The Year-End Rap: AI, a K-Shaped Economy, and the Risks Beneath a Strong Market

A Remarkable Year in the Markets

From the bottom in April, the market rose an astonishing 37 percent. NASDAQ swings were even larger. It has been quite a year.

In this final wrap of 2025, I want to step back and analyze the year, its dominant trends, and the themes shaping the market as we head into 2026. All data discussed reflects market levels as of Friday, December 12th, 2025.

For the full year, the S&P 500 is up 16 percent and NASDAQ is up 20 percent. Those numbers alone tell only part of the story.

How the Year Unfolded

The year began quietly. Investors broadly believed that President Trump’s policies would be pro market. By February 19th, equities were modestly higher.

Then came tariffs.

On April 2nd, dubbed Liberation Day, President Trump unveiled sweeping tariff plans. From February 19th through April 8th, the market fell 19 percent. As investors concluded that the tariff impact might be less severe than feared, equities rallied sharply for the remainder of the year, driven primarily by AI.

From the April lows, the S&P 500 rose 37 percent. NASDAQ rose even more. Volatility defined the year, but the recovery was powerful.

The Four Defining Themes of 2025

I see four dominant themes that defined the market this year.

1. The K-Shaped Economy

US GDP is on track to grow slightly above 2 percent in 2025. On the surface, that looks healthy. Under the hood, it is far less reassuring.

Two percent GDP growth represents roughly 500 billion dollars. AI-related capital spending alone exceeds 400 billion dollars. Strip out AI and the rest of the economy is barely growing.

This is the definition of a K-shaped economy. That divergence shows up clearly in markets. Despite a strong year for indices, 37 percent of S&P 500 stocks are down for the year. The equal-weight S&P is up only 10 percent versus 16 percent for the cap-weighted index. The biggest companies got bigger.

2. AI Revolution or AI Bubble

AI was the tide that lifted many boats after April. Hyperscaler spending and Nvidia’s explosive revenue growth dominated the narrative.

This is not a fictional boom. The largest companies in the world are spending real money to meet real demand. Nvidia growing revenue more than 60 percent year over year says everything about the intensity of current investment.

In the near term, I expect the AI growth story to continue into at least the first half of 2026. But there are two risks investors must watch.

The first risk is structural. Power is now the binding constraint. Data centers are being completed without access to electricity. US power demand growth has jumped from roughly 1 percent annually to closer to 3 percent. That incremental demand equals the power consumption of multiple large cities. New power generation takes years to plan and build. If power shortages delay data center deployments, chip orders will slow and the virtuous cycle will weaken.

The second risk is intellectual and far more important. The entire AI hardware boom rests on one assumption: larger language models must keep scaling and that scaling must keep delivering meaningful gains.

Cracks have started to appear. Critics noted that ChatGPT-5 was not dramatically better than ChatGPT-4. Some prominent researchers now argue that scaling alone is hitting diminishing returns and that the industry must return to fundamental research.

If this view gains traction, hyperscalers may rethink endless scaling. If that happens, chip demand will slow and the AI-driven growth engine will stall. In my view, this is the fundamental risk to the entire AI story.

3. The Growth of Private Credit

Private credit has quietly become a major concern. The market has noticed.

Despite a bull market, alternative asset managers like Apollo and Blackstone are down double digits. The reason is fear around private credit.

The private credit universe has grown from roughly 2 trillion dollars in 2020 to around 3 trillion dollars in 2025. US banks have lent nearly 300 billion dollars to private credit providers.

There have already been warning signs, not from traditional losses but from allegations of fraud. Collateral misrepresentation cases like Tricolor and First Brands have rattled confidence.

Unlike mortgages, private credit lacks transparent, public data. If problems emerge, they are likely to appear suddenly. I do not believe private credit will cause the next recession, but when a recession eventually arrives, weaknesses in this market will be exposed.

4. Affordability

Inflation has slowed, but the cumulative impact of years of price increases remains painful. Affordability has become a defining political and economic issue.

Housing illustrates the problem perfectly. Lower mortgage rates may help at the margin, but affordability is fundamentally a supply problem, not a demand problem.

Federal proposals like down payment assistance or 50-year mortgages will only push prices higher. The real barriers are local zoning laws, minimum lot sizes, long permitting timelines, and heavy impact fees that can add 10 to 15 percent to construction costs.

Until policymakers confront these local supply constraints, the housing affordability crisis will persist.

Sector Performance Reflects the Themes

Sector performance in 2025 mirrors these four themes clearly.

Technology led the market, up 26 percent, but dispersion was extreme. Nvidia and Palantir soared, while software and tech consulting lagged as investors questioned long-term moats in an AI-driven world.

Financials performed well due to a steepening yield curve, M&A activity, and the absence of recession. Large banks rallied sharply. Alternative asset managers lagged due to private credit concerns.

Communication services was the best performing sector, up 32 percent, driven by AI exposure and the Warner Brothers bidding war. Cable and wireless companies lagged.

Consumer discretionary underperformed due to housing weakness and stress among lower-income consumers. High-end brands thrived while value-oriented names struggled.

Healthcare delivered modest gains, with drugmakers outperforming and health insurers suffering from rising loss costs.

Industrials benefited from AI infrastructure spending. GE Vernova was the standout as demand for power generation surged. Housing-related industrials lagged.

Staples were among the weakest sectors as consumers pulled back and tariffs pressured margins.

Looking Ahead to 2026

We enter 2026 with a market still heavily dependent on AI-related spending. The story likely continues in the near term. Beyond that, the two AI risks I outlined will matter enormously.

If investors begin to treat those risks as real rather than theoretical, the market could experience a significant pullback. AI has lifted many boats. The reverse would not be gentle.

The key lesson is timing. Being too early is often as costly as being wrong. Trends of this magnitude last longer than most expect.

For long-term investors who can tolerate volatility, staying invested may make sense. For those with shorter time horizons or limited risk tolerance, flexibility matters. Holding some cash is not a failure. Money market funds yield around 4 percent and provide optionality without risk.

Final Thoughts

This was an extraordinary year. The market rewarded conviction and punished complacency. As we move into 2026, the signals are there if you know where to look. Today, those signals increasingly appear in real-time conversations on platforms like X and Reddit.

Thank you for engaging, questioning, and challenging these ideas throughout the year. This has been the final wrap of 2025. I look forward to continuing the conversation in 2026.

Thanks for reading this week’s wrap.
If you’d like to catch my interviews and market breakdowns, visit The Real Eisman Playbook or subscribe to the Weekly Wrap channel on YouTube.

This post is for informational purposes only and does not constitute investment advice. Please consult a licensed financial adviser before making investment decisions.