
Old Market Indicators Useless Since ChatGPT Launch? Fund Manager Launches ETFs for New Metrics
This article was created with the help of artificial intelligence.
Key Takeaways
- Kristof Gleich of Harbor Capital Advisors argued at a press briefing that traditional economic indicators have failed to reflect the stock market since ChatGPT's launch in November 2022.
- Harbor Capital Advisors launched the 800VDC AI Datacenter Ecosystem ETF (NYSE ARCA: HUMM) with a net expense ratio of 0.49 percent in response to the alleged decoupling of market indicators.
- OpenAI released ChatGPT on November 30, 2022, and reached 100 million users within two months.
- Spenser Lerner of Harbor Capital estimated planned infrastructure investments by hyperscalers and datacenter operators through the end of the decade at just under one trillion U.S. dollars.
- The Manufacturing Purchasing Managers' Index (PMI) moved largely sideways after November 2022 according to Harbor Capital, while the S&P 500 rose significantly, though 24/7 Wall St. could not independently verify this.
Kristof Gleich, president and chief investment strategist of Harbor Capital Advisors, claimed at a press briefing reported by news site 24/7 Wall St. on October 3, 2026, that traditional economic indicators on which investors have relied for decades have failed to reflect the stock market since ChatGPT's launch in November 2022. In response to this decoupling, the fund house launched the 800VDC AI Datacenter Ecosystem ETF (NYSE ARCA: HUMM) with a net expense ratio of 0.49 percent.
Manufacturing PMI Diverges from Market Since November 2022
Gleich presented the Manufacturing Purchasing Managers' Index (PMI) as evidence, a survey that measures the growth or decline of factory activity. Historically, the PMI had moved in parallel with the S&P 500, but has decoupled since November 2022: while factory indicators remained largely sideways, the market rose significantly. The news site 24/7 Wall St. noted that it could not independently verify this divergence and characterized it as Harbor Capital Advisors' argument.
According to Gleich, the AI economy requires "different frameworks, different benchmarks, new metrics, and a different language to describe, measure, and invest in it." The S&P 500 was trading at 7,722.72 points as of October 2, 2026, following a gain of 0.73 percent.
One Trillion in AI Infrastructure Investments by 2030
Spenser Lerner, head of Multi-Asset Solutions at Harbor Capital, divides the expected AI boom into phases. The first phase through the end of the decade focuses on building infrastructure. Lerner estimated the planned investments by hyperscalers and datacenter operators in AI infrastructure at just under one trillion U.S. dollars. The 800VDC AI Datacenter Ecosystem ETF was developed explicitly as a response to the conviction that traditional economic indicators no longer work in the post-ChatGPT economy.
ChatGPT Reached 100 Million Users in Two Months
OpenAI released ChatGPT on November 30, 2022, to the general public. The tool reached 100 million users within two months. Unlike earlier AI applications, users could directly ask questions and received clear, conversational answers instead of a list of links. Naseem Husain, Senior Vice President and ETF strategist at Horizons ETFs, explained in August 2026 that "with the launch of ChatGPT in late 2022, the true extent of its disruptive potential became clearer globally in 2023."
Commercial Interests Behind the Indicator Thesis
The sources explicitly point to commercial interests behind Harbor Capital's argument. While the claim about defunct indicators contains real analytical statements, there is a commercial motive: Harbor sells products based on this thesis. Investors, particularly those nearing retirement, should consider both the analytical argument and the commercial interest behind it.
Already in early 2023, AI-driven ETFs came to market. However, tests revealed limitations: ChatGPT was unable to identify existing AI ETFs such as AIEQ at that time, despite "AI Powered" being in the product name. The limitation was attributed to ChatGPT's "limited knowledge" of events after 2021.