Earticle

본 연구는 국내 주식시장의 위험-수익 관계를 종합적으로 분석하고 투자전략 측면에 서 활용하는 방안을 논의한다. 이를 위해 장기 및 단기 총 변동성, 체계적 변동성, 고유 변동성, 시장베타 8개의 위험측도를 사용하여 1990년 7월부터 2021년 12월까지 의 국내 유가증권시장의 위험-수익 관계를 분석하고, 그 결과를 이용하여 국내 저위험 관련 ETF의 수익성을 개선하기 위한 다양한 실험을 수행하였다. 주요 결과는 다음과 같다. 첫째, 전체 기간에서 국내 주식시장의 저위험 이상현상은 일별 수익률을 사용하 여 측정한 단기 변동성 지표들에 대해서만 유의하였다. 저위험 프리미엄의 크기는 롱-숏 포트폴리오는 소형 성장주 그룹에서, 매입 포트폴리오는 소형 가치주 그룹에서 크게 나타났다. 둘째, 위험-수익 관계와 그 유의성은 위험측도와 기간별로 상이하였 다. 셋째, 위험측도별 저위험 프리미엄 사이의 인과관계를 분석한 결과 국내 주식시장 의 저위험 이상현상은 고유 변동성에 의한 것임을 확인하였다. 마지막으로, 실증 분석 결과를 바탕으로 국내 주식 저위험 ETF 기초 지수의 수익성을 개선하기 위한 31가지 수정안을 분석한 결과 투자성과가 유의하고 강건하게 향상되는 것을 확인하였다. 본 연구는 강건성과 위험 측도별 저위험 프리미엄간 관계 등 다양한 측면에서 국내 주식 시장의 저위험 이상현상에 대한 이해도를 높이고, 투자상품 개발을 위해 실증 결과를 활용하는 실무적인 방안을 구체적으로 제시하였다는 의의를 갖는다.

This study comprehensively analyzes the patterns of risk and return relationship in the Korean stock market and applies them to portfolio management strategies. Total eight number of risk measures - long and short term total volatility, systematic volatility, idiosyncratic volatility, and market beta - are employed to analyze the risk-return relationship in the KOSPI market from July 1990 to December 2021. Furthermore, based on the results of the analysis, we conduct various empirical experiment to improve the profitability of the low-risk related ETFs traded in the Korean stock market. The key findings are as follows: First, we observe that the low-risk anomalies are significant over the entire period only for the short-term volatility measures. Small-growth groups show greater low-risk premiums for long-short portfolios, whereas small-value groups exhibit higher low-risk premiums for long-only portfolios. Second, the low-risk anomalies are not robust across the risk measures and measurement periods. Third, as a result of the causality analysis between the risk premiums from the different risk measures, we find that the low-risk anomalies in the Korean stock market are primarily driven by idiosyncratic volatility anomaly. Lastly, we find that the definition of underlying indices of the low-risk related ETFs currently traded in the Korean stock market does not adequately reflect the empirical patterns of the risk-return relationship. If we modify them in accordance with our empirical findings, the profitability of the low-risk indices significantly improve with robustness. This study has contributions in that it deepen the understanding of the low-risk anomalies in the Korean stock market and proposes practically useful applications of them to enhance portfolio management strategies.

Motivated by the mixed findings of the existing literature, we investigate the performance of the volatility targeting strategy in four Asian equity markets based on a block-bootstrap simulation. We found no clear evidence of outperformance of the strategy against a 50/50 constant-mix portfolio, except in the Chinese market where volatility and tail risk are highest. The lack in performance has been attributed to an inability to detect the negative relationship between volatility and returns by all of the volatility forecasting models we consider. This may affect the capacity to reduce downside risk and tail risk but consistently causes a substantial drag on performance. Owing to forecasting errors, the strategy appears to perform considerably better and easier to implement in practice when using a model with low sensitivity to volatility changes along with a rebalancing buffer.

이 연구에서는 국내 코인거래의 80% 이상을 차지하는 가상자산인 비트코인, 이더리움, 리플, 에이다, 도지코인의 2017년 9월부터 2022년 12월까지의 일별 거래자료를 이용해, 코인별로 원화 마켓과 달러화 마켓 간의 가격불균형을 측정하고, 여기서 공통요인을 추출해 가격불균형 지수를 만든 후, 이것의 시계열 특성과 관련 요인을 분석하였다. 분석결과, 가격불균형 지수는 원화가격 지수와 동조화된 움직임을 보여 가상자산 원화 가격이 평가절상(평가절하)될 때 가격불균형의 공통요인이 증가(감소)함을 알 수 있었다. 가격불균형 지수는 특정금융정보법 개정안 시행후 통계적으로 유의하게 상승한 것으로 나타나, 가상자산거래에 대한 규제가 도입되면서 차익거래가 위축되어 가격불균형이 빠른 시간 안에 해소되지 못하였음을 알 수 있었다. 가격불균형의 공통요인은 국내‧외 코인시장의 투자심리 및 국제금융시장과 관련이 높고, 코인의 대체재로 여겨지는 실물자산 및 주식의 움직임과는 관련이 없는 것으로 나타났다. 한편, 개별 코인에서 관찰되는 고유한 가격불균형은 해당일의 원화수익률, 원화거래대금 등과 같은 코인별 수요 요인에 영향을 받는 것으로 나타났다. 차익거래가 발생하지 않은 것으로 추정되는 코인-거래일 표본이 차익거래가 발생한 것으로 추정되는 표본에 비해 김치 프리미엄과 원화거래대금이 크고, Amihud 척도가 적은 것으로 나타나, 원화 마켓 가격불균형이 발생할 때 가격차이에 순응하는 모멘텀 트레이딩이 발생하는 경향이 강함을 알 수 있었다.

