Earticle

Levi and Welch (2020) argue that market beta and asymmetric downside beta are highly correlated and that most of downside beta's explanatory power stems from market beta. This study re-examines the relationship between stock returns and downside beta in the Korean stock market. We test beta spreads including beta asymmetry (i.e., down beta - up beta spread) and relative downside beta (i.e., down beta - market beta spread) to control for the market beta. We find negative correlations between downside betas and stock returns even after controlling for market beta and firm characteristics. We also find that regardless of market conditions (i.e., bear, neutral, and bull markets) high beta spreads are associated with low returns. Accordingly, zero-cost portfolios that purchase the stocks in the low quintile of beta spreads and sells the stocks in the top quintile generate significantly positive alphas. These findings underline that downside beta contains the unique information beyond market beta in the Korean stock market.

This study investigates whether pre-meeting vote disclosures by large asset owners can amplify the influence of their votes at shareholders' meetings. We utilize the newly adopted 2019 rule that mandates Korea’s National Pension Service (NPS) to disclose voting decisions before the shareholders’ meeting. Firms subject to this pre-meeting disclosure must meet either of two conditions: the NPS holds at least 10% of the firm’s voting shares or the firm’s weight in the NPS domestic equity portfolio is at least 1%. Employing a regression discontinuity design and difference-in-differences analyses, we find that the NPS pre-meeting disclosure causes other institutional investors to increase their conformity to the NPS votes. We further find that this effect mainly originates from the resolutions that the NPS votes against management; this result is pronounced among the institutional investors who are more likely to face problems related to conflicts of interest.

Educational background in Korea can represent personal traits, since academic tracks in the Korean educational system are determined earlier in the individual’s lifetime than in other countries. This study investigates whether CEO educational background affects the firm’s external financing choices. We find that firms with CEOs who majored in science or engineering are less likely to issue equity, and the investment of firms with such CEOs is more responsive to cash flow. We also find that equity issue announcements by firms with such CEOs exhibit more negative market reaction, suggesting that those have bad signals, for instance, poor conditions to issue debt even though they prefer debt to equity. Overall, our study suggests that Korean CEOs’ educational background, particularly science or engineering, is related to managerial confidence, which is the belief that the market undervalues their firms. However, we cannot completely rule out the endogeneity concern arising from CEO-firm matching.

본 연구는 홍콩 HSCEI지수를 기초자산으로 하는 한국 ELS(Equity Linked Security) 발행량이 홍콩 옵션시장의 변동성 스큐에 미치는 영향을 분석하였다. 국내 ELS의 대표적인 구조는 3년 만기의 6개월마다 자동행사 옵션을 부여한 상품으로서 중위험 중수익을 추구하는 상품이다. ELS를 자체 운용하는 경우 헤지 운용과정에서 다양한 위험에 노출되게 되는데, 그중 대표적인 위험이 변동성 변화에 기인하는 베가위험이다. 일반적으로 녹-인 풋옵션이 내재된 ELS는 발행시점에 발행자의 입장에서 풋옵션 매수 포지션과 유사하여 변동성 헤지거래는 옵션 매도를 통하여 이루어진다. 2014년 이후 국내 증권사는 ELS 평가시 변동성 곡면 도입과 동시에 베가위험을 곡면의 형태로 산출하기 시작하였다. 따라서 이후 특정 행사가격과 잔존만기에 해당하는 옵션에 대한 매도 수요가 증가하였다. 이를 확인하기 위해 ELS 발행량이 외가격 풋옵션과 등가격 옵션의 변동성 차이로 정의된 변동성 스큐에 미치는 영향을 분석한 결과, ELS 발행량과 변동성 스큐에는 강한 음(-)의 상관관계가 나타났다. 특히 2014년 이전과 이후를 비교하였을 때, 이전대비 2014년 이후 강한 상관관계가 있음을 확인하였다. 이는 국내 증권사의 변동성 곡면 도입 이후 ELS 발행이 증가함에 따라 특정 행사가격의 옵션 매도 수요가 증가하여 나타난 결과임을 시사한다.

