The “AI” or “Aïe, Aïe, Aïe” Year.

January 5, 2026

“Aïe, aïe, aïe” is a French expression used to convey sudden pain, worry, dismay, or mild distress, similar to “Ouch, ouch, ouch!” or “Oh dear, oh dear!” in English, indicating something hurts or there’s a problem, often used humorously or in anticipation of something bad.

• On January 6, 2025 I wrote in Fear:

“Americans are wealthy, employed and spending merrily; the 500 bps rise in interest rates has had little effect on GDP growth and inflation seems to be in check amid a productivity boom that could last several years.

Comparisons with the dot.com era should be tempered by the fact that the S&P 500 is a fundamentally higher-quality index today: higher margins, better, well-established and profitable leaders and lower net leverage, non only debt-wise but also operationally as most of the leaders are asset light and generate large amount of free cash. (…)

The fear is thus about the amplifier effect a tariff war could have if and when the economy slows.

But the amplifier is not what it used to be, particularly if higher tariffs are limited to goods, which now account for 31% of total personal expenditures. (…)

Still no crystal ball but solid consumer and construction spending will keep the economy humming, with the risk tilted on the high side. Inflation between 2.5-3.5%. (…)

The only sure prediction: it won’t be a tranquil year.”

• On April 7 (Fearless), after the S&P 500 lost 21% in 45 days:

“Suddenly, it’s a bear market!

Fear everywhere.

As equities tanked and bond yields cratered, few people were bold enough to speak their mind, fearing to get pilloried on some social media platform and/or lose some rights, a license, even a Congressional seat.

Then, China spoke.

China will fearlessly reciprocate with like for like duties. Bring it on Trump! Titan vs Titan. How fearless are you really? (…)

Trump fearlessly shouted: “CHINA MISMANAGED THE SITUATION, THEY PANICKED – THIS IS SOMETHING THEY CAN’T AFFORD TO DO!” (…) “TO THE MANY INVESTORS COMING INTO THE UNITED STATES, MY POLICIES WILL NEVER CHANGE.” (…)

Can he actually be fearless enough to find ways to backtrack while he might still have some credibility left? (…)

Wanna bet on that and buy equities here?

After all, the S&P 500 Index is down 20%. Its trailing P/E has declined from 25.7 at the end of 2024 to 19.7 on today’s pre-opening of 4900. Sounds cheap! (…)

There are a few mitigating factors to consider:

• oil prices, which peaked at $91 in 09/23 and started the year at $72, are below $60. Slowing wages and lower oil prices will keep services inflation (+3.8% a.r. in Jan-Feb) contained while goods prices (temporarily?) rise.
  

• bond yields closed at 4.0% Friday, down from 4.8% in January and lowest since October. Mortgage and other consumer rates should also decline.
  

• the U.S. private sector (corps and households) is not over-leveraged.

So, an optimist could say, probably not a recession, and maybe manageable inflation. (…)

On forward EPS, the 17.6 P/E is on its historical median, justifying accumulation if one believes that the current $278.96 forward EPS will hold amid the prospective economic chaos. (…)

Safer would be 17.5x $220-250 = 3850-4375, down another 10-20%. Realistic?

Only if you believe that a fearless Trump will keep fighting, against the American people, against Congress and litigations and against just about every business people. Does he really believe that a “little disturbance” will soon give way to a new era of prosperity? To the point of jeopardizing his slim majority and his place in history? (…)

There will be tariff negotiations and face-saving off-ramps provided. (…)”

• On April 28 (Fearful):

“The precipitous drop in world oil prices, mainly reflecting lower anticipated demand, is providing welcomed relief to consumers just before tariffs start biting.

Trump blinked several times last week, fearful that financial panic could jeopardize his “grand plans”. It really took Scott Bessent, “increasingly influential in Trump’s quasi-imperial court”, to explain that the worldwide stagflation underway would be blamed on him.

So he blinked on tariffs, he blinked on China and he blinked on Powell (…).

FEARFUL brings fear right back to the bully after Mr. Market spoke loudly and clearly. The Trump administration is now fearful of showing its backtracking, its improvised strategy, its surrender to the unanticipated fearlessness.

