Long-term And Short-term Components Of Beta Factor: Impact On Equity Pricing

In the classic capital asset pricing model,the beta coefficient is a constant,and the correlation between the risk premium of the portfolio and the market risk premium does not change over time.However,as the classic asset pricing model cannot explain many financial visions,many scholars have begun to study the beta time-varying.For example,Chen(1995)believes that the beta coefficient is time-varying and its change is related to the business cycle.Glosten(2012)believes that the beta coefficient is very small,but changes over time.Beta’s time-varying has gradually become the consensus of the academic community.At the same time,many studies have shown that the impact of different frequencies of information in financial markets on yields is different.For example,Gilbert,Hrdlicka,Kalodimos,and Siegel(2016)show that there is a huge difference between stock high frequency(daily)and low frequency(quarterly)beta.Studies by Engle and Lee(1999)and Engle and Rangel(2010)show that models with high and low frequency volatility better capture the dynamics of stock returns than single-frequency models.The choice of financial market information frequency is important for properly measuring risk and studying risk and return relationships.Ignoring this may result in poor estimates of systemic risks and associated risk premiums.In view of the Chinese stock market,there is currently no relevant literature on the pricing performance of asset pricing models containing time-varying betas with different frequencies.This paper selects China’s A-shares as a sample,constructs a portfolio in Fama and French(1993)literature,and studies the impact of time-varying betas on long-term,medium-term and short-term frequencies on equity prices,which helps to strengthen time-varying beta and capital.The in-depth understanding of the asset pricing model provides a new idea for portfolio management and is of great significance for portfolio risk management.This paper defines a time-varying three-component beta model containing three beta components in the long-term,medium-term,and short-term.The short-term component is calculated from the daily frequency data of the lag phase,and the medium-term component is calculated from the 5-year daily frequency data or the 1-year daily frequency data,and the long-term beta is calculated from the 10-year monthly data.The three-component beta model first obtains the values of the three beta components of the long-term,medium-term and short-term of each asset portfolio through the OLS time series regression of the rolling window;then,in each time period,the long-term,medium-term,time-varying Cross-sectional regression is performed on the value of the short-term beta component and the yield of the portfolio.In order to measure the pricing performance of the three-component beta model to explain the cross-sectional yield difference of different asset portfolios,this paper selects the capital asset pricing model and the Fama three-factor model as the benchmark model for measuring the pricing performance of the three-component beta model,and defines several Compare the three-component beta model,the capital asset pricing model,and the metrics of the pricing performance of the Fama three-factor model.The asset portfolio selected in this paper is a combination of 25 assets classified by size and book value ratio constructed by Fama and French(1993).The sample selected all the shares of the A-share transaction.Time series regression requires an estimated beta value and cross-sectional regression time interval from 2009 to 2018.The conclusions of this paper are as follows:(1)The single-frequency time-varying beta is better than the constant beta for the difference in the cross-sectional yield of the portfolio.(2)The cross-sectional pricing effect of the three-component beta model with time-varying betas containing long-term,medium-term and short-term frequencies is comparable to that of the Fama three-factor model and is significantly better than the CAPM model.The empirical results show that the time-varying betas estimated by daily frequency and monthly frequency data have a strong explanatory effect on the cross-sectional difference of portfolio yield.The time-varying beta estimated by daily frequency and monthly frequency data is an important part of system risk.(3)This paper provides some visual explanations of the net contribution rate of long-term and short-term components in the pricing performance of the three-component beta model.The short-term and medium-term components contribute more and the long-term composition is smaller.