This paper examines the effect of issuance of ELS(Equity Linked Security) with the Hong Kong HSCEI index as an underlying asset, on the volatility skew in the Hong Kong options market. The most common structure of ELS includes an autocallable option every 6 month and 3-year time-to-maturity, which pursues medium-return along with medium-risk. When trading ELS as an issuer, traders are exposed to various risks in the process of hedging - typically the vega risk caused by changes in volatility. In the case of ELS with knock-in put option, the issuer’s position is generally similar to taking a long position in a put option at the moment of issuance, so vega hedge is normally executed through selling vanilla options. As the ELS with global stock indices as underlying assets has become popular since 2014, local securities companies have begun to adopt volatility surfaces to evaluate the fair price of ELS for the purpose of increasing the book sizes and enhancing trading performance. Accordingly, the demand for selling options with specific strike prices is matched with the autocall barriers. We analyze the effect of ELS issuance on market volatility including the VHSCEI index. The volatility skew is defined as the difference between the volatility indexes of the OTM(Out-of-The-Money) put option and those of the ATM(At-The-Money) option. Since the first autocall barriers of ELS are located on 80% to 95% of the strike price, the volatility index of the 90% strike put option is used. So far, most studies related to ELS examined the effect of issuance on the trading volume and volatility of underlying assets. Even in the previous studies on the effects of delta and gamma (or vega) risk hedging on the volatilities, the effect on only short-term volatility such as realized volatility or VKOSPI index was analyzed. Also there are various studies on the way of calculating the volatility surface depending on maturity and strike price. But there was no studies on the effect of the issuance of structured products in one country on the option market in another country. This study is the first study on the effect of issuance of structured products in Korea on the volatility skew in the Hong Kong options market. The detailed results are as follows. First, as the issuing volume of ELS increases, market volatility such as VHSCEI index decreases, and in particular, VHSCEI index decreases more relative to the realized volatility. This is consistent with the results of previous studies. Also, issuance of ELS lowers the volatility index of the 90% put option compared with the VHSCEI index. Second, with respect to the volatility skew of the same maturity, a strong negative correlation between the ELS issuance volume and the volatility skew since 2014 is observed. It can be interpreted that the volatility of OTM put options decreases compared with the ATM ones as the selling pressure for the OTM increases compared to the ATM. These results are the same for all OTM put options. As a result, if the volume of financial products with the same structure increases significantly, and in particular, if the underlying assets, maturities, and strike prices of the financial products are concentrated, the impact of volatility on the options market can be expected to be larger. It is necessary to sophisticate the issuing and book-managing strategies to diversify the risk of the ELS issuer and to enhance the product’s yield in the future.

행동재무학에 따르면 주식시장에서 관심효과는 투자자의 의사결정에 중요한 영향을 미친다. 투자자의 관심이 높은 주식은 과신과 같은 투자자의 비합리적인 편향이 증가하며 이러한 심리적 편향은 해당자산에 대한 과잉반응을 야기하며, 과신하는 투자자는 공격적인 거래를 하며 투자자의 관심은 높은 거래량을 야기한다. 본 연구는 국내 주식시장을 표본으로 투자자의 관심효과로 인한 과신과 거래량의 관계를 조사한다. 우리는 개별자산에 대한 투자자 과신의 정도를 측정할 수 있는 변수를 제시하고, 투자자의 과신 변수를 이용하여 국내 유가증권시장에서 투자자의 과신이 주식수익률에 미치는 영향을 분석한다. 그 결과, 투자자의 과신이 높은 자산일수록 낮은 미래수익률을 가지는 현상을 확인한다. 이러한 현상은 기업특성요인, 투자자의 과잉반응, 변동성과 도박성향 특성을 고려한 후에도 유지된다. 더 나아가, 주식수익률과 투자자의 과신의 관계는 개인투자자의 거래비중이 높은 주식과 시장에서 더욱 뚜렷하게 나타나며, 기간에 따라 투자자의 과신이 주식수익률에 미치는 영향이 변화한다. 따라서 국내 주식시장에서 개별자산에 대한 투자자의 과신은 주식수익률에 유의한 음의 영향을 미치며, 횡단면적으로 유의한 음의 예측력을 가짐을 확인한다.

