V-Shaped Disposition Effect, Stock Prices, and Post-Earnings-Announcement Drift: Evidence from Korea

Abstract

This study investigates the impact of the V-shaped disposition effect on asset prices in the Korean stock market, which is characterized by a high proportion of retail investors. By utilizing a specified dataset containing stock-level information on the trading activities of different types of investors, we find evidence to support the presence of V-shaped net selling propensity in the Korean stock market. In addition, we find that net selling pressure has a positive effect on the cross-section of subsequent stock returns, and this relationship appears only when accounting for individual trading. Furthermore, this net selling propensity of retail investors delays the incorporation of good news into stock price, while helps stock price reflect its bad news. We show that good (bad) news lead to positive (negative) drifts in stock prices following earnings announcements in the presence (paucity) of investors exhibiting the V-shaped disposition.

Keywords:

• V-shaped net selling propensity
• Disposition effect
• Individual investor
• Post-earnings announcement drift
• Korean stock market

JEL CLASSIFICATION:

• G12
• G14
• G40

Notes

1 When there is little change in stock prices after stock purchases, speculative investors have no reason to reexamine their stock positions, revise their beliefs, or trade due to small changes in stock prices. However, substantial gains (or losses) compel them to reexamine their stock positions and trade accordingly (Ben-David and Hirshleifer 2012).

2 Based on trading amount data, we calculate the time-series average of the proportion of trading by individual investors in the Korean stock market (until 2017) and find that they account for 55% of the trading in the KOSPI market, 91% in the KOSDAQ market, and 66% in both the markets combined. The higher proportion in the KOSDAQ market could be due to the larger proportion of listed stocks that are more likely to be traded by individual investors, that is, smaller stocks, with a lower price, greater volatility, and a skewed return distribution (e.g., Han and Kumar, 2013).

3 Most of the trading activities in our database are classified under three types of investors: individual (retail), domestic institutional, and foreign investors. Since foreign investors are mostly foreign institutions, we group domestic and foreign institutions into ‘institutional investors’.

4 Bernard and Thomas (1989) first argued that PEAD is a delayed price response due to investors’ under-reaction, following which many studies have attempted to explain the reasons for a stock price drift following earnings announcements, which include arbitrage risk (Mendenhall, 2004), transient institutional trading (Ke and Ramalingegowda, 2005), divergence of opinions (Garfinkel and Sokobin, 2006), disposition effects (Frazzini 2006), trading costs (Ng et al., 2008), illiquidity (Chordia et al., 2009), investor inattention (Dellavigna and Pollet, 2009; Hirshleifer et al., 2009), ex ante earnings volatility (Cao and Narayanamoorthy, 2012), streaks in earnings news (Loh and Warachka, 2012), informational risk (Zhang et al., 2013), and anchoring bias (Birru, 2015; George et al., 2015).

5 A few studies analyze PEAD in the Korean stock market based on psychological bias, but only Goh and Jeon (2017) explain it using anchoring bias. Other studies attempt to identify the underlying factors of PEAD in the Korean stock market with a focus on audit quality (Nah and Lee, 2009), the trading volume around earnings announcements (Choi and Kim, 2009), individual investors’ trading activities (Lee and Choe 2012), and information uncertainty (Lee et al., 2015).

6 Speculative motive of investors can result from superior information or overconfidence. This paper does not discuss how investors become speculative. Instead, we examine the impact of this speculation-based V-shaped net selling propensity on stock prices and interaction with respect to under-reaction to earnings news.

7 The average purchase price is calculated as total buying amount divided by total buying trading shares.

8 This estimation window enables one to account for the different investment horizons of investors (Grinblatt and Han 2005).

9 In terms of whole investors, the buy and sell trading volume should be the same. However, when classifying the types of investors, the buy/sell trading volume can be different. For example, assume that individual investors buy 100 shares of stock A and sell 50 shares of the same stock. Simultaneously, institutional investors buy 50 shares and sell 100 shares of stock A. In this case, the total market buy/sell trading volume are 150 shares, but the buy and sell trading volume of individual (institutional) investors are 100 and 50 (50 and 100) shares, respectively.

10 We obtain similar empirical results using Amihud’s (2002) illiquidity measure as a proxy for market depth.

11 We obtain the same qualitative and quantitative empirical results using applying An’s (2016) asymmetric sensitivity for unrealized gains and losses (-0.23) to net selling pressure.

12 Our main empirical results do not change qualitatively when we isolate domestic institutional investors from foreign investors.

13 For example, Samsung Electronics accounts for more than 20% in the KOSPI200 index. Further, there are almost 2,000 common stocks are listed in the Korean stock market, but the Korea accounts for only 2% of the global market. In fact, there are many tiny stocks listed in the Korean stock market with a market cap of less than $1 billion. To resolve this issue, Asparouhova, Bessembinder, and Kalcheva (2010) suggest a gross return weighting scheme that minimizes confounding microstructure effects. They prove that this weighting scheme is a consistent estimate of the equally weighting scheme.

14 Unreported empirical results are provided upon request.

15 This methodology is similar to the overhang spread and the negative overhang spread constructed by Frazzini (2006).

16 Evans (2009) shows that retail investors comprise of only 30% of the US stock market. Further, recent data of the New York Stock Exchange (NYSE) reveal that trading by individual investors represent less than 2% of total trading volume of NYSE-listed firms on average.

17 See http://data.worldbank.org/indicator/CM.MKT.TRNR