Up next
Author
Share article
The post has been shared by 0 people.
Facebook 0
X (Twitter) 0
Pinterest 0
Mail 0

📊 Full opportunity report: Introducing Forezai · TradingAgents — a committee of LLMs decides paper-trades on ThorstenMeyerAI.com — validation score, market gap, and execution plan.

TL;DR

Forezai · TradingAgents is a new project that operationalizes a multi-LLM committee to simulate trading decisions on paper. It extends prior research on parametric strategies, focusing on AI collaboration rather than prediction accuracy. The system aims to test whether LLMs can outperform random decisions in simulated trading environments.

Forezai · TradingAgents has introduced an operational version of a multi-LLM committee designed to generate paper-trading decisions, marking a significant step in AI-driven trading research.

The project is a fork of the existing TradingAgents framework, which uses specialized large language models (LLMs) to analyze market data through multiple roles, including analysts, debate agents, and decision-makers. Unlike previous research that focused on parametric strategies, this system automates the entire decision process, including scheduling trades, managing positions, and logging results.

It incorporates an autonomous loop that runs daily, executing simulated trades via multiple modes, including a local Python broker and a sandboxed Alpaca paper-trading account. The system features a web dashboard for monitoring performance metrics such as equity curves, win rates, and trade reasons. Crucially, it is designed to prevent real money trading unless explicitly overridden, emphasizing research rather than live deployment.

Introducing Forezai · TradingAgents — Thorsten Meyer AI
AGENTS
● ANNOUNCEMENT / MAY 2026
THORSTEN MEYER AI · FOREZAI · § 03
FOREZAI · 03
TRADINGAGENTS · LAUNCH
Research Series · Companion to Polybot Week 1-2 · 2026-05-17

Introducing Forezai · TradingAgents.
A committee of LLMs
decides paper-trades.