What comes next?”

Well, the AI spending stampede came next, along with a 35% plunge in oil prices, slowing wages and the growing realization that rising productivity would keep a lid on tariff inflation.

Lucky Trump, lucky us!

So long, American exceptionalism For the first time, investors are talking about ‘US risk’

(…) You have to poke around to see this clearly. On the face of it, US stocks took a hit in April, when President Donald Trump tariffed all of planet Earth, and rebounded when he backed down a week or so later, ending the year up by about 16 per cent. What’s not to like? 

True believers in the notion that the US has won at capitalism will tell you this proves that orthodoxy and institutions have prevailed, and we can just get back to the normal state of the world, where American assets are all that matters.

For US investors, based in US dollars, that is fair enough. But the rest of the world is now clearly looking at the country’s markets through an entirely new lens.

“Risk in the US is higher than it was,” one senior executive in the Canadian pensions industry told me earlier this year — a realisation that hit hard in the second quarter of 2025. “There’s less certainty in regulatory compliance, in tax. Things we thought were contracts are open to politics. You have to think about risk in a different way.” (…)

In a couple of decades of writing about markets, I have never heard investors talk about “US risk” before. Now it is perfectly normal. 

This grand sweep of history, the point at which high finance and geopolitics meet, is one thing. But asset managers are also responding to the cold, hard reality that 2025 was a stinker for overseas investors in US dollar assets.  “The US does not understand that the dollar hit this year was very painful,” said Vincenzo Vedda, chief investment officer at German asset manager DWS.

“People are rethinking how to manage that.” (…) The drop in the dollar at the start of 2025, however, means euro-based investors in that same index are up just 2.6 per cent this year. That’s not a typo. They are in negative territory on both the Dow Jones Industrial Average and on the Russell 2000 index of smaller US stocks. 

What’s more, normally, non-American investors can rely on the dollar to rise in times of stress, under the unwritten haven status it has enjoyed for decades. This has been a very handy stabilising mechanism, dulling or even neutralising the pain in times when US stocks are falling.

Now, investors no longer trust that this will work, particularly as the president seeks to assert his easy-money doctrine on the Federal Reserve.

This leaves them with limited choices — they can try to hedge away the currency risk by betting against the dollar, which is often expensive and rarely scalable across a large portfolio, or they can look elsewhere to balance out returns, or a bit of both. (…)

But quite aside from the bubble risk in that [AI] sector, which all investors acknowledge to some extent, they need to spread the political and currency risk in a way they have not done before. (…)

US asset managers tend to believe April 2025 was a blip and we are all back to business as usual. The truth is rather different.

In February 2024, I wrote American Exceptionalism: Don’t Extrapolate. I hereby reprint some of my comments and update the data.

Trends are not always our friends. After a while, we tend to take them for granted, they become “natural”, ingrained, widely expected. Without a complete understanding of how they happened, we may be surprised when they end and reverse.

This chart from J.P. Morgan Asset Management shows that

over the past 50 years, there have been different regimes of U.S. vs. international outperformance. In other words, outperformance comes in waves. After a long period of U.S. outperformance, it is worth considering whether we may be transitioning to a new wave.

Source: FactSet, MSCI, J.P. Morgan Asset Management

Growth arises from many sources. The American economy benefits from several advantages compared to its world competitors. To list a few:

• population growth, including immigration
  

• education
  

• productivity, dynamism, flexibility
  

• innovations
  

• energy
  

• dollar

These attributes have long been mainstays of the American economy over time and cycles.

One additional source of growth has emerged since 2008: the U.S. government has significantly intervened in the economy, boosting its expenditures from 21% of GDP to its current 25.5%, doing so with borrowed capital as opposed to higher revenues. (…)

The jump in leverage since the GFC has been nothing short of spectacular: the federal public debt exploded from 62% of GDP in 2007 to 120%, in 15 years! (…)

In effect, the U.S. government’s increased spending provided additional revenues to the private sector without any offsetting contribution extracted from corporations or citizens.

The budget space provided by the huge decline in interest rates since the mid-90s was entirely used to raise spending and debt.

