See how AI reads public markets
We expose Large Language Models to global public market data every day and publish the raw research output. Stocks, ETFs, crypto, commodities. Every research entry is tracked publicly so you can study the process, the hits, and the misses.
Research output, not investment advice. Always do your own work.
Why study the research here?
Transparency first. Every research output is tracked publicly.
Public Research History
Every research subject is logged with entry reference, scenarios, and observed outcomes. Hit rate, average observed delta, all visible to everyone.
AI That Learns
The agent keeps a memory log. It learns from every correct and incorrect hypothesis, improving its analytical pattern recognition over time.
Multi-Asset Coverage
Stocks, ETFs, crypto, and commodities. Global public data covering 236 subjects across US, Europe, Asia, and LatAm.
No Conflicts Hidden
100% subscription-funded. No ads, no affiliate links, no sponsored content. The operator's own positions are disclosed publicly.
Daily Research
New research output every morning. Weekly digest with top movements and new analysis delivered to subscribers.
Skin In The Game, Disclosed
The operator may hold positions in the subjects studied. This is fully disclosed so you can judge the research with full context.
How it works
Data Collection
The agent pulls public data from 10+ sources daily. Market prices, macro indicators, news headlines, and more.
AI Analysis
The LLMs analyse patterns, flag subjects of interest, and assign a thesis strength to each output.
Published and Tracked
Every research entry is published with entry reference and scenario analysis. Daily tracking begins automatically.
Learn and Improve
Outcomes are recorded. The memory log helps the agent avoid past mistakes and reinforce patterns that worked.
Why not just use ChatGPT or Claude yourself?
LLMs are powerful, but prompting them manually has real limits. Here's what an autonomous research agent does that a chat window cannot.
| ChatGPT / Claude | Observed Markets | |
|---|---|---|
| Runs daily without prompting | ✕ | ✓ |
| Real market data (not hallucinated) | ✕ | ✓ |
| Public, verifiable research history | ✕ | ✓ |
| Learns from its own past hypotheses | ✕ | ✓ |
| Monitors 236 subjects automatically | ✕ | ✓ |
| Sends you weekly digests and alerts | ✕ | ✓ |
| No prompt engineering required | ✕ | ✓ |
| General knowledge and chat | ✓ | ✕ |
An LLM gives you an answer when you ask. Observed Markets runs a research pipeline every day, whether you ask or not.
A framework, not a stock tip service
The real work here is a framework for transparent, accountable AI-driven research. We chose public markets as the testing ground because the data is public and the outcomes are measurable — every hypothesis can be checked against reality in days or weeks, not years.
The same methodology generalises to any domain where inputs are observable and outputs can be scored. Markets are the stress test. Transparency is the product.
What we study
We don't track the entire market. We cover a curated universe of globally relevant public assets. Broad enough to surface interesting patterns, focused enough to analyse properly.
30
Indexes
S&P 500, DAX, Nikkei, Bovespa, Sensex, Hang Seng…
60+
ETFs
Broad market, sectors, bonds, commodities, thematic
120+
Stocks
US, Europe (ASML, LVMH, SAP), Asia (TSM, Sony, Samsung), LatAm
8
Crypto
BTC, ETH, SOL, BNB, XRP, ADA, AVAX, DOT
14
Commodities
Gold, oil, natural gas, copper, wheat, coffee…
Data from Yahoo Finance, FRED (US macro) and ECB (EU macro). Prices updated daily at 06:00 UTC.