Limited Arbitrage,investor Sentiment And ETF Arbtrage Return

Arbitrage activities are always accompanied by the mispricing of financial assets and are the foundation for the pricing of financial derivatives.According to the “law of one price”,financial assets and their derivatives share fundamental information and cash flow,and the prices of the two should be consistent.Otherwise,arbitrageurs will buy low-priced assets and sell high-priced assets at the same time until prices of financial assets and their derivatives converge.As a financial derivative product,ETF is also priced based on this principle.The ETF and the underlying asset share the same fundamentals and are independently traded in their respective markets.When the ETF and the underlying asset react to market-related information differently due to the irrational behavior of investors,the price of the ETF and the underlying asset will deviate.Violating the law of one price,at this time,the entry of arbitrageurs will eliminate this mispricing and make the price of ETF consistent with the underlying asset.However,in the real capital market,perfect arbitrage is almost non-existent,and investors’ arbitrage activities are more or less restricted,leading to asset mispricing and it can not be eliminated in time.When ETFs or underlying assets are affected by non-fundamental factors,the ETF price deviates from the underlying assets,resulting in a discount or premium.Due to the existence of arbitrage restrictions,investors can not fully eliminate this mispricing in time.As time goes on,information investors enter the market,the mispricing of ETFs and underlying assets will be corrected,and the prices of ETFs and underlying assets will back to fundamental value.ETF discounts and premiums are the result of investor irrational behavior and arbitrage restrictions,reflecting the degree of mispricing of ETFs and underlying assets.On the one hand,the irrational behavior of investors leads to mispricing of ETFs or underlying assets,causing the ETF to deviate from the underlying asset prices,resulting in discounts and premiums.On the other hand,arbitrage restrictions prevent arbitrageurs from participating in ETF arbitrage,making ETF discounts and premiums impossible to be timely corrected.Facts have proved that the degree of arbitrage restrictions is changing all the time,and mispricing in the capital market will not exist forever.Arbitrage restrictions only allow smart arbitrageurs to delay correction of mispricing.This delayed correction of mispricing means the potential prediction of ETF return.But some research found that only the net purchase and redemption of ETFs has a predictability for ETF return,and the predictability of ETF discounts and premiums on future ETF return is not significant.Then,is the ETF discount and premium predictable for future ETF return in the Chinese A-share market? What is the internal mechanism of the ETF discount and premium to predict future ETF return? What is the impact of the explosive growth of ETFs on the Chinese financial market? Should the management vigorously develop derivative products such as ETFs?In order to answer the above questions,this thesis sorts out the reasons for the formation of ETF discounts and premiums and the relevant literature on future ETF return forecasts,combines theoretical models and logical reasoning to analyze the conditions and internal mechanisms of ETF discounts and premiums to predict future return.The predictability of discount and premium to future ETF returns,and the internal mechanism of discount and premium to predict future ETF returns from the perspectives of arbitrage restrictions and investor sentiment would be test through empirical analysis.In order to ensure the universal applicability of the research conclusions,cross-border ETFs in major global financial markets are taken as comparison objects,and the predictability of future return of ETF discounts and premiums is studied in different arbitrage restriction environments.As a supplement to the research,we further discussed the economic consequences of the continuous expansion of the number and scale of ETFs and explored the impact of ETF ownership on the efficiency of stock information.The research results show that:（1）ETF discounts and premiums have significant predictive properties for future ETF reutrn.The long-short hedging strategy based on ETF discounts and premiums can obtain 0.91% of returns per week.After a series of robustness tests be done,such as controlling ETF size,liquidity and sub-samples,the conclusion remains unchanged.（2）Arbitrage restrictions and mispricing caused by investors’ irrational behavior are the main factors that drive discounts and premiums to predict future ETF returns.On the one hand,as the severity of arbitrage restrictions increases,ETF mispricing will increase.The more severe the arbitrage restrictions,the more severe the discount and premium of ETFs will be more predictive of future ETF return.On the other hand,investor sentiment is also the main reason that affects ETF mispricing.The higher the investor sentiment,the greater the degree of ETF mispricing,and the stronger the predictability of ETF discounts and premiums on future ETF return.（3）A longshort hedging strategies based on cross-border ETF discounts and premiums can earn return 2% per week,far exceeding domestic ordinary ETFs,indicating that the greater the arbitrage limit,the higher the degree of ETF mispricing.The more predictable the ETF discounts and premiums on future ETF return..We further used cross-border ETF discounts and premiums to construct one-way long-short strategies and purchase and hold strategies,and obtained excess returns of one-way long-short strategies to test the relationship between investor sentiment and excess returns of strategies,and found that the underlying cross-border EFT assets correspond to the industry Investor sentiment has a significant impact on strategy excess returns,and investors’ reactions to positive and negative sentiment show asymmetry.Strategic excess returns are mainly driven by negative sentiment.Investors are more likely to overreact when sentiment is pessimistic.（4）By discussing the impact of ETF ownership on the information efficiency of underlying assets,it is found that ETFs have transmitted information to the underlying assets through purchase and redemption activities.The increase in ETF ownership can not only reduce the participation of noise traders in underlying assets,and make the stock price reflect value information more accurately,but also alleviate the degree of stock prices delayed response to information,thereby improving the information efficiency of underlying assets.ETF ownership mainly affect the information efficiency of underlying assets by reducing the transaction cost,alleviating shortsale restrictions,and encouraging more analysts to mine basic asset-related information.ETF purchase and redemption activities have played an important role in the influence of ETF ownership on information efficiency,and promoted the underlying assets to reflect market and industry information.The research contributions of this thesis are mainly in three aspects: First,it is found that ETF discounts and premiums can predict future ETF returns,and this phenomenon is widespread in major global financial markets.Through analysis of cross-border ETF arbitrage in finanacial market of China,the United States,Singapore and China’s Hong Kong,it is found that whether it is the A-share market or other developed financial markets,the ETF discount and premium are negatively correlated with future ETF return,but the predictability of the A-share market is significantly higher than that of other markets,which proves that the efficiency of the A-share market is lower than that of developed financial markets such as the United States.It provides new ideas and practical guidance for investors to use this market ineffectiveness to obtain excess returns and develop investment strategies.It will help the financial market eliminate mispricing and improve the efficiency of market information.Second,an arbitrage equilibrium mathematical model is used to derive the conditions for predicting future ETF retrun by ETF discounts and premiums,and explain the driving factors for ETF discounts and premiums to predict future ETF return.Further use of empirical analysis methods,we explains the predictability of ETF discounts and premiums in major global markets on future returns from the perspectives of investor sentiment and arbitrage restrictions leading to ETF mispricing and delayed corrections.The mathematical model is used to rigorously derive the relationship between the ETF discount and premium and the future ETF return,which makes up for the lack of rigorous logical interpretation.Third,from the perspective of ETF’s impact on the trading environment of underling assets,we explores the impact of ETF ownership on stock information efficiency,rather than simply discussing the relationship between ETF arbitrage and the information efficiency of underlying assets,which not only enriches the existing research on ETF’s information efficiency and help the management to decide whether to vigorously promote financial derivatives such as ETFs.