After two weeks of finding out most parametric strategies don’t work, the obvious next research question: can multi-agent LLM judgment do any better?
A fork of the open-source TradingAgents framework (TauricResearch): thirteen LLM agents in four stages — four parallel analysts · a bull-bear debate with research-manager arbitration · a three-voice risk team · a two-layer trader + portfolio-manager decision. The fork keeps the agent graph intact and adds the operational layer the upstream doesn’t ship: an autonomous loop · a multi-broker abstraction · a local web dashboard · Codex OAuth · MCP plug-ins · 520+ unit tests. The question is narrower than “do LLMs predict the market” — that prior is “no, with high confidence.” The narrower question is: when LLMs are structured into specialised adversarial roles, does the committee produce decisions at least no worse than a coin flip after fees? Honest priors before running: it might fail too. If it appears to work, the most likely explanation is variance.
Thorsten Meyer AI Forezai · TradingAgents Apache-2.0 fork · upstream cited Companion piece · ~2,500 words Polybot · § 03
This is not financial advice. Nothing in this announcement should be used to inform real trading decisions. The software described trades simulated money by default. If you reconfigure it to trade real money, you should expect to lose that money — regardless of how clever any individual agent’s reasoning looks. Algorithmic trading is zero-sum after fees and structurally hostile to part-time retail strategies.
13 agents
Specialised roles in four stages
Analysts · Debate · Risk · Decision
78% / -33%
Polybot prior: fleet win rate
combined with -33% bankroll
520+
Passing unit tests across engine,
services, HTTP routes (starting baseline)
€0 floor
LLM cost on Codex OAuth
(falls back to public API per token)
FOREZAI / TRADINGAGENTS· APACHE 2.0 FORK· UPSTREAM TAURIC RESEARCH· LANGGRAPH· 13 AGENTS / 4 STAGES· 4 PARALLEL ANALYSTS· BULL-BEAR DEBATE· 3-VOICE RISK TEAM· TRADER + PORTFOLIO MANAGER· 5-TIER FINAL RATING· ALPACA PAPER + LOCAL + SHADOW· LIVE ENDPOINTS HARD-REFUSED· FASTAPI + REACT VIA CDN· CODEX OAUTH· MCP PLUG-IN REGISTRY· 520+ UNIT TESTS· POLYBOT WEEK 1: 21 EXPERIMENTS· WEEK 2: -33% BANKROLL· 78% FLEET WIN RATE· HONEST RESEARCH, NOT EDGE· FOREZAI / TRADINGAGENTS· APACHE 2.0 FORK· UPSTREAM TAURIC RESEARCH· LANGGRAPH· 13 AGENTS / 4 STAGES· 4 PARALLEL ANALYSTS· BULL-BEAR DEBATE· 3-VOICE RISK TEAM· TRADER + PORTFOLIO MANAGER· 5-TIER FINAL RATING· ALPACA PAPER + LOCAL + SHADOW· LIVE ENDPOINTS HARD-REFUSED· FASTAPI + REACT VIA CDN· CODEX OAUTH· MCP PLUG-IN REGISTRY· 520+ UNIT TESTS· POLYBOT WEEK 1: 21 EXPERIMENTS· WEEK 2: -33% BANKROLL· 78% FLEET WIN RATE· HONEST RESEARCH, NOT EDGE·
FIG. 01 — THE 13-AGENT COMMITTEE
Thirteen specialised roles · four stages · biases made to argue in public
The architecture forces the system to articulate its reasoning rather than relying on what a single context window happens to recall
Stage 1 · Four analysts in parallel4 agents
Market
Structure, ranges, regime indicators
News + Insider
News flow, filings, insider activity
Fundamentals
Balance sheet, earnings, ratios
Social Sentiment
Social-media tone, retail signal
↓
Stage 2 · Bull-bear debate + research-manager arbitration3 agents
Bull researcher
Argues upside thesis from analyst reports
Bear researcher
Argues downside thesis from same reports
Research manager
Arbitrates · writes single synthesis
↓
Stage 3 · Three-voice risk team3 agents
Aggressive
Looks for upside · accepts variance
Conservative
Looks for downside · protects capital
Neutral
Balances · forces downside articulation
↓
Stage 4 · Two-layer decision2 agents
Trader
Three-tier proposal · buy / hold / sell
Portfolio manager
Five-tier rating + price target + horizon · sees arguments only, never raw data
The portfolio manager only sees the arguments, never the raw data — which forces the committee to make its reasoning explicit rather than relying on a single context window’s recall. The upstream framework ships the agent graph; it does not ship the operational machinery to run that graph on autopilot, observe its results honestly, store them for later inspection, or prevent the operator from accidentally trading real money. That gap is what the Forezai fork fills.
FIG. 02 — THE POLYBOT PRIOR · WHY THIS IS A DIFFERENT BET
Two weeks of paper-trading prediction markets · the trap underneath the headline numbers
25 experiments · 78% fleet-wide win rate · -33% bankroll · most parametric strategies are structurally negative-expectation when measured honestly
The flattering number
78%
Fleet-wide win rate · week 2
“You can win four out of five trades and still go broke, because the one loss is bigger than the four wins put together.” Win rate without P&L context is a mechanical illusion.
The honest number
−33%
Fleet bankroll · week 2 close
The strongest possible demonstration of the trap. A parametric trading strategy that looks compelling in a backtest will almost always fail to survive a fresh sample. Most “edges” are mechanical artefacts.
Week 1: 21 parallel strategy experiments · early winners mostly mechanical illusions · exactly one strategy (a fair-value taker on BTC) showed the mathematical signature of real edge over a few hundred settled trades. Week 2: same fair-value strategy with more data collapsed. A separate mid-week hypothesis (market-making) also failed cleanly. Fleet ended week 2 at roughly negative thirty-three percent of bankroll. The honest research finding wasn’t on the winning side — it was on the losing side. Adding more parameters to Polybot wouldn’t change that. TradingAgents is asking a separable question.
FIG. 03 — WHAT THE FORK ADDS · THE OPERATIONAL LAYER
Six layers the upstream framework doesn’t ship
Same agent graph, intact. The fork makes it a research instrument rather than a tech demo.
01 · Loop
An autonomous loop
Scheduler · watchlist · auto-trader maps ratings to paper orders · allow-list filtering · per-ticker cooldowns · sector caps · cash checks · position manager evaluates open positions every 60s for TP / SL / max-hold. Append-only audit logs.
02 · Brokers
Multi-broker abstraction
Three modes: local Python broker (yfinance fills, JSON-persisted) · Alpaca paper-trading adapter · “shadow” mode running both in parallel with divergence view. Real Alpaca live endpoints are hard-refused at multiple layers.
03 · Dashboard
A local web dashboard
FastAPI backend · React via CDN, no Node toolchain · SVG equity curve · rolling-peak drawdown · win-rate by rating / ticker / model · exit-reason breakdown · LLM cost vs realised P&L joined by run ID. Runs locally; nothing sent to a cloud service.
04 · Codex
Codex OAuth
Runs the engine on a ChatGPT Pro subscription via the Codex backend. LLM cost floor effectively zero if you already have ChatGPT Pro. Token stored encrypted locally. Falls back to the regular OpenAI API if you’d rather pay per token.
05 · Alerts
Multi-channel alerts
Slack · Discord · SMTP email · configurable filter on rating events and order fills · append-only history kept locally. Webhook URLs masked in API responses so a screenshot can’t accidentally leak credentials.
06 · MCP
MCP plug-ins
Registry for adding Anthropic Model Context Protocol servers (Kensho · Aiera · FactSet · Morningstar · LSEG) as analyst tools. Plug-ins advertise category (fundamentals · news · market data · social) · probe endpoint tests credentials.
