Informed Trading and Expected Returns
Stocks with the greatest information asymmetry have annualized returns that are 10.8 percentage points higher than stocks with the least information asymmetry.
In Informed Trading and Expected Returns (NBER Working Paper No. 18680), co-authors James Choi, Li Jin, and Hongjun Yan use daily institutional ownership data from the Shanghai Stock Exchange to examine whether information asymmetry affects expected stock returns. They argue that focusing on China is useful because there is likely to be significant variation across companies in how much private information is shared with select investors, largely as a result of the state of Chinese legal institutions and regulations.
The authors first show that stocks bought heavily by institutions subsequently outperform stocks sold heavily by institutions. Thus, institutions appear to have a strong information advantage over individual investors, and that is true for stocks of all sizes. Moreover, the authors confirm that the institutional sector's future information advantage is larger in stocks that it previously traded more aggressively. Therefore, the aggressiveness of institutional trading in a stock, as measured by prior institutional ownership volatility, can be used as an ex ante predictor of future information asymmetry in this stock.
Sorting stocks based on this predictor of information asymmetry, the authors find that the 20 percent of stocks with the greatest information asymmetry have future annualized returns that are 10.8 percentage points higher than the 20 percent of stocks with the least information asymmetry. This difference remains significant for ten months after the initial sorting monththe same amount of time that the difference in institutional information advantage between the two portfolios lasts. There is no evidence of subsequent return reversals. They conclude that information asymmetry increases the cost of capital.
--Claire BrunelThe Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.