HTX Research's Latest Report Analyzes AI Assets in the US Stock Market: Technology Still in Early Stages, Capital Expenditure and Valuation Entering Later Cycle

Source: HTX Research

Recently, HTX Research, the research department under Huobi HTX, released its latest report titled "Industrialization and Bubble Cycle of Intelligence: Token Economics, Capital Expenditure, and Re-evaluation of AI Assets in the US Stock Market." The report analyzes the valuation status, cyclical position, and risk-return structure of various segments of the AI industry chain from the perspective of token economics. It argues that the AI industry and AI stocks are currently not in the same time cycle. The technological diffusion of large models, agents, and multimodal products is still in its early stages, while the enterprise application cycle is in the early to mid-stage; however, the capital expenditure cycle of hyperscalers has already entered the mid to late stage, with stock valuations, market concentration, and trading sentiment closer to the latter half of a bull market.

Variables Determining Stock Prices

Over the past three years, the market first traded on the supply scarcity of GPUs, HBM, servers, networks, and data centers, followed by trading on the capabilities leap of frontier models, inference models, and coding agents. Entering 2026, the core variables determining stock returns are shifting from model parameters and capital expenditure scale to the production costs of tokens, task success rates, user consumption intensity, enterprise workflow penetration rates, and whether massive AI investments can ultimately convert into free cash flow.

This shift is driven by a change in the magnitude of capital expenditure. J.P. Morgan Asset Management estimates that the capital expenditure of five US hyperscalers will be approximately $69.7 billion in 2026, with AI capital expenditure rising from about 33% of these companies' operating cash flow in 2023 to an estimated 93%. When capital expenditure consumes most of the operating cash flow, the market's focus must inevitably shift from revenue growth to capital return rates.

Bubbles Exist in Financial Structures, Not in the Industry Itself

================

The report provides a more precise dissection of the term "AI bubble." Cloud revenue, coding agent usage, chip revenue, and enterprise demand are all genuinely growing; AI technology itself is not a false narrative. However, capital expenditure, external financing, data center projects, private model valuations, and some overvalued second-tier assets have shown clear signs of bubble characteristics.

The way to assess the cost-effectiveness of targets also needs to be adjusted. Surface-level price-to-earnings ratios cannot directly represent true valuation levels: Alphabet's P/E ratio is distorted by investment returns, and Amazon's current accounting profits do not represent normalized valuations. What truly offers cost-effectiveness are companies that best match normalized valuations, competitive barriers, cash flow, and AI optionality.

At the current price and cyclical position, the report believes that Alphabet's comprehensive odds are the most prominent, possessing both full-stack capabilities and multiple growth engines; Microsoft's fundamentals have the highest win rate, but valuation expansion space is limited; Meta's surface valuation is well-matched with revenue growth, but capital expenditure risks are significant; TSMC is the "shovel seller" with the most stable quality in the supply chain, needing to bear geopolitical tail risks; NVIDIA remains the most attractive core semiconductor target after growth adjustments; Amazon has significant AWS and Trainium optionality. Oracle and Micron are high-odds but low-win-rate cyclical assets, while AMD, Arista, and Vertiv have excellent business quality, but current prices demand near-perfect execution.

The AI Mainline is Changing the Configuration of Crypto Users

==================

The influence of AI as a common mainline in global capital markets is reflected not only in US stock pricing but also in changing the asset allocation behavior of crypto users. When core AI targets like NVIDIA, Micron, TSMC, Broadcom, Meta, and Alphabet enter the daily investment portfolios of crypto users alongside gold, oil, ETFs, and pre-IPO assets, the boundaries between the two markets begin to blur. Funds can flexibly switch between BTC, ETH, leading AI stocks, gold, and ETFs based on macro environments, industry trends, and risk preferences, with more users now understanding crypto and US stocks as different allocation directions within the same global risk asset system.

Huobi HTX is one of the early platforms systematically promoting this direction in crypto trading. According to data disclosed in August 2026, the platform's TradFi contract segment has accumulated a trading volume exceeding $2.5 billion, supporting over 170 TradFi-related assets, covering US stocks, ETFs, gold, silver, oil, AI chips, storage, aerospace, and pre-IPO thematic assets like OpenAI and Anthropic.

The key to this model lies in the platform's large number of crypto users who have completed registration, identity verification, and asset deposits. These users typically hold stablecoins like USDT directly, eliminating the need to open traditional securities accounts or transfer funds to another financial system, allowing them to trade TradFi assets within the same account. When risk preferences decline, they allocate to gold, ETFs, or large tech stocks; when risk preferences rise, they increase the proportion of crypto and high-beta AI stocks, keeping funds within the same platform.

The Competitive Boundaries of Trading Platforms are Changing

=============

The growth of TradFi business indicates that future competition among trading platforms will expand from spot, contracts, liquidity, and listing speed to comprehensive competition around crypto, US stocks, ETFs, commodities, pre-IPO, wealth management, and AI investment tools. Platforms with true long-term competitiveness will upgrade their core capabilities from single trading abilities to global asset distribution capabilities.

This also confirms a larger judgment proposed by the report: AI is changing not only model capabilities and computational demands but also the flow of funds, asset allocation methods, and the organizational forms of financial products. Huobi HTX's early layout in the TradFi direction resonates with HTX Research's ongoing tracking of AI mainlines and cross-market capital flows—identifying cyclical positions, judging capital flows, and understanding the interconnections between different assets are not only core to investment research but also constitute a source of first-mover advantage in actual business decisions. As AI drives global financial markets into a new phase of integration, institutions that can simultaneously understand industry cycles and capital flows are more likely to secure advantageous positions in the next round of competition.

Note: The content of this article does not constitute investment advice, nor does it constitute any offer, solicitation, or recommendation of investment products.

About HTX Research

HTX Research is the dedicated research department under Huobi HTX, responsible for in-depth analysis across a wide range of fields including cryptocurrencies, blockchain technology, and emerging market trends, writing comprehensive reports, and providing professional assessments. HTX Research is committed to providing data-driven insights and strategic foresight, playing a key role in shaping industry perspectives and supporting informed decision-making in the digital asset space. With rigorous research methods and cutting-edge data analysis, HTX Research consistently stands at the forefront of innovation, leading the development of industry thought and facilitating a deeper understanding of the ever-changing market dynamics. Visit us.

This article is contributed and does not represent the views of BlockBeats.