Bubble Talk Rising, Will AI Continue as Hottest Bet in 2026?

The market had been continuously declining due to liquidity concerns caused by the U.S. government shutdown and financing worries surrounding the OpenAI supply chain. However, now that the U.S. government has reopened and NVIDIA has reported strong earnings, the market still isn't buying it.

What exactly is wrong with the U.S. stock market? With only five weeks left in 2025, has it become a complete waste of time? Does AI still have a future in 2026?

I. Looking Ahead to 2026, What Exactly Are Trump's Demands?

This year has one last chance for a rate cut—the FOMC meeting on December 10. From a macro perspective, the market's short-term nerves seem to be dominated by rate cuts. Therefore, when Federal Reserve officials continuously release conflicting signals, the market starts to exhibit an "internal friction" trend.

However, if we consider the U.S. midterm elections next year and the long-term strategic development of the U.S., we can basically identify two fundamental needs:

a. Immediate Realistic Issues: Inflation, more specifically, the call for rent cost reduction through rate cuts.

Currently, the largest expenditure for residents is rent and mortgage payments. Especially now, the proportion of monthly rent to monthly income is almost at a historical high in recent years, accounting for 22% of disposable income. In this situation, the slowdown in price growth and residents' ability to bear the cost of living are almost two separate issues.

High housing prices and soaring mortgage rates have jointly pushed up mortgage and rent costs. In the short term, the key to improving rental affordability is to drive down mortgage rates through rate cuts. Currently, U.S. mortgage interest rates are basically standing above 5.5%. Rate cuts can efficiently improve rent and mortgage costs.

Take the election of the new mayor of New York as an example. His core policy agenda, in Dolphin Research's view, is essentially "administrative intervention" in prices—freezing rents, free public transportation, and government-run grocery stores.

Trump's policies have also become noticeably more pragmatic:

1) Patching up the trade war: No longer imposing tariffs on certain core price-related categories, such as agricultural products and fertilizers;

2) Applying extreme pressure to resolve the Russia-Ukraine conflict and stabilize energy prices;

These measures, which essentially control prices, are likely to continue into the 2026 midterm elections.

b. The Vast Ocean of Economic Development: AI needs cheap capital for financing

As of the end of 2025, U.S. AI capital expenditures, among the companies observed by Dolphin Research, are still primarily supported by equity financing and cash reserves. Only a few new cloud companies or companies with unhealthy balance sheets are using debt financing to fund capital expenditures.

However, by 2026, U.S. stock market giants will also begin to announce bond financing, and the market will increasingly see cases where debt financing is used to support AI capital expenditures.

During this period of high private capital expenditure, if interest rate costs remain high, it will inherently increase the return risk of these capital expenditures. From a strategic development perspective, during the high-density investment period in AI infrastructure, the U.S. needs cheap capital to support the widespread advancement of AI infrastructure.

When these two issues are combined, the U.S. government has a very clear and strong demand to push for rate cuts, both in the short and long term.

II. Is the December Rate Cut Really That Critical?

c. Federal Reserve Candidates: Will There Be a More Fiscally Aligned Federal Reserve?

In May next year, the current Federal Reserve Chairman Jerome Powell's term will expire, and a new person will need to take over. Trump's two core requirements for the candidate—"loyalty + growth"—are to coordinate with fiscal easing.

The selection process is currently being led by Treasury Secretary Besant, and the candidates have been narrowed down to five people. The frontrunner is Hassett, while the least likely candidate is Michelle Bowman. Compared to Powell, they are all more dovish.

d. Looking Ahead to Next Year, Is It Suitable for a Rate Cut?

There are still some key macro data missing, but based on the employment and inflation data that have been released, the 120,000 non-farm payrolls in September are not low, but the increase in labor force participation and the previous downward revision of employment figures indicate that the unemployment rate is indeed rising (4.44%). Meanwhile, the month-on-month growth in residents' wages is basically in the range of 0.2-0.4%, showing very stable performance. Core inflation remains at 0.2%, which is also good.

Although these figures are somewhat outdated, they overall reflect a state of weak employment and low inflation risk.

If U.S. inflation (excluding tariff impacts) remains at a monthly rate of 0.2%, the corresponding annualized inflation rate will be around 2.5%. The current policy rate is 3.75%-4%. It seems that from now until the end of 2026, the total rate cut will be around 75 basis points, and it is unlikely to exceed 100 basis points.

The remaining issue is the timing of this 75-100 basis point rate cut: If the rate is cut in December, there will be only two rate cut opportunities left in 2026. If the rate is not cut in December, a rate cut is inevitable by the end of January 2026.

