The Challenge

How this experiment works

Overview

Three AI models — Claude, ChatGPT, and Gemini — each manage a real $2,000 investment portfolio for one year. Every Monday they independently research current market conditions, decide how to allocate their portfolio, and execute real trades via a trading API.

Their performance is tracked against a passive benchmark portfolio. Every trade, decision, and line of reasoning is public.

Rules

  • 01Each AI starts with $2,000 in a real brokerage account
  • 02Trades are made once per week, every Monday at market open
  • 03Any US-listed security is fair game: stocks, ETFs, short positions
  • 04No single position may exceed 40% of the portfolio
  • 05Cash is a valid allocation
  • 06Each AI receives the same market data and prompt every week
  • 07Each AI also receives its own prior 4 weeks of decisions as memory
  • 08The experiment runs for one full year

The Competitors

Claude

Anthropic

claude-opus-4-7

ChatGPT

OpenAI

gpt-5.5

Gemini

Google

gemini-2.5-pro

The Benchmark

A passive portfolio that makes no active decisions, rebalanced monthly:

65%VTI — US Total Market
25%VXUS — International Stocks
10%BND — US Bonds

The Prompt

Every Monday, each AI receives this exact prompt with live data injected:

You are an active portfolio manager competing against other investors over a 1-year period. Your sole objective is to generate the highest possible total return — to outperform your competitors. You are judged on results, not process. You have full discretion: you may take concentrated positions, rotate aggressively, short securities, or hold cash. There is no prize for playing it safe. A portfolio that simply tracks broad market indexes will not win this competition. You are managing a real portfolio with $[TOTAL_VALUE] in total assets. Current date: [DATE] Current holdings: [SYMBOL] | [SHARES] shares | $[VALUE] | [%] Cash available: $[CASH] Market context (last 7 days): - S&P 500 (SPY): [YTD]% YTD, [WEEK]% this week - NASDAQ (QQQ): [WEEK]% this week - VIX: [VIX] - Notable events: [SUMMARY] [Prior 4 weeks of your own decisions, if available] Your task: 1. Assess the current market environment and what it means for your strategy 2. Decide your target portfolio allocation — think like an active manager seeking alpha 3. You may: buy any US-listed stock or ETF, short stocks, buy/sell options 4. Hard rule: no single position may exceed 40% of portfolio 5. You may hold cash if you see no good opportunities Return ONLY valid JSON: { "market_outlook": "2-3 sentence market assessment", "reasoning": "3-5 sentences explaining your alpha thesis", "target_allocations": [ { "symbol": "SPY", "pct": 40 }, { "symbol": "CASH", "pct": 60 } ] }

How Trades Work

Each AI returns a target allocation as a list of symbols and percentages. The system then:

  1. 1.Validates the response (allocations sum to ~100%, no position exceeds 40%)
  2. 2.Compares target allocations to current holdings
  3. 3.Calculates the required buys and sells to reach the target
  4. 4.Executes notional (dollar-based) orders via the trading API — e.g. "buy $400 of AAPL"
  5. 5.Logs every trade with the AI's stated reasoning, publicly visible here

Trades are market orders executed at open. If an AI returns malformed JSON, the system holds the current portfolio and logs the failure. Brokerage and execution handled via a trading API.