Product updates, market insights, and trading analysis

Across 311 public outcomes, signals scoring 50+ had a 55.1% win rate versus 36.6% below 50. From March to June, the eligible cohort shrank while its resolved win rate rose.

TradeHorde’s first 112 resolved public forecasts show a clear early pattern: high-conviction signals outperform, low-conviction signals lose money in aggregate, and filtering weak trades may be the real edge.

Most market regime detection systems measure the wrong thing. Here's what 1,300+ scored predictions against $SPY taught us.

Multi-model market analysis that generates forecasts with conviction scores tracked to outcomes.