Fu Peng: Outlook on the Core Market Themes for 2026 — Focusing on the US Treasury Yield Curve and Three Possible Scenarios for AI Application Implementation [Fu Peng Talks 13]

“See the world from the trading desk—Fupeng reviews finance and economics”
^{《Fupeng Talks·Season Six》Fully Upgraded!Subscribe Now}

This year marks the tenth anniversary of Wallstreetcn's annual column 《Fupeng Talks》!

Season Six of 《Fupeng Talks》 is fully upgraded. In addition to regular videos, a new column group chat will update the latest market views & chart data interpretations (community updates first)

Beyond the usual "Fupeng Talks" content, we've added column online reading docs, sending everyone a daily trading early briefing, noteworthy research reports, recommended e-reading booklists, and extended reading materials. Everyone can also ask more questions and communicate within the group for common progress. Users who join the column, please scan the QR below to add the assistant and enter the exclusive "Fupeng Talks" discussion group.

Core Themes for the 2026 Market—Tracking U.S. Stock Valuations, U.S. Treasury Yield Curve, and the Implementation & Verification of AI Applications

See the world from the trading desk, Fupeng reviews finance and economics. This video was recorded on December 13, 2025.

In previous column community chart releases, I drew a chart on the U.S. Treasury yield spread, U.S. stock volatility, and valuation, discovering that many investors still misunderstood it. Therefore, in this article, I will reorganize the core data and internal logic of the chart to help everyone understand the current market pulse.

This chart contains two sets of key data:

First: the CAPE Ratio, which reflects U.S. stock valuations, is presented here on an inverse axis, with the current value corresponding to about a 40x valuation—note that on the inverse axis, the lower the value, the higher the valuation;

Second: the spread between 3-month and 10-year U.S. Treasuries. The key reason for using this indicator is that, over the past year, the U.S. Treasury yield curve has steepened significantly. To facilitate direct understanding of positive/negative relationships, the chart’s right axis marks the current structure of the yield curve. The trend shows short-end rates have continued to fall, while the 10-year yield has remained within a converging triangle for four years, with long-end rates stable. This creates a curve steepening driven by declining short-end rates.

This rate structure is closely related to the Fed’s "precautionary rate cuts" policy.

What are "precautionary rate cuts"? The core logic is to prevent the 10-year Treasury yield from breaking below the converging triangle’s lower trendline—if short-end rates keep falling and drag long-end rates lower too, it directly signals recession and would sharply impact U.S. stocks.

To understand the Fed's deeper thinking behind this policy, you need to look back at the market conditions of 2021–2022: At that time, U.S. stock valuations were high, market volatility in 2022 rose rapidly, the inverted yield curve widened—essentially the market was pricing in possible shocks ahead of time.

The immediate cause of increased market volatility was that the 10-year Treasury yield shot up to 3%–4% or higher during the Fed's rate-hike cycle, with high rates suppressing overvalued stocks. Notably, this coincided with the critical turning point of the AI sector’s lifecycle. Those who have followed "Fupeng Talks" will recall the main 2020–2021 market narrative: tracking ARKK (aka "Wood Sister") and Nvidia, noting that 2022 would be a turning year for Nvidia and the AI supply chain. That year, Nvidia shifted from "valuation-driven" to "devaluation" and then to "value-driven", and the AI sector moved from hype into a phase of certain growth.

This 2022 transformation came with sharp asset repricing: Nvidia's price fell 70%, ARKK 75%, Bitcoin crashed from $69,000 to $20,000—a collective pressure on high-value assets. But this deep adjustment highlighted the core value of the AI supply chain: Nvidia first demonstrated the commercial value of AI, industry shape became clearer, and upstream infrastructure investments gained real ground, driving both capex and revenue growth.

After its 70% plunge, Nvidia began a path of valuation recovery and value enhancement, with its market cap successively breaching $2T and $3T. ChatGPT's emergence at the end of 2022 was a clear sign of sectoral shift: After valuations reset, upstream AI capex was much more certain, evident in Nvidia's ongoing share-price climb and the historic downtrend in market volatility, forming a regime of low volatility.

For U.S. stocks, the deep 2022 devaluation laid the foundation for further growth, with the market entering a "dual-engine from numerator & denominator" virtuous cycle. Nvidia-led numerator (corporate earnings) provided core momentum, and denominator (valuation) continued rising, together driving solid market performance. In this phase, the yield curve had a Back structure, amid Ukraine war, energy crisis, and post-pandemic inflation—all coupling with fiscal stimulus, war shocks, and AI capex to push short-end yields higher.