From the perspective of boosting economic growth through strong year-end consumption, a rate cut in December and no rate cut in January seems to be a reasonable choice. However, even if there is no rate cut in December, a rate cut is highly likely within a month after December 10.

From the perspective of U.S. government policy, if there is no rate cut in December, the dovish expectations for the 2026 rate cut schedule can still be guided by releasing information about the new Federal Reserve Chairman candidate.

Given the current market pullback, the upcoming release of U.S. fiscal funds, and the gradual increase in fiscal stimulus in 2026, even if there is no rate cut in December, it seems more like a macro narrative where all the bad news has been priced in.

III. Industry Investment: Is There a Bubble in AI or Is It the Core Conflict of Trading?

After clearing the macro issues, from an industry perspective, the current soul-searching question remains whether the current AI capital investment plans can match the potential value created by scene penetration. In other words, is there a bubble in current AI investments? Are current AI valuations showing signs of being extrapolated?

Dolphin Research believes that this question is difficult to answer definitively in the short term. Its value lies more in the discussion itself, as reflecting on this question is a process of returning to ideal investment thinking.

AI valuations have pulled back, but from the perspective of current economic input, the process of AI capital expenditure, whether it is the construction of data centers, the construction of numerous factories, the updating of network equipment, or the infrastructure of electricity, is essentially a "great reconstruction" of industrial infrastructure in the AI era, a process of reindustrialization of infrastructure.

Compared to concerns about the replacement of some jobs in the mid-to-late stages of AI terminal scene implementation, 2026 looks more like the first half of the reindustrialization process of AI capital expenditure implementation. Before AI replaces human labor, AI investment will first create a batch of infrastructure job demands. 2026 is likely to be a year driven by AI infrastructure, boosting domestic demand and job creation.

Combined with the rate cuts and fiscal stimulus in 2026, there will not be significant issues with U.S. domestic demand. This process may instead be a period of AI pullback and traditional industry valuation recovery.

In other words, for U.S. stock investments in 2026, during the AI infrastructure investment era, the AI industry chain needs valuation pullbacks to wait for the realization of terminal scene value. However, the industrial reconstruction brought by AI will drive demand transmission in traditional industries, and the recovery of domestic demand will instead provide opportunities for excess returns for traditional companies outside the AI chain.

IV. Portfolio Returns

Last week, Dolphin Research's virtual portfolio Alpha Dolphin did not adjust its positions. The portfolio fell by 2% for the week, outperforming the CSI 300 (-3.8%), the MSCI China Index (-5.5%), and the Hang Seng Tech Index (-7.2%), but slightly underperforming the S&P 500 Index (-1.9%).

This was mainly because a significant portion of Dolphin Research's portfolio consisted of cash, gold, and U.S. Treasuries, which were relatively risk-resistant during the market downturn. However, Pinduoduo's earnings report, due to its high weight, somewhat dragged down the portfolio's performance.

Since the portfolio began testing (March 25, 2022) until last weekend, the portfolio's absolute return was 109%, with an excess return of 94% compared to the MSCI China Index. From the perspective of asset net value, Dolphin Research's initial virtual asset of $100 million has exceeded $213 million as of last weekend.

V. Individual Stock Gains and Losses Contribution

Last week, Dolphin Research's virtual portfolio Alpha Dolphin fell along with the market. Pinduoduo's decline was more significant among the holdings. The specific stock gains and losses are explained as follows:

VI. Asset Portfolio Distribution

The Alpha Dolphin virtual portfolio holds a total of 18 stocks and equity ETFs, with 7 standard allocations and the rest underweight. Non-equity assets are mainly distributed in gold, U.S. Treasuries, and U.S. dollar cash. Currently, the ratio between equity assets and defensive assets such as gold/U.S. Treasuries/cash is approximately 55:45.

As of last weekend, the Alpha Dolphin asset allocation and equity asset holding weights are as follows:

VII. Key Events Next Week

This week, the earnings season for Chinese concept stocks continues, with Alibaba, Meituan, and new energy vehicle companies continuing to release their reports. The AI investment logic line for Chinese concept stocks and the outlook for new energy vehicles in 2026 are worth paying close attention to:

Risk Disclosure and Statement of This Article:Dolphin Research Disclaimer and General Disclosure

For recent articles on Dolphin Research's portfolio weekly reports, please refer to:

"This is the Most Down-to-Earth, Dolphin Investment Portfolio Has Started Running"