Short selling is the practice of selling securities or other financial instruments that are not currently owned, and then subsequently repurchasing them in the near future to return borrowed securities to their owner. Short selling can play a positive role in the market. It increases market liquidity, contributing to efficient price discovery, and helps to expand the selection of investment strategies and risk management. However, short selling can also play a negative role, with naked short selling sometimes used to manipulate the market. During the global financial crisis triggered by the subprime mortgage crisis in 2007, the financial authorities in many developed countries implemented short-selling restrictions to stabilize financial markets, protect investors, and aid efficient price discovery. Most studies on short selling concern their relation with stock prices (returns), with less attention paid to the relation between short selling and volatility. In this study, we investigate the relationship between short-selling activities and stock return volatility in the Korean stock market using daily data on individual stock returns, stock trading volume by investors, and short-selling activities by investors. The sample period ranges from January 2007 to November 2012. Three types of traders are considered: individual investors, institutional investors, and foreign investors. We use several volatility measures: historical volatility, range-based volatility (Garman-Klass volatility or intraday volatility), implied volatility (VKOSPI or the volatility index), and conditional volatility, which is estimated from the GJR-GARCH model. Three kinds of short-selling activity measures are adopted as the major explanatory variable: logarithm of the money value of short-selling volume, the short-selling/stock trading volume ratio, and the share of short-selling volume by trader type. Miller (1977) and Diamond and Verrecchia (1987) suggest that policies that impose short- selling constraints increase volatility in the stock market. Such a policy was enacted during the sample period, allowing us to test the prediction that short-selling transactions as a whole increase volatility in the Korean stock market. However, the daily short-selling trading volume was found to exert a negative effect on the volatility of KOSPI index returns. In the case of short-selling restrictions, it turns out that such restrictions serve to reduce intraday volatility, conditional volatility, and the volatility index (VKOSPI), although the effect of a short-selling ban on volatility remains inconclusive. There are two short-selling ban periods in our sample period: the eight months from October 1, 2008 to May 31, 2009 (BN1) and the three months from August 10 to November 9, 2011 (BN2). The sample data show BN1 and BN2 to have a higher and lower volatility level, respectively. After empirical analysis, the short-selling ban was found to increase intraday volatility and the VKOSPI in BN1 and to decrease them BN2, which implies that such a policy measure does not have deterministic effects on volatility. When we classify short-selling trading volume by trader type, the short-selling volume of individual investors is observed to stabilize daily range-based volatility and the VKOSPI, whereas that of institutional investors destabilizes such volatility measures as historical volatility, range-based volatility, and the VKOSPI. Foreign investors’ short-selling volume increases none of the volatility measures. The finding that the short-selling trading activities of individual investors stabilize the stock market and those of institutional investors increase stock market volatility, whereas, in the stock market more generally, individual investors increase stock market volatility and institutional investors stabilize the market, is an interesting one. It implies that individual investors have an information advantage in short-selling trading, considering that their share of such trading volume hovers around 3%. This finding is consistent with that of Jung, Kim, and Lee (2013), who report that the short-selling trading activities of individual investors do not increase stock market volatility in Korea. Foreign investors stabilize the stock market through both short-selling transactions and stock trading activity. With regard to policy effects, we find that short-selling restrictions reduce intraday volatility, the VKOSPI, and conditional volatility. The short-selling ban initially prompted by the global financial crisis appears to have had limited effects, although the second prohibition clearly reduced volatility. Therefore, it can be concluded that short-selling regulations can stabilize the market if implemented in appropriate and timely fashion. A remaining puzzle is why individual investors’ short-selling activities reduce volatility, whereas those of their institutional counterparts increase it. The issue requires further analysis, and is left to future research.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

