ETF折溢价可以预测其未来收益吗?——基于错误定价的视角

Can ETF s Premium Predict Its Future Returns?Based on a Perspective of Mispricing

本文通过套利均衡模型刻画了ETF折溢价的形成过程,分析了ETF折溢价预测未来ETF收益的条件,并利用投资组合和Fama-Mecbeth回归检验了ETF折溢价对未来ETF收益的可预测性。研究结果显示:ETF折溢价是噪声冲击和投资者对信息反应不足导致的错误定价,这种错误定价主要发生在ETF层面,错误定价的修正表现为折溢价对未来ETF收益有负向预测作用,溢价率越高,未来ETF收益越低,投资者利用多空对冲策略每周可以从ETF错误定价修正中获得0.94%的收益。同时,ETF折溢价对未来ETF收益预测具有持续性,平均而言,ETF折溢价对未来ETF收益的预测可以持续4周,意味着市场完全修正ETF错误定价至少需要4周时间。本文进一步从套利限制的角度解释了ETF折溢价对未来收益预测的持续性,发现套利限制越大,折溢价对ETF未来收益预测的持续性越强,投资组合获得的收益也越高。

The paper depicts the formation process of ETF premium through an arbitrage equilibrium model to analyze a condition for predicting future ETF returns based on ETF premium,and examines the predictability by the method of portfolio analysis and Fama-Mecbeth regression.We find that ETF premium is a mispricing from noise shocks and investor underreaction to information,the mispricing occurs mainly at the ETF level,which implies that ETFs are much more likely to be mispriced than their underlying assets.The mispricing correction on ETF will display a negative return predictability that the future ETF return will decrease as the premium on ETF increases,and investors can gain 0.94%per week from ETF mispricing corrections by a long-short hedging strategy.Meanwhile,the ETF premium has a persistent predictive effect on future ETF return.On average,the persistence of return predictability can last for 4 weeks,which indicates that it takes at least 4 weeks for the market to fully correct ETF mispricing.In addition,we further explain the persistent predictability of ETF premiums on future returns from the perspective of arbitrage restriction,and we find that the greater the arbitrage restriction is,the longer the effect last and the higher the returns is.