Honest-by-design touches: every generated report prepends “Research, not advice” and appends a footer with version, commit, provider, models used, run ID, and cost. Closed trades carry the same metadata. 520+ passing unit tests across engine, services, and HTTP routes. The intent: when the system loses money, the journal makes it impossible to pretend it didn’t.
FIG. 04 — HONEST PRIORS · BEFORE RUNNING THIS IN ANGER
Three priors stated before the data starts arriving
The bias of the project: when the data says no, the dashboard says no, the article says no
1
It might fail too. LLMs are not oracles, and a sophisticated framework around language-model outputs does not change the underlying error rate of the model. Sample is still everything. The framework’s outputs are subject to the same statistical noise as any prediction system over small samples.
Highest likelihood
2
If it appears to work, the most likely explanation is variance. The same trap that caught the first article’s candidate edge applies here. A high win rate over fifty trades means much less than it looks. Without out-of-sample confirmation, a flattering early sample tells you almost nothing about whether the system has real edge.
Second-most likely
3
If it appears to work for the right reasons — empirical win rate matches stated confidence, and alpha-versus-benchmark persists across non-overlapping samples — that would be a meaningful research finding. Whether that happens, I don’t know. The point of putting it in the open is that the data will say.
Genuinely open
This is explicitly not a launch announcement for a product anyone should connect a real brokerage account to. The Alpaca live endpoints are hard-refused at multiple layers in the code, and the design choice is deliberate. The right next step is data, not deployment. The bias of the whole project is straightforward: when the data says no, the dashboard says no, the article says no, and no one tries to retroactively rescue the thesis. That’s the contribution.
FIG. 05 — WEEK THREE · WHAT THE METHODOLOGY WILL MEASURE
Four concrete measurements before publishing findings
The hope: write the week-three article from a position of “here’s what the data says”. The fear: another candidate falsified at higher sample. Both outcomes are publishable.
M1 · Sample discipline
Small watchlist for a few weeks before publishing
A handful of tickers across two or three sectors. Long enough to gather sample, narrow enough to keep attention on what’s actually happening per agent. Avoid the noise of a 65-ticker autonomous loop until the smaller version has been read carefully.
M2 · Calibration view
Stated confidence vs. realised win rate
When the system says “75% confident”, do the trades actually win 75% of the time? Same measurement applied to Polybot’s fair-value model. If the model is systematically over-confident, that bias dominates everything downstream.
M3 · Cost accounting
Cost per ticker · per rating · per profitable trade
With Codex OAuth the marginal LLM cost is effectively zero. With the public OpenAI API, each run is hundreds of agent turns. The honest question: does this scale economically if you ever did run it at real cost?
M4 · Non-overlapping windows
Alpha vs benchmark · out-of-sample
Not within-sample alpha — trivially inflatable. Hold out one period entirely, run the system on the next, then check whether the held-out result matches the in-sample stats. If they diverge sharply, the in-sample was curve-fit.
Open under Apache-2.0 with upstream cited from every relevant surface. Not open: the operator’s running results, the specific watchlist, the per-agent prompt customisations, the alert channels, the trade journals — kept local for the same reason Polybot’s per-experiment data is kept local. Publishing exact configurations encourages people to copy them with real money, which is the opposite of what an honest research project should do. Summary findings will be published. Recipes will not.
The bet is on a different mechanism, not a different parameter setting. The point is not to find a money-printing AI. The point is to put honest measurements of these systems into the public record — so the next person looking at the space starts a step further along than the last.
Thorsten Meyer AI · Introducing Forezai · TradingAgents · § 03
Source dossier & project notes
  • Forezai · TradingAgents fork — Apache-2.0 license preserved · GitHub published under same author as Polybot · upstream cited from README · marketing pages · generated report footers · forezai.com
  • TradingAgents upstream — TauricResearch team · multi-agent stock-research framework on LangGraph · the agent graph is intact in the fork · upstream license requirements (notice file · attribution · patent-grant clauses) preserved if anyone forks again
  • Polybot · Week 1 — 21 parallel strategy experiments on Polymarket 5-minute Up/Down · early winners mostly mechanical illusions · one BTC fair-value taker showed mathematical signature of edge over a few hundred settled trades
  • Polybot · Week 2 — fair-value strategy collapsed at higher sample · market-making hypothesis failed cleanly · fleet ended at ~negative 33% of bankroll · 78% fleet-wide win rate combined with deeply negative P&L · “you can win four out of five trades and still go broke”
  • Architecture · Stage 1 — Four analysts in parallel: market structure · news + insider · fundamentals (balance sheet + earnings) · social-media sentiment · each writes a short structured report independently
  • Architecture · Stage 2 — Bull-bear debate: two researcher agents argue opposing theses from the analyst reports · a research-manager agent arbitrates and writes a single synthesis
  • Architecture · Stage 3 — Three-voice risk team: aggressive (upside, accepts variance) · conservative (downside, protects capital) · neutral (balances) · forces explicit articulation of downside before any order is proposed
  • Architecture · Stage 4 — Two-layer decision: trader agent produces three-tier proposal (buy / hold / sell) · portfolio-manager agent synthesises into final five-tier rating with price target and time horizon · PM sees arguments only, never raw data
  • Operational layer — Autonomous loop · multi-broker abstraction (local Python / Alpaca paper / shadow mode) · FastAPI + React-via-CDN web dashboard · Codex OAuth on ChatGPT Pro · multi-channel alerts (Slack · Discord · SMTP) · MCP plug-in registry (Kensho · Aiera · FactSet · Morningstar · LSEG)
  • Engineering baseline — 520+ passing unit tests across engine, services, HTTP routes · honest-by-design touches: “Research, not advice” prepended to every report · footer with version, commit, provider, models used, run ID, cost
  • Hard-refused live trading — Real Alpaca live endpoints hard-refused at multiple layers · operator must deliberately override the refusal in more than one place to actually risk real money
  • Week three methodology — Small watchlist · calibration view (stated confidence vs realised win rate) · cost accounting per ticker / rating / profitable trade · out-of-sample alpha across non-overlapping windows
  • Stance — Summary findings published · recipes kept local · the bias is straightforward: when the data says no, the dashboard says no, the article says no, no one tries to retroactively rescue the thesis
Colophon