A random-end (RE) trading mechanism is an integral part of the call auction mechanism. The ending time of a call auction is not fixed, but can be extended in certain circumstances for a brief randomly-chosen span of time less than or equal to the “maximum duration.” RE mechanisms are intended, in part, to discourage the placement of spoofing orders. In this paper, we investigate the main characteristics of a specific RE trading mechanism, that which the Korea Exchange (KRX) employs at the opening and closing call auctions, and its effects on price discovery and stabilization. We are aware of four papers related to RE trading mechanisms. Medrano and Vives (2001) is a theory paper, whereas Hauser, Kamara, and Shurki (2012) is an event study analyzing market quality just before and after the adoption of the RE trading mechanism. Hence, neither paper illuminates the detailed functioning of RE mechanisms. The other two papers, Zimmermann (2013) and Eom and Park (2014), directly analyze the economic functions and effects of real-world RE mechanisms. The present paper is in line with Eom and Park (2014) in that it examines the KRX RE trading mechanism, but differs from it in that it performs cross-sectional analyses and analyzes spoofing orders, the discouragement of which was the main motivation for the mechanism’s adoption. Our sample comprises all 1,567 stocks listed on the KOSPI and KOSDAQ markets after filtering out a typical group of special-purpose stocks. We analyze all RE invocations occurring among the sample stocks during the daily opening and closing call auctions from January 2009 to December 2010. We use both daily and intraday data during our sample period. For real- time RE occurrence data, we first reconstruct the real-time order book from KRX Trade and Quote data, and then match our reconstructed order book against the complete set of RE occurrences provided by the KRX to validate the reconstruction process. We obtain the following results. First, at both the open and close, RE was invoked at least once for most stocks listed on the KOSPI and KOSDAQ markets. However, it was invoked most frequently for smaller, more volatile, lower-priced, financially distressed, or poorly managed stocks. The RE occurrence rate is consistent with market volatility. That of KOSDAQ is 2.1% on average, which is higher than that of KOSPI, at 0.8%. The disparity ratios between the projected price and potential price and between the potential price and opening (closing) price are 8.3% (8.2%) and -27.5% (-30.2%), respectively. The reversals in these ratios indicate that the RE trading mechanism makes a meaningful contribution to price stabilization for the opening and closing prices. The realized duration between the scheduled opening (closing) time and the real opening (closing) time is, on average, 1 minute 30 seconds, which indicates that the RE time is not drawn from a uniform distribution. Second, the opening process on the days on which RE occurred exhibits a different pattern from that on the days it did not occur, suggesting that sophisticated investors can identify the discrepancy and trade strategically, whereas other investors trade normally. Also, at the closing call auctions, the ratios of cancellation and correction orders to normal orders were overwhelmingly higher than those at the opening call auctions. Third, although the potential prices overshot to some extent, they were close to the average market equilibrium price formed during the five minutes after RE occurrence. Moreover, some of the overshooting was corrected during the RE period. These results imply that the RE trading mechanism contributes to opening price discovery and stabilization. Finally, many more spoofing orders were observed for the stocks for which RE occurred than for those for which it did not. Further, among the stocks for which RE occurred, spoofing orders were more than twice as numerous on the days of occurrence, with spoofing much more pronounced in the closing call auctions. In addition, at both the open and close, spoofing orders fell sharply after RE invocation, suggesting that the RE mechanism helps to discourage spoofing. In sum, our results suggest that the current KRX RE mechanism fulfills its purpose to some extent, that is, by improving price discovery and stabilization and reducing spoofing order submission, although more could be achieved. To increase the efficacy of its RE trading mechanism, the KRX should consider adjusting the current parameters. For example, it could set multiple RE invocations in succession, make the disparity rate more flexible to deal with situations as they develop, expand the RE period, and so on. The major European exchanges, which have been praised for providing some of the most sophisticated RE trading mechanisms in the world, would be a good benchmark.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

