-appreciated offsetting factors have played a role.

In the May 18 edition, I highlighted how healthcare in America had been propping up economic activity. Indeed, this year the sector has kept household spending buoyant and has accounted for a staggering 83 per cent of all jobs created in the US in 2025 so far.

But the AI boom has been this year’s primary economic crutch.

AI accounted for at least half of the US’s 1.6 per cent growth rate over the first six months of 2025. Investment in IT equipment has offset broader weakness in capital expenditure, while data centre building has countered low activity in construction.

I was too focused on ascertaining the future productivity effects of AI to appreciate the huge near-term appetite just to invest in the technology’s infrastructure. This meant I also underestimated the voracious demand for already richly valued US tech stocks, contributing to my lowball predictions for the S&P 500 (April 27 edition).

AI has supported activity more generally. The World Trade Organization estimates semiconductors, servers and telecommunications equipment drove nearly half of all global trade expansion in the first half of the year.

A weak US dollar has been another global shock absorber, boosting emerging markets’ performance in particular.

It has been a hard year to make predictions. Diagnosing the economy has been likened to reading tea leaves in a hurricane. (I outlined why in the October 12 edition.)

Still, analysts weren’t wrong about the risks — such as tariffs, uncertainty and financial markets priced for perfection — which explains the gloomy mood at the start of 2025.

But they did misjudge second-order effects, overlooking the adaptability of economic agents, existing sources of resilience and positive offsets. Indeed, many of the five optimistic scenarios I outlined in the March 23 edition did come to pass.

There are lessons here for next year, but they shouldn’t be overlearnt. Resilience this year does not mean risk has disappeared. The US Supreme Court ruling on the legality of Trump’s tariffs as well as the midterm elections could change the president’s behaviour. His existing policies, from tariffs to his immigration crackdown, could bite harder over time. Reform efforts in major economies could falter. And, if the AI boom does implode, it would be among the most unsurprising market crunches ever.

That’s precisely why Free Lunch on Sunday will keep stress-testing the dominant macro assumptions in the year ahead.

Send the narratives you’d like me to tackle in 2026 to freelunch@ft.com or via X @tejparikh90.

Food for thought

Does political power in the US translate to wealth? This working paper analyses capital gains on Capitol Hill.

Free Lunch on Sunday is edited by Harvey Nriapia