Is Trend Still Your Friend? A Microstructural Account of the Demise of Short-Term Trend-Following

Battle Count: 84

This paper is directly and fundamentally relevant to quantitative trading. It addresses the central question of why short-term trend-following strategies (a core systematic trading strategy) have failed since 2008/9, provides a novel microstructural explanation based on volatility-normalised tick size, and offers practical implications for CTA strategy design, capacity estimation, and execution. The findings directly impact how quantitative traders should allocate across contracts, design execution strategies, and understand the structural limits of trend-following in modern electronic markets.

The core methodology (EWMA signal construction, volatility normalization, tick-size tiering, portfolio simulation) is conceptually straightforward and implementable with standard financial data infrastructure. However, the full analysis requires: (1) daily settlement data for ~100 futures contracts spanning 30 years, (2) intraday 5-minute bar data with order-book snapshots for ~53 contracts, (3) CTA industry AUM data, (4) SG CTA Index data, and (5) electronification data. The microstructural diagnostics (book/trade imbalance correlations, price-impact analysis) require high-quality intraday data. The tick-size tiering procedure is simple but requires careful causal implementation.

The paper provides detailed methodology for signal construction, portfolio weighting, and tick-size tiering. However, key data sources are proprietary: 5-minute bar data from Capital Fund Management (CFM), SG CTA Index from Société Générale Prime Services, and CTA AUM from BarclayHedge. The daily settlement data for 101 futures contracts is described but not publicly linked. The electronification data is courtesy of Robert Almgren (Quantitative Brokers). Reproduction would require access to similar proprietary datasets.