成交量長期變動率與股價關係：考慮財經狀態

本研究主要探討長期成交量的變動率是否隱含預測未來股票報酬率的資訊。以TESLA作為研究對象，將TESLA之日交易量修改為累積一段時間，即為長期成交量，採用Chen (2012)所使用的線性迴歸檢定對樣本進行統計分析。本研究發現到成交量於累積1期時，無任何顯著預測力，而分別累積至40、80、120期時，有顯著的預測力。本研究在分析長期成交量對預測未來報酬率的相關性後，加入景氣循環、股市漲跌、COVID-19、異常交易量四種不同財經狀態下的預測進行分析。實證結果顯示，長期成交量預測未來報酬率在不同狀態之下皆有影響，投資人可結合過去長期成交量的變動與股價關係，再加入不同的財經狀態比較，做為未來投資決策的參考。

關鍵字

長期成交量；變動率；股票報酬率；財經狀態

並列摘要

This study primarily investigates whether the rate of change of long-term trading volume contains information for predicting future stock returns. Using Tesla as the research subject, the daily trading volume of Tesla is transformed into cumulative volume over a period of time, which represents the long-term trading volume. The statistical analysis of the sample was conducted using the linear regression test employed by Chen (2012). The findings indicate that there is no significant predictive power for trading volume with a lag of one period, but when accumulated over 40, 80, and 120 periods, it exhibits significant predictive power. After analyzing the correlation between long-term trading volume and predicting future returns, the study incorporates four different economic and financial state: Business cycle, stock market fluctuations, COVID-19, and abnormal trading volume for further analysis. The empirical results demonstrate that long-term trading volume predicts future returns under different state. Investors can combine the historical changes in long-term trading volume with the relationship between volume and stock prices, and then compare them under different economic and financial state as a reference for future investment decisions.