流動性衡量方法:在資產訂價模型的應用

Abstract

「流動性」一直都是都是投資人做投資決策時的重要考量因素。故如何精準地衡量流動性，就是一個重要的議題。此篇論文根據過去Amihud (2002)和Pastor and Stambaugh (2003)等文獻，提出六種衡量流動性的方法，分別是HLVS、ILLIQ、the Bid-Ask Spread、LM6、LM12，和Market Liquidity Measures等。我們將利用此六種方法，做樣本期間自1986年1月至2015年12月共360筆的追蹤資料回歸分析( Panel data regression analysis)，發現此六種流動性方法對市場上公司的股票報酬皆具有顯著的解釋力，其中以Bid-Ask Spread和ILLIQ兩種方法，對於預期的股票報酬具有更高的解釋能力。另外，藉由美國產業標準分類碼(SIC code)，我們將全美上市公司分成十個產業，並依照上述方法做追蹤資料回歸分析，我們發現金融保險及不動產業、批發業，和零售買賣業這三個產業，流動性對於預期股價報酬確實具有較高的解釋力。又以上述六種流動性方法做資產訂價模型，並與CAPM、Fama-French三因子訂價模型相做比較，發現在不同的情況下，此些流動性方法互有不同的表現，而其中以 HLVS、ILLIQ 和the Bid-Ask Spread 對於預期報酬和異常報酬有較顯著的解釋能力。

Everyone knows that “liquidity” is an important issue because it is a major consideration for investment decisions. In this paper, we introduce six liquidity measurements, like HLVS, ILLIQ, the Bid-Ask Spread, LM6, LM12, and Market Liquidity Measures (γ). These measurements are constructed by previous papers and calculated by the daily data from the CRSP. Our finding shows that they really have ability to explain the expected returns whatever is all sample period or just the sub-periods. Furthermore, following the SIC code’s classification, we find that the industry of Finance, Insurance & Real Estate, Retail Trade, and Wholesale Trade will be the most affected by the liquidity, and the industry of Agricultural, Forestry, Fishery, and Husbandry, and Construction have less significant effect. These results are almost consistent with our expectations. Also, we use those liquidity measurements to compare with CAPM and Fama and French three factors model. The results demonstrate that HLVS, ILLIQ, and the Bid-Ask Spread show better performances to significantly explain the expected returns and anomaly returns in most of the situations.