Reexamination of the Size Anomaly on Taiwan Stock Exchange

• Main Authors: Tung-Wei Sung, 宋東威
• Other Authors: Tung-Liang Liao
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/36475264797499926889

碩士 === 逢甲大學 === 企業管理所 === 96 === This study uses average returns, compound returns, and monthly cross-section regressions to investigate the relation between realized returns and firm size in the Taiwan stock markets from 1982 to 2006. Size effect indicates that small-cap stocks have substantially higher returns compared to large-cap stocks. Lately, studies of U.S. and Great Britain stock markets suggested that a “reversed small-size effect” or a “U-shaped pattern” might exist in these markets during 1990’s. Studies of Taiwan stock market also shown similar results the same period. This study used both realized returns and beta-adjusted returns to re-investigate the size anomaly and obtained the following finds: (1) A reversed size effect were observed during 1982-1986. (2) In general, data of 25 years study period shown an insignificant “reversed J-shape” pattern. (3) A “J or U shape” pattern was found in 1999, and hence is recognized to have both small-cap and large-cap size effect or, on the contrary, to have no size effect by some of previous studies. (4) Small-cap portfolio had asynchronous price behavior to market movement and therefore might possess some special risk factors that β failed to capture. (5) Small-cap portfolio, in the long run, has highest cumulative average returns as well as cumulative average abnormal returns. (6) Linear regression analysis using beta-adjusted data shown that there are 13 years had small-cap size effect while other 4 years had large-cap size effect during 1985-2006. In sum, three analyses report no consistent relationship between realized returns and firm size. These findings seem to show that the use of size in asset pricing or in forming the portfolios is doubtful.