
Old Market Indicators Broken Since ChatGPT Launch: Fund Manager Releases New Indicator ETFs
This article was created with the help of artificial intelligence.
Key Takeaways
- Harbor Capital Advisors claims that traditional economic indicators such as the Manufacturing PMI have lost their historical correlation with the S&P 500 since ChatGPT's launch in November 2022, though independent sources have not verified this divergence thesis to date.
- In response to the diagnosed market shift, Harbor Capital launched the 800VDC AI Datacenter Ecosystem ETF (NYSEARCA: HUMM) with a net expense ratio of 0.49 percent, designed to represent the new AI-driven economy.
- Spenser Lerner of Harbor Capital estimates that hyperscalers and data center operators will invest roughly one trillion U.S. dollars in AI infrastructure through the end of the decade, representing the first phase of the AI boom.
- Harbor Capital's commercial interests raise questions: the firm sells ETF products that directly benefit from its own argument about the breakdown of traditional indicators.
- Experiments with ChatGPT as an investment adviser showed mixed results: a June 2023 test achieved around six percent returns over roughly three weeks, but experts warn of parallels to earlier speculative waves such as the GameStop hype.
In November 2022, OpenAI released the AI chatbot platform ChatGPT. What was initially regarded as a technological milestone represents, according to Harbor Capital Advisors, a fundamental break in how financial markets function. Kristof Gleich, president and chief investment officer of the firm, argues that established economic indicators have lost their role as reliable stock market forecasters from that point forward.
Thesis on the Collapse of Forecasting Power
Gleich cites as the central example the Manufacturing Purchasing Managers' Index (PMI) – a survey of manufacturing activity trends that for decades showed strong correlation with the S&P 500. The PMI measures monthly sentiment among purchasing managers in the manufacturing industry and is regarded as a leading economic indicator. After November 2022, this historical relationship diverged: the PMI remained relatively stable while the market rose significantly. 24/7 Wall St. points out that this divergence thesis could not be independently verified.
"The AI economy requires different frameworks, different benchmarks, new metrics and frankly a different language to describe it, measure it and invest in it," Gleich explained according to Harbor Capital Advisors. The argument goes: traditional industrial indicators do not adequately capture value creation from software, data centers and cloud infrastructure.
New ETFs as Response to Diagnosed Market Shift
Harbor Capital Advisors responded to its own analysis by launching new fund products. The most prominent is the 800VDC AI Datacenter Ecosystem ETF (NYSEARCA: HUMM) with a net expense ratio of 0.49 percent. The fund aims to represent companies in the AI data center ecosystem – from chip manufacturers to cooling systems to power suppliers for data centers.
The product development raises questions about commercial interests. 24/7 Wall St. puts it plainly: "Harbor's argument is real. So is the commercial interest behind it." The firm sells ETFs that directly benefit from its own thesis that old indicators no longer work and new investment instruments are needed.
One Trillion U.S. Dollars for AI Infrastructure
Spenser Lerner, head of multi-asset solutions at Harbor Capital, divides the AI boom into two phases. The first phase encompasses infrastructure buildout through the end of the decade. Hyperscalers – providers of massive cloud computing capacity such as Amazon Web Services, Microsoft Azure and Google Cloud – and data center operators are investing roughly one trillion U.S. dollars in AI infrastructure according to Lerner's estimates. This sum flows into high-performance chips, specialized cooling systems, power supply and network infrastructure.
The second phase that Lerner mentions is not described in detail in the research material. It is said to contain inflationary components, but details on timing or specific characteristics are lacking.
ChatGPT as Investment Adviser: Experiments and Recommendations
Parallel to the debate over market indicators, German financial media tested ChatGPT as an investment adviser in the course of 2023 and 2024. In December 2025, the AI presented concrete individual stock recommendations – with explicit notice that these do not constitute purchase recommendations:
- Microsoft (NASDAQ: MSFT): The cloud and AI platform with Azure and Copilot has strong financial resources. ChatGPT justifies its mention with the broad, high-margin platform and ongoing AI leverage. The price-to-earnings ratio is above 30; the AI does not provide valuation details.
- ASML (EURONEXT/NASDAQ: ASML): The Dutch semiconductor equipment manufacturer holds a quasi-monopolistic position in EUV lithography for cutting-edge chips. ChatGPT characterizes ASML as "picks and shovels in the semiconductor cycle" with high technological barriers to entry. Criticism: extreme volatility and strong dependence on the chip cycle.
- Nestlé (SIX: NESN): The defensive consumer goods portfolio offers stable dividends and global brand positioning. Criticism: the stock has stagnated for years, margins are under pressure, growth is modest.
- LVMH (Euronext Paris: MC): The leading luxury conglomerate with over 70 brands benefits from structural growth in the luxury segment, strong brands and high margins, according to ChatGPT.
ChatGPT emphasizes that it cannot provide personal investment advice. The AI offers "general standard building blocks" for long-term portfolios. Observers criticize that ChatGPT tends to rely on well-known names rather than undervalued positions and ignores valuation details.
Experiments with Mixed Results
An experiment from June 2023 tested ChatGPT's ETF selection in the AI sector. Over roughly three weeks, the portfolio achieved around six percent returns. However, experts who took positions on AI stock ETFs as early as April 2023 warned against euphoria. They compared the enthusiasm to earlier waves of speculation such as the GameStop hype and warned of possible significant corrections after meteoric rises.
After ChatGPT gained prominence in November 2022, several ETF providers, including WisdomTree, developed AI-focused funds. The wave of enthusiasm for the AI sector began immediately after the launch and continues to this day.
Market Environment and Technical Context
On October 2, 2026, the S&P 500 was trading at 7,722.72 points, up 0.73 percent. The NASDAQ Composite stood at 27,190.86 points (+1.19 percent), the DAX at 25,273 points (+1.16 percent). These levels provide current context for Harbor Capital's argument that traditional indicators have not adequately captured market developments since November 2022.
Critical Assessment: Data and Interests
Harbor Capital Advisors presents a clear thesis: ChatGPT's launch fundamentally changed market dynamics. However, the central claim – the divergence between PMI and S&P 500 after November 2022 – has not been independently confirmed to date. The firm simultaneously sells products that benefit from this argument. The 800VDC AI Datacenter Ecosystem ETF positions itself as a solution to a problem that Harbor itself diagnosed.
Whether traditional indicators have actually lost their predictive value or whether the period since late 2022 represents a cyclical anomaly cannot be conclusively determined with the available data. The investment estimate of roughly one trillion U.S. dollars for AI infrastructure through the end of the decade comes exclusively from Harbor Capital itself.
ChatGPT as an investment adviser remains an experimental tool. The AI offers standard recommendations with established names, however, ignores valuation levels and provides no analysis of risk-return profiles. For retail investors in the German-speaking markets: AI tools can help with information gathering, but do not replace your own analysis or professional advice.