As a result, interest expense now represents 15.4% of the government budget, up from 9.2% in 2010. Some late 2025 estimates project this share reaching 17% for the start of fiscal year 2026. (…)

This means that the federal government is losing considerable leeway to adjust spending to economic needs and its discretionary expenditures are unlikely to provide the same economic impetus as they did since the GFC.

The pandemic prompted the U.S. government to further boost spending and the Fed to flood the economy with liquidity. Americans merrily spent their pandemic bounty. Meanwhile, broken trade channels and the trade dispute with China are inciting businesses to reshore production, encouraged by significant government subsidies and increased protectionism.

Manufacturing construction doubled (+$110B) since mid-2022, ten times faster than GDP, while manufacturing shipments and new orders stagnated. Actually, manufacturers spent twice more building plants in 2022-23 than during all previous 20 years.

Manufacturing construction actually peaked in 2023 and turned negative in H1’25:

Manufacturing capacity utilization peaked at 80% in April 2022, highest in 22 years, from 76% pre-pandemic. It has since dropped to 76.5% and is likely to get lower as the more recent projects get completed. It was 75.3% in November 2025.

Most of this new capacity is not to meet new demand, creating overcapacity, mainly in China, that is now fighting for new orders, likely displacing other production in a deflationary domino effect.

The war in Ukraine also benefitted defense spending in the U.S.. (…)

Investors are paying top valuations for large cap stocks, clearly extrapolating the past without appreciating that the true American exceptionalism actually is all the exceptional factors that boosted its economy since 2009 and oblivious to the rising risk from its indebtedness as interest rates normalize.

We may well be in a melt-up fueled by Goldilocks sentiment and passive investing mechanically boosting the Magnificent 7 stocks, but the risk/reward ratio has reached a mined no-man’s land area. (…)

Since 2009, this author has been discouraging country diversification: in single words, Europe was seen as unmanageable, Japan as unscrutable and China as uninvestable. By comparison, the U.S. was very likable, and mostly reasonably valued.

Nobody knows how long the U.S. will remain such a magnet for capital and what will trigger the change in sentiment. But alternatives are now more interesting, allowing for at least some diversification.

Back in 2009, U.S equities were selling at a discount to the world as the GFC made it “uninvestable” to many (one reader called me a “bloody fool” after I wrote a very bullish post in March 2009 at 666!). The current premium is the largest ever as the U.S market has become “the only one”. (…)

To be sure, many alternatives carry their own stigmas: the Eurozone is still largely disfunctional and China is still China. But big opportunities generally hide in plain sight, particularly when nobody wants to look.

The S&P 500 ranks last in Goldman Sachs’ Peter Oppenheimer 2025 returns table. S&P 500 earnings actually outperformed most other markets, all of which enjoying significant P/E expansion while the US P/E stayed flat. Not meaningless.

(Factset, Datastream, STOXX, Goldman Sachs)

There is now nowhere to hide on valuation:

Even within the S&P 500, although one could argue that (1) the top 5’s P/E is not as bubbly as perceived, and (2), it has recently dropped to relatively reasonable levels if the AI era is only beginning.

Peter’s “top 5” stocks are NVDA, AAPL, MSFT, AMZN and GOOG (29.6% of the S&P 500).

To complicate things, KKR takes the top 12 AI-related stocks, adding META, TSLA, AVGO, AMD, CRM, NFLX and ORCL (together 10.4% = 40%), to reveal a different valuation landscape after adding many higher P/E stocks.

It’s this 40% that’s at risk if, per KKR (20% probability), “demand for ‘compute’ (a term used here to encompass infrastructure, processing power, and energy resources required to power modern data centers and AI models) fails to keep pace with rapid supply expansion, and AI monetization proves slower than expected, exposing cracks in the AI capex cycle.”

I am not worried about demand for compute. In fact, it will likely exceed expectations (see below). But the infrastructure buildout needs to deal with many potential roadblocks (power supply, equipment supply, costs, Chinese competition).

KKR’s glass is half-full:

While negative supply shocks including tariffs and reduced immigration pose headwinds, we do not anticipate a hard landing in 2026.

