| The stock market,which is at the core of finance,is closely related to the development of the entire market economy and is called the "barometer" of economic conditions.Stock return prediction plays an important role in various fields such as risk management and asset pricing,and at the same time,it can help investors avoid certain risks and seize investment opportunities to obtain more considerable investment returns.Therefore,the study of stock return predictability,which has attracted the attention of many scholars and investors,has both academic and practical significance.Among a range of variables used to predict yields,oil prices have been attracting increasing research interest due to their important role in global resource and basic industrial inputs.This article focuses on the excess returns of stocks in the S&P 500 index,using monthly data for up to 90 years.By comparing with a series of macroeconomic variables,this paper explores the predictive ability of oil price changes and their asymmetric forms to stock excess returns in and out of the sample.In terms of econometric methods,this paper uses a series of popular constraint strategies to make predictions.Finally,based on the main empirical analysis,from the perspective of economic cycle,financial crisis and geopolitical risk,the performance of different forecasting strategies in different periods is studied to explore their cyclical characteristics.This paper finds that,based on the unconstrained method and three popular forecasting constrained strategies,oil price changes have stronger and more robust predictive power than their asymmetric forms.A new constraint method on the three-sigma rule can achieve higher prediction accuracy than other constraint methods.Compared with the univariate macroeconomic model,adding oil price changes can improve the average predictive performance of 14 macroeconomic indicators.In addition,the forecast performance of the two oil variables varies in different periods related to economic cycles,geopolitical risks and financial crises.Finally,the sources of predictability of oil price variables on stock returns can be explained from the discount rate channel and the sentiment channel.Compared with the existing literature,firstly,this paper adopts a series of constraint methods to improve the predictability of stock returns with this simple but effective strategy.And among the three prediction constraint methods,a new constraint method on the three-sigma rule’s prediction performance is more significant and robust.Second,because of financial shocks or external shocks,there is an impact on the relationship between oil and stocks.This paper finds that the predictive ability of oil price variables varies not only with financial shocks such as economic cycles and financial crises,but also with geopolitical risks,which is a further improvement of relevant research conclusions.Finally,based on the empirical results,the economic explanations for the predictive power of two oil price variables on stock returns are explored.The practical significance of this research lies in: the empirical results further verify the predictive ability of oil price variables on stock returns,and discover oil investment strategies that are better than most macroeconomic variables,and explore more accurate and professional forecasting methods for financial practitioners basis is provided. |