We construct the “betting against beta” (BAB) factor in Frazzini and Pedersen (2014) using Korean stock market data, and investigate the relationship between leverage constraints and stock returns. Frazzini and Pedersen (2014) extend Black’s (1972) capital asset pricing model (CAPM) with restricted borrowing by explicitly considering leverage and margin constraints that vary across investors and time. According to their model, investors who cannot use leverage overweight the high-beta assets in their portfolios, leading to the overpricing of those assets and subsequent lower returns. Unconstrained investors, in contrast, underweight (or short-sell) overpriced high-beta assets and buy low-beta assets with leverage. Frazzini and Pedersen (2014) show that their model with leverage constraints produces a security market line that is flatter than that predicted by the CAPM, with the slope depending on the tightness of the leverage constraints that investors face on average. Frazzini and Pedersen (2014) derive several testable predictions from their model, and they propose a BAB factor as a central portfolio whose pattern of returns captures the asset pricing effect of leverage constraints implied by their model. The BAB factor is a portfolio that holds low-beta assets, leveraged to a beta of one, and shorts high-beta assets, de-leveraged to a beta of one. The returns on the BAB factor correspond to a zero-cost (self-financing) investment strategy that longs low-beta assets and shorts high-beta assets, with the weights on the long and short positions adjusted in combination with leverage to ensure that the factor’s ex ante beta is zero. Frazzini and Pedersen’s (2014) model predicts the BAB factor to have a positive expected return and a realized return that decreases when leverage constraints become more binding. They test their model’s predictions by constructing BAB factors in a wide range of asset classes, including 20 international equity markets, Treasury markets, and corporate bonds and futures markets, and find evidence to support its predictions. We construct a BAB factor using the stocks listed on the Korea Stock Exchange, and investigate whether the relationship among leverage constraints, the beta, and risk-adjusted returns exhibits patterns consistent with the predictions of Frazzini and Pedersen’s (2014) model. The sample period is from April 1990 to Mach 2014. We find the average monthly return on the BAB factor to be significant both economically (1.32%) and statistically (t-statistic of 2.85). Its alphas (risk-adjusted abnormal returns) are also of sizable magnitude (CAPM alpha of 1.33% and Fama-French three-factor alpha of 1.25%) and statistically significant. These findings are consistent with the overpricing of high-beta stocks and underpricing of low-beta stocks arising from leverage constraints, as predicted by Frazzini and Pedersen’s (2014) model. The BAB factor is constructed using beta-ranked portfolio weights: lower-beta stocks have greater weight in the low-beta portfolio (a long position in the BAB factor), and higher-beta stocks have greater weightin the high-beta portfolio (short position in the BAB factor). This weighting scheme is designed to strengthen the beta’s effect on the alpha (i.e., the alpha is decreasing in the beta) in returns on the BAB factor. We examine the effect of different portfolio weighting schemes on the average returns on the BAB factor by constructing the factor using more conventional value-weighting and equal-weighting schemes. The results show the risk-adjusted return on the BAB factor to be substantially reduced in size and to loseits statistical significance when value-weighting is used, which indicates that the effect of leverage constraints on the underpricing of low-beta stocks is likely to be concentrated in small capitalization stocks in Korea. Accordingly, we classify stocks into three groups by market capitalization and construct a BAB factor within each group. We find that the BAB factor constructed within the small capitalization group shows a larger average (and risk-adjusted) return with stronger statistical significance. Moreover, when the proportion of trading by institutional and foreign investors is considered in addition to market capitalization when forming the BAB factor, we find the risk-adjusted return on the factor to be sizable and statistically significant when it is constructed using small capitalization stocks with a lower proportion of trading by institutional and foreign investors. These results suggest that the pricing effects of leverage constraints in Korea are concentrated primarily in small capitalization stocks heavily traded by individual investors. We also investigate the hypothesis of Frazzini and Pedersen’s (2014) model that the realized return on the BAB factor decreases when leverage constraints become more binding. We find that changes in the 91-day CP-CD spread and 3-year AA-rated corporate bond-government bond spread are negatively associated with contemporaneous returns on the BAB factor in Korea. These results are consistent with Frazzini and Pedersen’s (2014) finding that changes in the TED spread are negatively associated with contemporaneous returns on the BAB factor in the U.S. Taken together, our findings suggest that the pattern of returns on the BAB factor in Korea captures the asset pricing effect of leverage constraints faced by investors, as postulated by Frazzini and Pedersen (2014).