In our view, the Fed is easing policy into an earnings upswing, productivity growth remains above-trend, private sector balance sheets are healthy, and the global backdrop is turning more reflationary. Against this setup, we expect
corporate earnings to rise by 11% in 2026, with higher nominal GDP growth and margin expansion powering the next leg of the recovery.

We believe valuations can largely hold steady, supported by easing tariff concerns (barring Supreme Court outcomes), accommodative financial conditions, further (albeit fewer) Fed rate cuts, and ongoing tailwinds from prior global monetary easing.

We continue to view the revival in labor productivity growth as the ‘secret sauce’ to a more sustainable earnings recovery, keeping unit labor costs contained and enabling higher, non-inflationary growth alongside margin expansion.

Maybe more important, though, today’s S&P 500 represents a higher-quality index compared with 10-20 years ago. Specifically, it is characterized by stronger margins, lower net leverage, better credit ratings, and a more asset-lite sector mix (though, we acknowledge that this is beginning to shift for hyperscalers).

These structural improvements, in our view, justify somewhat higher valuations. We would also note that valuations are not a short-term market-timing tool; expensive markets can remain expensive in the absence of a clear catalyst. Key downside catalysts to monitor going forward include

a) a non-linear rise in layoffs stemming from further labor market deterioration;

b) an unwind of the AI capex cycle;

c) excessive investor exuberance and positioning; and

d) a renewed inflation surge prompting markets to reprice for rate hikes (a late-1960s analog).

While we remain mindful of these risks, we do not assume they materialize under our base case [60% probability] at this stage.

Despite elevated valuations, we forecast the S&P 500 to rise about 9%, reaching 7,600 by end-2026. For 2027, we currently assume a more modest return of 6-7%, implying a price target of roughly 8,100.

About compute demand, Bain & Co. just updated its AI adoption survey:

Between the third quarter of 2024 and the third quarter of 2025, the percentage of companies ranking AI as a top-three strategic priority rose from 60% to 74%. The share who ranked it as their No. 1 priority more than doubled, with 21% of respondents ranking it on top. 

The percentage of companies saying that AI posed a very high risk of disruption in their industry more than doubled between the fourth quarter of 2024 and the third quarter of 2025—the same period when agentic AI became the focus of discussion and experimentation.

Most use case categories are seeing an increase in the percentage of pilots moving to production at scale. Software development is a clear leader, with 40% of pilots moving to scale—a good indicator that it fits well with AI capabilities. There’s also a solid second tier of domains where between one-fifth and one-third of use cases are scaling, including customer service, sales, marketing, and knowledge worker efficiency.

As companies move AI out of pilots and into production, most report greater satisfaction with the results. Our survey found that of the 59% of companies that are meaningfully adopting generative AI, the technology met or exceeded expectations in about 80% of cases across domains.

About 62% of those who said generative AI met or exceeded expectations also cited improved business results or successful transformation due to its deployment, and 78% of those respondents said these results resulted in measurable revenue increases or cost decreases. In all, about 23% of all respondents said the use of generative AI had delivered more revenue or lower costs.

 Interestingly, satisfaction appears to increase as companies progress from using AI as an assistant to assigning it task or agentic workflow automation. Respondents using AI for agentic workflow automation were twice as likely to say it exceeded goals as those who use it as an assistant, and only half as likely to report disappointment.

Where AI failed to meet expectations, companies told us the technology could address some work tasks but not others. About 33% of these unsatisfied respondents said the technology worked at the pilot level but didn’t scale. About the same percentage said it was more expensive to develop than anticipated.

Three years after generative AI began its ascent as an essential business technology, executives across industries show few signs of losing interest. More companies are rating AI as a top priority, and they are getting more serious about developing strategies and managing effective deployment.

Indeed, the rapid embrace of AI exceeds the pace of any technology we’ve seen yet and is likely to persist as long as businesses continue to find innovative and productive applications for AI.

Many of the concerns that have held back adoption of generative AI have begun to see a very gradual reduction, including those related to in-house expertise, quality and accuracy, return on investment, and data readiness. The notable exception remains concerns around data security and privacy, which have risen over the past year, especially among companies that have moved from pilots to production.