Set in Source Serif 4 (display, italic accent), EB Garamond (body), IBM Plex Sans (UI labels), IBM Plex Mono (mastheads, ticker, stamps, plug-in tags). Paper-cool gray-cream #e6e7e4.

Chromatic register: structural-slate dominant (committee-architecture analysis), empirical-clay for the Polybot prior forensic, labor-rose for the cautionary “expect to lose money” stance and the regulatory disclaimers, transition-bronze for the week-three forward-shape methodology, alternative-sage for the open-source / honest-measurement positive signal.

Key frame:

FOREZAI / TRADINGAGENTS 13-AGENT COMMITTEE POLYBOT PRIOR OPERATIONAL LAYER EXPECT TO LOSE MONEY CALIBRATION VIEW BULL-BEAR DEBATE HARD-REFUSED LIVE APACHE 2.0 HONEST MEASUREMENTS

Implications of AI-Driven Multi-Agent Trading Systems

This development demonstrates a move toward operationalizing AI research in finance, moving beyond theoretical analysis to practical experimentation with autonomous trading systems. It highlights how structured multi-LLM committees can be used to generate, justify, and potentially improve trading decisions without relying on traditional predictive models.

While the system currently trades on simulated data, its design provides a foundation for future research into AI collaboration, transparency in decision-making, and the potential for AI to assist human traders or manage portfolios in more complex environments.