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

This paper evaluates the (out-of-sample) prediction performance of bankruptcy prediction models using Korean firms. Based on the source of information, we classify these models into accounting-based, market-based, and hazard categories. Note that hazard models are based on both accounting and market information. We consider five bankruptcy prediction models in this study; two accounting-based models, one market-based model, and two versions of a hazard model. One of the accounting-based models employs multivariate discriminant analysis (MDA), and the other employs logit analysis. The example of the first accounting-based model is the Altman (1968) Z-score model, and that of the second is the Ohlson (1980) O-score model. Most studies in the Korean literature use the accounting variables and coefficient estimates that Altman (1968) and Ohlson (1980) use for U.S. firms to predict bankruptcy, which may result in bias and inaccuracy because accounting variables may have different economic implications in different countries. Accordingly, in this study, we select new accounting variables that better fit Korean firms with respect to discriminant power and goodness-of-fit, and re-estimate the coefficients. For the market-based model, we use the KMV default-to-distance (DD) model. DD indicates the distance from the mean of the firm’s current asset value to its default point. The greater the DD, the smaller the probability of default. For the hazard model, we use the bankruptcy prediction model developed by Campbell, Hilscher, and Szilagyi (2008) (CHS), which uses both market and accounting information and has become popular in the finance literature. It has also been shown empirically that the CHS model is effective in predicting bankruptcy for firms in a variety of countries. We consider two versions of this hazard model. The first is one that adopts the same variables as those selected for U.S. firms in the CHS model and re-estimates them using Korean data. The second is a modified version of the CHS model. The variables are newly selected for Korean firms, and the coefficients are then re-estimated using data on these firms. The modification involves the addition of effective variables and exclusion of irrelevant variables based on empirical analyses. We estimate the five foregoing bankruptcy prediction models using data on all Korean firms in the seven years (in-sample period say, 2001~2007) prior to the one-year forecasting period (out-of-sample period say, 2008). By rolling over year by year, we repeat the in-sample estimation using the seven-year data prior to the one-year out-of-sample period. The sample period is 2001 to 2013. Thus, we obtain forecasting results from the five models for six years (i.e., 2008, 2009, 2010, 2011, 2012, and 2013). We then evaluate their bankruptcy prediction performance. Three methods are used to evaluate the models’ prediction accuracy for Korean firms: the hit ratio, receiver operating characteristic (ROC) curve, and information content test. The hit ratio is calculated as the ratio of the number of bankrupt firms in the portfolio to the number of all bankrupt firms in each forecasting year. Portfolios are formed by assigning all firms into one of ten decile portfolios based on the seven-year in-sample estimation results of each model. The hit ratio and ROC curve are traditional comparison methods that classify firms dichotomously (bankrupt or not). The information content test, in contrast, assesses whether different models convey information on actual bankruptcy. Of the five models evaluated, the hazard model modified for Korean firms performs best in predicting actual bankruptcies in the out-of-sample period with respect to all three bankruptcy prediction performance measures. The unmodified hazard model performs second best, and the DD model worst. There is little difference in prediction performance between the two accounting-based models, i.e., that using logit analysis and that using multivariate discriminant analysis, although their performance varies depending on the performance evaluation method used. This paper contributes to the literature in the following ways. First, by modifying the accounting and market variables used in the CHS model for Korean firms and re-estimating the model, we present a new hazard model that better fits Korean firms and outperforms other models in bankruptcy prediction accuracy. To the best of our knowledge, this is the first study to evaluate the performance of bankruptcy prediction models using Korean data. Second, this study is also the first in the Korean literature to include data over a period encompassing the 2008 global financial crisis. As there were many bankruptcies during that period, its inclusion enhances the reliability of our results. We suggest that our modified hazard model be used in future studies in both academia and industry.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel

Institutional investors are obligated to actively exercise their shareholder rights to perform fiduciary duties as trustees. Korean institutional investors, however, do not perform these duties diligently because they indirectly invest in entrusted assets and are reluctant to exercise shareholder rights out of concerns over investing companies, political affairs, and institutional constraints. Foreign investors, in contrast, are more likely to exercise their shareholder rights because they are free from these concerns. This paper investigates the role of domestic and foreign institutional investors as corporate monitors using a daily mechanism rather than voting rights, which are exercised only on specific proposals at general shareholder meetings. We analyze how the agency costs and value of target firms are affected by the sale of a portion of their stocks by domestic and foreign institutional investors. Our sample comprises all institutional investors’ disclosures of an equity ownership change to abide by the “5% rule” in the period from April 2005 to December 2012. Under the 5% rule governing large share acquisitions of publicly held companies, a person or institution who holds 5% or more of the equity securities in a target company or thereafter changes its holdings by 1% or more must file a report with the Financial Supervisory Commission and Korea Exchange (KRX) within five days of the transaction or change date. We exclude from the sample announcements made by a target company between the trading date and announcement date of stock sales because the firm announcement effect and institutional investor disposal effect are likely to be intertwined. We also exclude disclosures that generate a positive market response because only unfavorable announcements are followed by monitoring-motivated stock sales. If a target firm makes an announcement before an institutional investor’s announcement of share disposal, we include that announcement only if it shows a statistically significant negative market response at the 10% level, as measured by the cumulative abnormal returns (CAR) around the stock sale disclosure date. CAR is calculated by summing all market-adjusted abnormal returns around the target firm’s announcement. If a target firm makes no announcement within five days of institutional investors’ announcement of a stock sale, the announcement is included in the final sample to ensure that we do not rule out the disposal of a portion of shares that is too small to attract an announcement obligation by institutional investors to punish management seen to be destroying shareholder value. A sale is regarded as a monitoring-motivated sale when the CAR around the announcement date of an institutional investor stock sale is positive at the 10% significance level. We expect that institutional investors’ sale of a portion of their stake reduces a target firm’s agency costs and increases corporate value if it has a “kick-in-the-pants” motivation and that strategic investors are more likely to reprimand and exert influence on management than financial investors. However, the opposite outcome is also possible if management subordinates shareholders’ passive stock sales to the more active exercise of shareholder rights. In multiple regression analysis, the dependent variables are agency costs measured by the ratio of operating expenses to sales, which tend to decrease when managers’ perks are restrained, and the asset turnover ratio, which tends to increase when assets are used more efficiently. Another dependent variable is firm value measured by (industry-adjusted) Tobin’s Q. The explanatory variables are a monitoring dummy indicating a positive market response, an institutional investor intention-to-hold stock dummy indicating a strategic investment, an unfavorable announcement-by-the-target-firm dummy, a foreign investor dummy, and a foreign investor intention-to-hold stock dummy. Two-stage least-squares regression analysis is performed with one year lagged data to consider the possible endogeneity of the monitoring dummy representing monitoring motivation. We find that an institutional investor’s sale of a portion of its stock is more likely to reduce operating expenses relative to sales when an unfavorable disclosure by the target company generates a positive market response to the institutional investor’s announcement of that sale and when foreign investors are labeled as strategic investors. Our results also show that stock sales by institutional investors improve firm value. For foreign investors, firm value is improved only when they report the purpose of their stock holdings to be management participation. Strategic institutional investors tend to have a negative effect on firm value except in the case of kick-in-the-pants-motivated sales, which suggests that we should encourage strategic foreign investors to improve the long-term performance of firms. In sum, institutional and foreign investors monitor firms through the sale of a portion of their shares, in addition to registering a “no” vote on firms’ proposals in shareholder meetings and engaging in face-to-face talks with management.

• EndNote
• RefWorks
• Scholar's Aid
• BibTeX

Export Citation Cancel