Roadblocks

The AI CORNER segment in my September 23, 2024 Daily Edge post was about AI’s Power Play: “While David and I seeked AI/LLMs investment opportunities, one of our main findings is that the world will be hard pressed to supply the amount of energy needed to feed data centers over the next 10 years.”

The reality is that AI demand exceeds expectations but power supply is lagging in the West. The scramble to build data centers, amid a growing shortage of power, is turning into a scramble to generate power at all costs. I bet none of that is in Western hyperscalers’ business plans.

Data centres turn to aircraft engines to avoid grid connection delays Supply chain shortages drive developers to use smaller and less efficient power sources to fuel AI power demand

(…) The need for on-site energy is booming as data centres face wait times of up to seven years to connect to the grid, as well as a backlash over their impact on utility bills. By installing power sources such as aeroderivative turbines and generators next to their data centres, developers can power the training and running of their artificial intelligence models without the immediate need for a grid connection.

GE Vernova is supplying data centre developer Crusoe with aeroderivative turbines that are expected to produce nearly 1 gigawatt of power for OpenAI, Oracle and SoftBank’s Stargate data centre in Texas.

Sam Altman-backed aviation start-up Boom Supersonic announced a deal to sell to Crusoe turbines that are expected to provide 1.2GW of power and are “virtually identical” to those built for its jets. Boom Supersonic intends to use earnings from power turbines to help fund its jet business.

“Three or four years ago I imagined we would do the airplane first and energy second,” chief executive Blake Scholl told the Financial Times. “But then I got a call from Sam Altman who said: ‘Please, please, please make us something.’”

The use of generators fuelled by diesel and gas is also increasing. Manufacturer Cummins has sold more than 39GW worth of power to data centres and nearly doubled its capacity this year. While generators are often used by data centres as backup power, Cummins’ data centres executive director Paulette Carter says they are seeing “growing interest in on-site primary power”.

Energy secretary Chris Wright has suggested commandeering existing backup generators to fortify the grid, telling Fox News in November: “We will take backup generators already at data centres or behind the back of a Walmart and bring those on when we need extra electricity production.”

The use of generators has prompted concerns over emissions, since smaller power sources tend to be less efficient. While local and federal regulators place limits on when backup generators can be used, these are being loosened in response to data centre demand.

In Virginia, where “data centre alley” is located, the Department of Environmental Quality is considering allowing data centres to run diesel generators more often, while the Environmental Protection Agency said data centres could use generators to maintain stable power. (…)

However, the cost of on-site power is likely to be higher than a simple grid connection, since such arrangements miss out on the economies of scale that utilities enjoy. Analysts at BNP Paribas modelled the price of power at a behind-the-metre gas plant Williams Company is building in Ohio, for which Meta will be a customer. The result was $175 per megawatt hour, which is roughly double the average cost of electricity for industrial customers.

The rush for power may also die down when hyperscalers slow their capital spending. “We’re in a very strong market right now, but it won’t stay like that forever,” said Mark Axford of Axford Turbine Consultants.

The AI risk

• Barclays reckons that investment in software, computer equipment and data centres boosted GDP growth by around one percentage point in the first half of 2025.
• The numbers show that business investment in AI may have accounted for half of the growth in gross domestic product, adjusted for inflation, during the first six months of this year.
• George Saravelos, Head of FX Research at Deutsche Bank tells clients that “AI machines – in quite a literal sense – appear to be saving the US economy right now.” He says that “in the absence of tech-related spending, the US would be close to, or in, recession this year.”
• Private business investment excluding AI-related categories has flatlined since 2019, according to Deutsche Bank.
• Commercial construction outside of data centres – shopping centres, office buildings and the like – is in decline.
• Bank of America estimates that Microsoft, Amazon, Alphabet and Meta will make US$344bn in capital expenditures this year. That’s 1.1% of GDP – up from US$228bn last year.
• Bank of America expects the big four tech companies to pour another US$404bn into capital expenditures next year, though the pace of growth is expected to moderate.
• Data centres now account for around 35% of Turner Construction’s backlog in the US, up from around 13% five years ago.
• The surge in demand is straining supply chains. Lead times for electrical generators, switchgear and other equipment have stretched by months in some cases. “Every element of the supply chain is being stressed right now”.
• If AI investment simply stopped growing, that could knock another 0.5 point off. If it collapsed entirely, a full percentage point disappears.
• The scale of AI-related borrowing compounds the risk.
• “If you take a fragile labour market and you kick it with a capex bust, you’re probably going to get a recession out of it”.

