Skip to content
All library documents

AI Equity Basket Rotation Using Leaders, Followers, News, and Price Statistics

Article Strategy library · Author: ianzeng123
Read the original

Summary

This system studies rotation within AI-related equity baskets traded through USDT perpetual contracts. It screens for equity contracts, uses an LLM to assign companies to one or more AI subsectors, and tracks relative returns and Z-scores to identify unusually strong basket leaders. It then checks historical candle data for leader-to-follower correlations, follow rates, and lag periods. When a leader breaks out, recent news is gathered and assessed by an LLM; only signals with a plausible news explanation, matching direction, and a validated statistical relationship can target a lagging follower. The system supports notification-only or automated trading, with hard and trailing stops.

The document describes a detailed workflow, not demonstrated profitability: it gives no backtest results or measured signal quality. It acknowledges risks from incorrect model classifications, noisy or late news, unstable historical relationships, contract funding and liquidity, and followers that fail to catch up. Account-level exposure controls and execution safeguards also remain areas for improvement, so the system is framed as a research tool requiring observation and review.

Key ideas

  • The system builds overlapping AI subsector baskets from screened equity perpetual contracts using LLM-generated company profiles.
  • Relative returns and Z-scores identify basket leaders, while historical candle statistics test whether followers tend to react later.
  • News is sought after a price breakout and serves as confirmation alongside the statistical relationship before a follower signal is allowed.
  • Notification-only and automated trading modes are available, with hard stops and profit-protecting trailing stops.
  • No profitability results are presented, and model errors, noisy news, unstable relationships, funding, and liquidity can undermine signals.

Tags

EquitiesFuturesMachine learningStatisticsMomentumRisk management

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.