According to behavioral finance, the effect of attention in the stock market has a significant impact on investor decision-making. Individual stocks of high investor attention increase investor irrational bias, such as overconfidence, and this psychological bias causes an overreaction to the asset. Overconfidence investors make aggressive transactions, and investor attention causes high trading volume. This study investigates the impact of investor overconfidence on stock returns in the Korean stock market based on the fact that investor confidence due to the effect of investor attention is closely related to trading volume. We present variables that can measure the degree of investor overconfidence in individual assets and analyze the effect of investor overconfidence on stock returns in the Korean securities market. In this study, we present the investor's overconfidence measure (IOC) using the highest trading volume over the past 12 months and the current trading volume. At the end of each month, we form a portfolio of deciles based on IOC to confirm the relationship between stock returns and investors' overconfidence. We focus on extreme portfolios to determine the impact of IOC on future returns. If the investor's overconfidence has a significant negative effect on the stock return, the investor's overconfidence has a relatively low stock return on an extremely high portfolio. As a result, the higher the investor's confidence, the lower the future return. It argued that the relationship between investor overconfidence and stock returns through a zero-cost portfolio strategy in which longs the lowest IOC portfolio and shorts the highest IOC portfolio, the IOC strategy has a significant positive return. This phenomenon is maintained even after considering firm characteristics factors, investor overreaction, volatility, and gambling tendency. The IOC's significant negative predictive power for stock returns is maintained through Fama-MacBeth cross-sectional regression analysis. Furthermore, We reaffirm the results of this study by the proportion of individual investors' trading, the separation of sub-periods, and the past period constituting the IOC. As a result of dividing the portfolio according to the proportion of individual investors, the higher the proportion of individual investors, the more pronounced the significant negative relationship between stock returns and investors' overconfidence. Moreover, the results of the KOSDAQ market where individual investors have a high proportion of transactions and high volatility has a higher return on IOC strategy. The sample period is divided into four sub-periods and analyzed. We confirm that the relationship between investor overconfidence and stock returns is relatively weakened after the financial crisis. In addition, the return on the IOC strategy increases as individual investors' market participation increases after the outbreak of coronavirus in the Korean stock market. These results reaffirm that the IOC fully reflects the psychological changes of individual investors. Finally, as a result of changing the past period constituting the IOC to 6, 18, and 24 months, the IOC strategy has a significant positive rate of return in all past periods. Accordingly, investors' overconfidence in individual assets in the domestic stock market has a significant negative effect on the stock returns and a significant cross-sectional negative predictive power. This study is meaningful in that it studied the effect of investor overconfidence on the stock market due to the effect of interest in individual assets in the domestic stock market. Furthermore, it has implications for related fields in that it presents measures indicating the degree of investor overconfidence in individual assets and presents empirical results. And report the results of controlling several variables closely related to investor overconfidence. These results support the existing argument that investors' overconfidence has a significant effect on stock returns.

본 연구는 신용정보 표본DB 원격분석시스템에서 제공하는 기업신용정보를 분석하여 기업의 채무불이행을 예측한다. 사업자 구분에 따라 분석대상을 나누고 표본DB에서 제공하는 기업신용정보의 활용에 따른 실험을 구성한다. 또한, 다양한 기계학습 기법을 모수 추정 방식에 따라 모수적 방법론, 비모수적 방법론, 준모수적 방법론으로 구분하여 예측성과를 비교한다. 표본DB 원시데이터를 활용한 분석보다 대출 및 연체 종류에 따라 가공한 자료를 활용하는 경우 각 기계학습 모형별 성능개선이 관측되었으나, 기업 차주의 특성정보와 기술신용평가 정보의 활용은 모형별 성능개선에 기여하지 못하였다. 모든 세그먼트에서 준모수적 방법론에 해당하는 심층신경망 모형에 대해 성능이 가장 우수한 것으로 확인되었으며, 트리계열이 아닌 비모수적 방법론의 경우 재현율이 낮게 관측되어 채무불이행 예측 문제에 적합하지 않았다. 기존 실무에서 사용되는 모수적 방법론을 활용한 경우보다 준모수적 방법론을 활용할 경우 분류 성능이 향상됨을 확인하였다. 본 연구는 실제 기업신용정보에 대해 구성된 표본DB를 활용하여 기업부실 예측을 시도한 최초의 연구이며, 기업신용정보를 활용하는 여신 금융기관과 신용정보사의 자료 활용 및 모형 구축에 대한 방향성을 제시한다.