Apollo Management illustrates the phenomenon and the inevitable slowdown in growth and, eventually, in contribution to GDP.

The data in Ed Yardeni’s left-hand chart is included in the data in the right-hand chart, suggesting that, in total, data center construction is no longer offsetting the weakness in total manufacturing construction.

 

Note also the wide gap since 2024 between nominal and real expenditures, highlighting how construction costs have exploded from the surge in demand.

Some other facts:

• JPMorgan Chase calculates that rising prices of AI stocks alone boosted consumer spending by US$180bn over the past year, as households spend a portion of paper gains.
• That’s 0.9% of total consumer spending, the upward line in the US “K” economy.
• Jonathan Millar, Senior US Economist at Barclays, estimates that a 20% to 30% stock market decline could shave one to 1.5% off GDP growth over roughly a year.

The Q3 US GDP data showed consumer spending up 3.5% after +2.5% in Q2. Impressive.

But the quarterly data hid some worrisome monthly trends that few pundits highlighted:

• Real personal income ex-transfer receipts was up only 0.1% MoM in September after 0.0% in August. Since March, this important income measure is –0.6% annualized and was up only 1.46% YoY in September (+1.4% for all of Q3, down from 1.9% in Q1 and 1.7% in Q2).
• Real disposable income, still up 1.9% YoY, only rose 0.1% MoM in August and 0.06% in September, +1.0% annualized for the last 2 available months.
• Growth in real expenditures slowed significantly from +0.37% in July to +0.25% in August and +0.04% in September, the last month with official data. On a YoY basis, spending in September was up 2.1%, after 3.1% in Q1 and 2.7% in Q2.
• Spending on Bank of America credit and debit cards rose 0.3% in October but were flat in November, in nominal dollars. The +1.3% YoY advance in November is actually negative in real terms.

Goldman Sachs analysis confirms that tariff driven price increases have negatively impacted real spending on goods:

Spending on services has been steady at around 2.3% YoY in 2025 but it is vulnerable to an income slowdown or to accelerating services inflation (+3.3% a.r. in Q3 vs +2.5% in Q2).

The government shutdown is still impacting most official data. We’re all flying semi-blind for a while. But many private sources suggest slowing consumer demand.

The latest employment numbers are worrying, showing only 17,000 new jobs per month since May, 10,000 in the last 4 months and –41,000 in the last 2 months.

Trump’s One
Big Beautiful Bill Act (OBBBA) stimulus might prove very timely, boosting income tax rebates early in 2026 (and cutting expenses late in the year). Trump will also try to find ways and means to send more money to consumers as we get closer to November. 2026 could be a volatile year, again…

The trends in employment are worrying. The weak, distorted official figures are confirmed by private Indeed Job postings which have declined 6.5% between January and December 12 while JOLTS data remained flat.

LinkUp data show a similar trend, in both goods and services, and “in every single industry except Retail-General Merchandise which rose a paltry 1% (so much for holiday hiring). (…) based on the huge drop in job vacancies in November, December job gains will certainly be negative, perhaps materially so.”

The 2026 outlook thus rests meaningfully on AI to:

• keep hyperscalers spending and boost adoption
• which would keep the economy humming
• which would keep employment positive
• which would keep consumers spending
• which would keep profits rising
• which would keep equities buoyant.

Sand or whatever in the AI machine would

• significantly slow the economy
• which would hurt profits
• which would hurt equities
• which would hurt spending even more
• which would kill the economy.

In truth, OpenAi is rapidly becoming too big to fail. Sam Altman understands that OpenAi, on the hook for $1.4T with no free cashflow for a while, critically needs to keep fickle VC investors excited. OpenAi is thus constantly scrambling to be best in most everything to keep the money flowing his way.

A daunting task, particularly because it’s built on a pile of debt, unlike its Western and (lower costs) Chinese competitors. A failing OpenAi would create significant economic and financial tremors until the winners, or the US government, pick up the pieces.

