Research On The Cause Of Idiosyncratic Volatility Based On Systematic Risk

According to traditional asset pricing theory,idiosyncratic risks can be completely dispersed,and there should be no correlation between idiosyncratic volatility and expected returns.However,academic articles and empirical test results have found that stocks with high idiosyncratic volatility tend to obtain lower expected returns,and there is a negative correlation between the two.This phenomenon,which cannot be explained by traditional economic theoretical models,is called ‘idiosyncratic volatility puzzle’.The current academic interpretations of ‘idiosyncratic volatility puzzle’ mainly focus on two aspects: risk compensation and investor irrationality,but the strength of the explanation is not ideal.This article takes a different approach and tries to explain the ‘idiosyncratic volatility puzzle’ from the perspective of the model itself.Specifically,a large number of articles proving the anomaly of idiosyncratic volatility have used the traditional unconditional model based on the CAPM model or the FF model,but the traditional model cannot fully explain the systematic risk due to its inherent flaws,leading to some systematic risks into the residual part,resulting in ‘idiosyncratic volatility puzzle’.In order to test the above conjecture,the article selected the common stocks of the New York Stock Exchange,the American Stock Exchange and the Nasdaq Stock Exchange between July 1962 and May 2014.After removing the missing values,12,813 stocks and 1403544 observations were obtained.First,the article uses the traditional Fama-French five-factor model to calculate and construct the idiosyncratic volatility portfolio,which proves the existence of the ‘idiosyncratic volatility puzzle’ under the traditional model and can obtain 0.6%-0.9% excess returns.Subsequently,the article replaced the FF five-factor model with the IPCA model proposed by Kelly et al.(2019),which can effectively solve the shortcomings of the traditional model and has a good prediction effect.After testing,the use of the IPCA model can effectively eliminate the phenomenon of ‘idiosyncratic volatility puzzle’,thus proving the article’s conjecture about the cause of the ‘idiosyncratic volatility puzzle’.Finally,the article uses the IPCA model to conduct a more in-depth exploration.The study found that among 36 company characteristics,some company characteristics do dominate other company characteristics,and a small number of important company characteristics can provide a good results.Replacing the latent factor of the IPCA model with the FF5 factor can also eliminate the ‘idiosyncratic volatility puzzle’.It shows that the main source of IPCA model explanatory power is time-varying β.