Chemical Process Simulation and the Aspen HYSYS Software

Chemical Process Simulation and the Aspen HYSYS Software

Used Book in Good Condition

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Background on AI and Algorithmic Trading Research

Previous research, notably by Thorsten Meyer and the TauricResearch team, revealed that parametric, rule-based trading strategies often fail to survive fresh market data, despite promising backtest results. This led to questions about whether less rule-bound, more collaborative AI approaches could do better.

The TradingAgents framework was initially designed to explore whether LLMs, structured as specialized roles, could produce decisions that are at least no worse than random, by forcing explicit reasoning and debate. The new Forezai fork extends this research by adding operational features that enable real-time, automated paper trading, making it a practical testbed for AI collaboration in market decision-making.

“This project moves the research from theoretical analysis to an operational environment, allowing us to observe how AI committees perform in simulated trading.”

— Thorsten Meyer

The Intelligent AI Investor: A Beginner’s Guide to Using AI Tools for Informed Investment Decisions, Risk Management, and Wealth Building (Trading & Investing Series Book 7)

The Intelligent AI Investor: A Beginner’s Guide to Using AI Tools for Informed Investment Decisions, Risk Management, and Wealth Building (Trading & Investing Series Book 7)

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Unanswered Questions About AI Trading Effectiveness

It remains unclear whether the committee-based approach will outperform traditional algorithms or human traders in live or extended simulation environments. The current system is limited to paper trading, and its long-term robustness and adaptability are still untested.

Furthermore, the impact of model biases, the quality of debate among agents, and the potential for overfitting to specific market conditions are areas requiring further investigation.

Advanced Stock Market Analysis Dashboard: Design and Implementation of a Real-Time Financial Analytics Platform

Advanced Stock Market Analysis Dashboard: Design and Implementation of a Real-Time Financial Analytics Platform

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Next Steps for Testing and Development

Future work will include extended backtesting across different market regimes, refining the agent roles and debate structures, and potentially integrating live trading with safeguards. Researchers will also analyze the decision rationale logs to evaluate the reasoning quality and identify areas for improvement.

Additionally, the team plans to open the system for broader testing and gather feedback on its decision-making transparency and performance in diverse simulated environments.

Amazon

automated trading strategy backtester

AmazonView Latest Price

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

Can this system be used for real trading now?

No, the current setup is limited to paper trading and is intended for research purposes. It is explicitly designed to prevent unintended real-money trading unless deliberately overridden by the operator.

How does the LLM committee make decisions?

The system involves multiple specialized roles that analyze data, debate opposing views, and synthesize a final decision. This structure forces explicit reasoning and debate rather than relying on single-model predictions.

What advantages does this multi-LLM approach offer?

It aims to improve decision robustness by incorporating diverse perspectives and explicit reasoning, potentially reducing biases and overfitting common in rule-based strategies.

Is this system intended to replace human traders?

Currently, it serves as a research tool to explore AI collaboration in trading decisions. Its goal is to inform future AI-assisted trading or portfolio management, not to replace human judgment outright.

What are the main limitations of this system?

Its effectiveness in live markets remains unproven, and it depends on the quality of the LLMs and the structure of debates. It also does not account for real-time market changes or unforeseen events beyond its simulation environment.

Source: ThorstenMeyerAI.com

You May Also Like
The CFO’s new operating system. Anthropic, OpenAI, and the consulting margin that just got compressed.

The CFO’s new operating system. Anthropic, OpenAI, and the consulting margin that just got compressed.

Anthropic and OpenAI are moving from model sales to integrated operating systems for CFO workflows, backed by PE investments and strategic alliances.
The Anthropic IPO Disclosure Document: What the S-1 Has to Say Before October

The Anthropic IPO Disclosure Document: What the S-1 Has to Say Before October

Ahead of its October Nasdaq listing, Anthropic’s S-1 reveals critical financial and operational details, including revenue recognition disputes and regulatory risks.
The mandate. Why the US conversational- finance surface does not translate to Europe.

The mandate. Why the US conversational- finance surface does not translate to Europe.

The US launches permissionless personal-finance surfaces, while Europe mandates licensing and consent, fundamentally changing market architecture.
The bank account in the chat. How personal finance became an agentic on-ramp.

The bank account in the chat. How personal finance became an agentic on-ramp.

OpenAI launched a preview of personal-finance tools in ChatGPT, enabling users to connect bank accounts and access real-time financial data, signaling a shift in consumer finance.