Deutsche Bank

Backstop???

• The president expects his Federal Reserve chair (to be announced… soon?) to lower interest rates if the economy is doing well. “Anybody that disagrees with me will never be the Fed Chairman!” Trump said on social media. Piling on with guidance, Treasury Secretary Scott Bessent backed the idea of reconsidering the central bank’s 2% inflation target.

Sure, what’s the problem with 3% inflation, unless you’re in the 80% of Americans making up the lower part of the K economy. This majority is characterized by stagnant real wages, increasing debt, and a lack of significant assets compared to the top 20% who drive the bulk of economic growth. Approximately 66% of Americans are reported to be struggling financially.

Needless to say, the KKR folks are clearly on the upper end of the K. From their lens, the economy is OK:

The U.S. economy has shown remarkable resilience in 2025, successfully navigating tariff pressures and geopolitical uncertainty. This resilience has stemmed from several underlying strengths:

1) an absence of overheated activity across traditionally cyclical areas including construction and inventory investment;

2) a cautious yet stable backdrop for consumer behavior;
and

3) productivity that extends well beyond the AI story, helping spin an ongoing ‘flywheel’ of innovation.

These dynamics have helped sustain household wealth through market gains and have bolstered corporate efficiency across industries. We believe these dynamics will continue to support growth through 2026 and 2027.

Although we maintain a constructive outlook, our optimism is tempered by recognition of evolving vulnerabilities. Specifically, we expect that cracks in the labor market, including moderating wage growth and fading hiring momentum, particularly in the public sector, will weigh on consumption in 2026, especially among lower- and middle-income households.

Coupled with a more constrained fiscal stance relative to 2025, these pressures are likely to limit the pace of economic growth. Even so, we expect the continued expansion of technology capex and broad-based productivity  improvements to
counterbalance these headwinds over the next two years.

Easy to say that the resilience “stemmed from several underlying strengths”. But the AI spending stampede, the plunge in oil prices and a booming stock market helped a lot!

RBA’s Rich Bernstein addresses the risk with the stock market which, let’s not forget, fed most of the upper part of K last year.

2025 was a historic year for speculation across the financial markets. The economy is healthy, and the banking system is functioning well, so the Fed’s rate cuts and the anticipation of future rate cuts have resulted in excess liquidity that the economy simply can’t absorb, and excess liquidity and leverage form the life blood of speculation.

Whether it was the equity market’s emphasis of AI, SPACs, and Meme stocks, the fixed-income market’s near-record narrow credit spreads, individual investors’ record use of options and levered ETFs, or the hoarding of cryptocurrencies, speculation was rampant in 2025.

Pushing the speculative envelope even further, several money management firms commented during the year that they are considering incorporating sports betting as an asset class. (I kid you not!) (…)

Performance since “Liberation Day”, when the most recent stock market rally began, reflects speculation. Chart 2 shows that quality and broader stock market measures substantially underperformed more speculative market segments. (…)

Investors forget that the Fed only indirectly affects the economy. When the Fed cuts rates, they do so explicitly to lower the cost of funding to the banking system and other lenders in hope that a lower cost of funding will spur additional lending in hope that the additional lending will spur economic growth.

Given that relationship, investors would be hard pressed today to find a bottleneck within the financial markets that is hindering lending and stymying economic growth which would justify the Fed cutting rates. Chart 7 depicts the Bloomberg Financial Conditions Index and financial conditions are clearly very generous relative to history.

We don’t think the Fed will be able to cut rates as much or as fast as investors currently believe because financial conditions are not hindering growth and, perhaps more importantly, inflation is simultaneously moving away from the Fed’s 2% inflation target. (…)

Unless the Fed changes its long-standing position and accepts higher inflation, it seems reasonable to suggest that current forecasts of how fast and how much the Fed will cut might prove overly optimistic.

Accordingly, our portfolios entering 2026 are geared toward more boring investments in which performance is less dependent on liquidity and on easier financial conditions.

The two themes that dominate our equity portfolios as we enter 2026 are dividends and quality. We think these two cornerstones of building wealth have been forgotten by many investors.

