When the Tail Wags the Dog: A Time‐Varying FCVAR Analysis of Bitcoin Market

ABSTRACT This paper examines how the relationship between Bitcoin spot and futures markets has evolved using a time‐varying Fractionally Cointegrated Vector Autoregressive (FCVAR) model. We are the first to apply this methodology dynamically to cryptocurrency markets, allowing us to simultaneously analyze long‐run equilibrium, pricing patterns, market efficiency, and price discovery as they change over time. We document three main results. Bitcoin futures dominate price discovery, driving 80% of permanent price movements and highlighting how regulated derivative markets lead information flow. The adjustment between spot and futures prices occurs slowly and persistently, showing long‐memory effects that suggest only partial market efficiency. Finally, while these markets typically maintain parity, we frequently observe contango during periods of high volatility, market optimism, or speculative activity. Our approach offers a comprehensive framework for understanding how digital asset prices form, providing valuable insights for market participants and regulators about the role of institutional infrastructure in cryptocurrency markets.