AI agents beat Wall Street on 60 of 90 earnings forecasts, averaging 2.5% misses vs 2.9%
rohanpaul_ai · x · 2026-10-12
- OpenStocks, built by Ruggero Gargiulo, replaces conflicted sell-side analysts with AI agents that have no clients to please, publishing their scores against reported numbers next to Wall Street consensus.
- How it works: before each earnings report, agents predict three figures (revenue, profit, EPS); once results land, whoever is closer wins each number. Anyone can submit an agent and climb the public leaderboard.
- Track record: since scoring began Aug 4, the official AI consensus (built from a single vetted agent) has won 60 of 90 reports, including 17 of the latest 20, with a typical miss of 2.5% vs Wall Street's 2.9%.
- The author argues investors will increasingly use an AI consensus alongside Wall Street's — if the public track record earns their trust.
Related event: OpenStocks Pits AI Agents Against Wall Street in Earnings Forecasts(3 posts)→
More from Venture
- a16z speedrun SR008 opens: up to $1M investment plus $10M+ in credits — andrewchen · 2026-10-12
- Palantir positioned to win as open-weight models go mainstream: Baird — eliano · 2026-10-12
- Chinese tech firms historically unwilling to pay strategic premiums for M&A, says Rui Ma — pstAsiatech · 2026-10-12
- Perplexity CEO: $750M annualized revenue, AI-native SMBs fueling new millionaire boom — AravSrinivas · 2026-10-12
- Stripe wants partners to co-design a 0-1 click pay-per-consumption flow — jeff_weinstein · 2026-10-12
- Study of 20,000+ results across 16 LLMs finds the #1 AI ranking factor: relevant brand mentions — VeryWellVersed · 2026-10-12