The compounding of dividends has historically been one of the easiest ways to build wealth yet is often overlooked. (…) on a total return basis, the S&P Dividend Aristocrat Index has marginally outperformed NASDAQ over the past 25 years with considerably less volatility.

Investors have commented for some time on the attractiveness of European and other non-US equities based purely on those regions’ undervaluation relative to the US, but ignored the regions’ inferior growth and were ultimately disappointed by performance. However, there appears to be a different rationale for non-US investing today in that non-US high quality companies now offer competitive if not superior earnings growth. Non-US developed market high quality stocks now offer investors the trifecta of superior growth, higher dividend yield, AND undervaluation.

As Chart 10 highlights, non-US quality has an expected total return (i.e., earnings growth + dividend yield) greater than that of the Magnificent 7 yet sells at a 30-40% discount. (…)

Chart 11 highlights that there have been only three previous periods in the last 30 years during which credit spreads have been this narrow, and major credit events occurred after each one. The late 1990s Asia/Russia crises, the Global Financial Crisis, and the Fed/inflation fears of 2022 were all preceded by similarly narrow credit spreads. (…)

We have had a structural allocation to gold over the past several years because gold has historically been a good hedge against uncertainty and uncertainty, by definition, can’t be predicted. We view gold as a “spare tire” in the portfolio.

Chart 12 shows the relationship between gold returns and uncertainty. Uncertainty has indeed risen over the past several years, and gold has accordingly appreciated.

Investors might want to ignore claims that cryptocurrencies are “digital gold.” As we explained in an earlier report, gold is driven by uncertainty, but cryptocurrencies’ performance is driven by liquidity. Gold has historically been a safe haven, but cryptocurrencies have been a means of speculation.

Rather than being similar, gold and cryptocurrencies are almost two diametrically opposed asset classes.

When sports betting is considered a new asset class, as it apparently is today, it’s easy to argue that speculation is dominating investors’ thoughts. Historically, it’s been prudent to keep portfolios simple and boring as speculation reaches a crescendo because boring suddenly becomes beautiful when speculation subsides.

About credit, please see Your Daily Edge of Jan. 2: Credit and Freedom

Richard’s piece is well worth reading in its entirety: 2026: Boring is beautiful. (Richard’s video).

Bore me not!

But after three years when the equity market’s rip-roaring run made a mockery of any bearish calls, sell-side strategists are marching in lockstep optimism, with the average year-end S&P 500 forecast implying another 9% gain next year. Not a single one of the 21 prognosticators surveyed by Bloomberg News is predicting a decline.

“The pessimists have just been wrong for so long that people are kind of tired of that schtick,” said veteran market strategist and longtime bull Ed Yardeni. He expects the S&P to finish next year at 7,700 — up 11% from Friday’s close — yet even he finds the lack of dissent a little concerning.

“That’s where my counter instincts come out: Things have been going my way for so long that it is kind of worrying that everyone else seems to have become optimistic,” he said. “Pessimism is on the out right now.”

If the Wall Street forecasters are correct in 2026, stocks are heading for their longest stretch of annual gains since the lead-up to the Global Financial Crisis. The highest targets among the cohort, if they materialize, would also mark the first time the S&P has seen four years of double-digit returns since the dot-com bubble of the 1990s.
For now, strategists seem to be leaning into a lesson learned the hard way over the past few years: Don’t underestimate the strength of the US stock market. (…)

“The macro set up is simply solid.” (Bloomberg)

Hmmm…

Economist Peter Atwater, who in 2020 popularized the idea of a K-shaped economy, said the current state of the US economy more resembles “a top-heavy Jenga tower,” a reference to the game of stacked wooden blocks where players attempt to remove one block at a time and place it at the top of the tower without collapsing the entire structure. (Bloomberg)

Watch the Sahm rule: When the three-month moving average of the national unemployment rate is 0.5 percentage point or more above its low over the prior twelve months, we are in the early months of recession. November: 0.43 up from 0.10 in July.

All the best during this other certainly-not-boring new year.

Crucially, for our own sakes, all the best to Sam Altman and OpenAi.

Watch Sam. Because in 2026, Sam rules!