Tracking the U.S. Treasury Converging Triangle

Looking at the 10-year Treasury yield’s converging triangle pattern, first the yield curve stayed in a Back structure, then gradually shifted to a Contango structure, the turn occurring in mid-2024—this perfectly matched Nvidia's $3T market-cap milestone.

At that time, certainty expectations for AI peaked, with Nvidia's CUDA monopolizing the AI upstream, causing investors to pile onto leverage frantically. After researching in Singapore, I warned institutional investors: the combination of ultra-low volatility and ultra-high certainty concentrates risks—leverage levels must be watched. Soon, this was proven: In 2024, our warning (a month and a half ahead) of Nvidia’s OTC leverage flash crash played out, with about $800B in positions under clearing pressure. Note: This crash was unconnected to AI fundamentals or Nvidia’s own value—it was a leverage stampede under certainty expectations.

After the flash crash, the market entered a one-year transition where both industry development and monetary policy were "precautionary". Valuations climbed to around 35x; Nvidia, post-deleveraging, resumed its upward share price, breaking $4T market cap, but the market structure changed—the most obvious shift is continued upward movement of the U.S. stock volatility center, even in November (normally stable), volatility kept rising.

It’s crucial to note that volatility is essentially "risk insurance", not a trading target and cannot be held long-term—its core function is hedging risk.

In the last 18 months, higher volatility "insurance costs" have been closely related to tariff changes, Nvidia volatility, start-year PLTR valuation reset, and the November 2025 market move, marking a change from "high certainty" to "reducing certainty/rising uncertainty", with significantly decreased market stability.

The key to lowered stability is a bottleneck in the AI supply chain—massive upstream capex is done, but commercial value at the downstream application end remains doubted.

This sectoral contradiction maps perfectly onto the yield curve’s characteristics: From the Fed’s viewpoint, short-term rates show labor-market worries; long-term rates show concern about inflation, with policies stuck in balancing "avoiding recession and avoiding inflation". From the supply-chain viewpoint, upstream capex is driving "structural inflation", yet it's unclear that the downstream application end will bring support—there’s fear of "deflationary drag" from applications.

To fulfill this balance, the Fed used short-end rate cuts to quickly steepen a Contango curve—"precautionary policy".

Currently, upstream AI industries remain robust (Nvidia and other core names fundamentally solid), but compared to the high certainty of 2023–2024, market confidence is diminished. Previously, the sector story was "to get rich, first build the road"—building out infrastructure; now, the conflict is "the road's done, is there any traffic?"—can applications generate scaled demand?

The drop in certainty is the core reason for increased volatility. Meanwhile, the Fed’s rate cuts have boosted valuations, and at a CAPE of 40x, U.S. stocks are now in a "high valuation + high volatility" regime, with upstream inflation and downstream profit at the crux of market tension.

Looking ahead, the fate of the yield curve will depend directly on the reality of AI application deployment, especially enterprise-grade heavy AI commercialization. If, in the next year, AI apps gain scale, forming an oligopoly and a tech moat, then upstream capex will get profit support from downstream, establishing a virtuous "upstream inflation + downstream inflation" cycle.

Under this scenario, curve steepening could mean long-end rates rise and short-term rates fall, with AI landing driving down volatility—because profit growth (numerator) offsets cost pressure (denominator), forming a healthy "profits cover costs" pattern.

Conversely, if AI applications fail to land, downstream doubt spreads while upstream rates get pushed up by other factors, curve steepening happens via long-end rates rising. This "no numerator confirmation + more denominator pressure" combo will greatly amplify volatility, and expensive U.S. stocks will face double jeopardy.

More extreme: If long-end rates break the triangle’s lower band and signal recession, it means "roads built, but no cars" comes true, the AI industry is exposed as a bubble, upstream capex NPL risk spreads through the market, yield curves tumble with high stock volatility, and recession risk erupts across the board.

Sorting these logics reveals that the market’s core problem isn’t just the Fed’s rate cut policy or short-term macro jobs data, but rather the "match between upstream investment and downstream output in the AI supply chain"—the essential challenge is whether, after productivity rises, production relations can adapt.

This key conflict will be the main theme for global capital markets in 2026. This sharing aims to clarify market misunderstandings, hoping investors take time this weekend to ponder the central logic and prepare for future decision-making.

Risk Warning and DisclaimerThe market has risks; investments must be cautious. This article does not constitute personal investment advice and does not account for an individual’s particular investment objectives, financial situation, or needs. Users should consider whether any opinions, views, or conclusions herein suit their circumstances. Investment based on this is at your own risk.