Self-exciting Threshold Auto-regressive Model For Evaluating Hedging Cost

In principle of no-arbitrage, price movements should best be described by a first order vector error correction model, with the error correction term being the price differential between spot and futures markets (the basis). Evidence from Chinese markets suggests that basis series follow a multi-order auto regressive model, which means that there are persistent arbitrage opportunities than should be in functioning markets. With the data base of Chinese copper futures market, using self-exciting threshold autoregressive (SETAR) model, we analyze whether such dynamics can be related to variable arbitrage cost. Our findings reveal that arbitrage opportunities are decreasing with the development and effectiveness of Chinese markets. Furthermore, a simple and general scheme is presented for establishing SETAR model to evaluate hedging cost. With the improved algorithm, both of threshold values and autoregressive coefficients may be optimized. The empirical research shows the scheme was practical and efficient.