This paper investigates the ability to predict corporate default rates using loan-sample data from the Korea Credit Information Service's financial big data open system (CreDB). The corporate loan from financial institution increases financial institution's credit exposure. Because measurement of the impact on the credit risk in the financial institution is used in determining the pricing model and structure of loan products, it is an essential factor for the financial institution that affects its profit structure. In terms of risk management, predicting delinquency using loan data is necessary for 5,000 Korean financial institutions. In several studies, bankruptcy forecasting was conducted on listed companies that disclosed financial and stock price information. However, this study increases the practical utility by extending the analysis target to individual entrepreneurs and small and medium-sized enterprises(SMEs). In addition, this study presents representative big data analysis results by utilizing loan, delinquency, and technology credit information of approximately 1.1 million corporations, which is 20% of almost 5.6 million domestic sole proprietors and non-listed corporations. For loan data, it includes ten monthly loan type codes and eleven overdue reason codes. Prediction targets are separated by individual and corporate entrepreneurs. Also, analyses are divided by use of the processed dataset. For efficient analysis, the data dimension was reduced by changing the table structure through nested iterative operations while expanding the variable composition from a table consisting of N rows to one column. To reflect the characteristics of the data as much as possible, exploratory data analysis and feature-engineering were performed to process the data. Also, classification models are classified by four groups using a parametric method that nine models train for classification. Group 1 consists of Logistic Regression and Linear discriminant analysis based on the parametric method, group 2 consists of several algorithms that calculate the distance for model learning. In addition, group 3 consists of tree-based algorithms, which are also non-parametric methods. Group 4 consists of the semi-parametric method, which is deep neural network. However, out of the total 438,697 corporations, 810 defaulted, accounting for only 0.2% of the forecast, so the target distribution is severely imbalanced. For this reason, before model fitting, under sampling of imbalanced data was performed. The bias of the sampled training and validation data is minimized by performing. K-fold cross validation as much as the level of K=5. Finally, the analysis result suggests a significant effect on classification performance when the processed data is used. However, this study suggests no significant effect on performance when loan owner's characteristics are included. Moreover, tech-credit rating (TCB) information gives any meaningful effect regarding the type of corporation. Also, classification with Deep Neural Network (DNN), which is based on the Semi-parametric method, makes the best performance of binary classification. Non-parametric and Non-tree based models are not appropriate methods for analyzing loan data. In the case of the DNN based on the semi-parametric methodology, the highest classification performance was confirmed for all analyses and entrepreneurs' classifications performed in this study. The neural network used in this study consists of 14 hidden layers. According to the neural network baseline design, the sigmoid function was applied to the activation function's initial value, the relu function was applied to the hidden layer, and optimization was performed through the Adam optimizer. In particular, the analysis of credit transaction information based on credit information of all financial institutions in Korea was conducted, and there is a possibility for alleviating information asymmetry of individual credit institutions regarding risk management targets. In addition, in the case of parametric methodologies used in classical studies and most used in practice, the average classification performance for major segments was inferior to that of semi-parametric methodologies. Furthermore, the difference between these performances is up to 16 percent. This paper suggests the direction of using loan-sample data. It is foundational research for financial institutions that are using loan data for credit risk management. It is necessary to expand research focusing on semi-parametric methodologies about corporate credit information analysis.