Virtual assets have experienced dramatic ups and downs over the past few years, since an anonymous developer named Satoshi Nakamoto introduced Bitcoin in 2008. Bitcoin is a peer-to-peer electronic cash system based on blockchain technology, enabling payments and transactions verification without a centralized custodian. Despite controversies surrounding Bitcoin's currency status (IMF, 2016; Yermack, 2015; Hendrickson et al., 2015), there is a consensus that virtual assets function as alternative investment assets providing portfolio diversification, inflation hedging, and exposure to new technology effects. (Glaser et al., 2014; Briere et al., 2015; Bouri et al., 2017; Choi, 2022) Consequently, the market for virtual assets has evolved significantly, with over 26,346 types of virtual assets issued by the end of June 2023, and their total market capitalization estimated to exceed $11.7billion (coinmarketcap). Despite recognition of the legitimacy of virtual assets by major countries such as the US, EU, Japan, and Korea, attempts to bring them under policy consideration have led to differing regulatory systems and relative scales of supply and demand across countries, resulting in price deviations. Notably, the daily average price ratio between the USD and KRW Bitcoin market exceeded 15% from December 2017 to February 2018, a phnomenon referred to as the 'Kimchi premium' in both popular press and academic research (Makarov ans Schoar, 2020; Yang, 2019). These deviations from the law of one prices is even more pronounced due to the untraceable cross-border trading of virtual assets via peer-to-peer (P2P) transaction facilitated by advanced blockchain technology (Nakanoto, 2008). It seems that constraints and frictions in the arbitrage capital flow have led to market segmentation and significant, persistent price deviations between countries. However, there has not been a systematic analysis of the arbitrage trading across borders and the price deviation between countries due to the lack of transaction data in the virtual asset market. This paper attempts to fill this gap in the previous studies. In this study, daily trading data from September 2017 to December 2022 for virtual assets such as Bitcoin(BTC), Ethereum(ETH), Ripple(XRP), Cardano(ADA), and Dogecoin(DOGE) were used. For each coin, the price deviations between the KRW domestic and the USD international market were calculated by subtracting the KRW/USD spot exchange rate from each coin's exchange rate, which is the ratio of its KRW closing price to its USD closing price on the same trading day. Subsequently, a price disparity index was constructed by extracting a common factor from the price deviations of each coin using principal componant analysis(PCA) methodology. This study examined the time-series characteristics of the disparity index and the economic factors related to the index. Our analysis showed that the disparity index co-moved with the KRW price index, indicating that price deviations occur during periods of a rapid appreciation in the KRW coin market. The disparity index increased significantly after the implementation of a new regulation such as the Travel Rule, implying that frictions in the arbitrage activities lead to a market segmentation and price deviations across countries. In the multi-variate regression analysis, where the price disparity index was regressed on cryptocurrency market sentiment variables, financial market variables, commodity and stock market variables, the disparity index was closely related to investor sentiment in both the KRW and USD coin market, as well as to the movement of international financial markets, while showing no significant relationship with commodity and stock markets. Next, for each coin, the idiosyncratic price deviation was extracted by subtracting the disparity index from each coin's price deviations. The unique price deviations between the KRW and the USD market observed in individual coins were influenced by demand factors specific to each coin, unlike the disparity index. Specifically, the idiosyncratic price deviations on the previous day, daily returns, and KRW trading volume exhibited positive relationships with idiosyncratic price deviations in the four coins, and these results were maintained even after controlling for other variable. Meanwhile, using the daily change in price deviation as a proxy variable, this study divided the entire sample into two groups: one estimated to have arbitrage trading and one without, and examined whether arbitrage trading contributes to resolving the price disparity. The sample estimated to have no arbitrage trading showed a higher Kimchi premium than the sample estimated to have arbitrage trading, indicating that arbitrage transactions contribute to resolving price deviations between the KRW and the USD market. The sample without arbitrage trading showed larger KRW trading volume and lower Amihud measure compared to the sample with arbitrage trading, exhibiting a stronger tendency towards momentum trading that follows price deviations, rather than engaging in contrarian trading for arbitrage purposes.

We investigate the impact of price limit expansion on stocks’ realized volatility, uniquely integrating both magnet effect and correction effect into our model. Our findings reveal that price limit expansion may increase or decrease stock volatility, with the magnet effect and stocks’ inherent volatility levels as primary driving forces, and the correction effect playing a minor role. Typically, a decrease in volatility occurs when a stock’s inherent volatility is moderate (i.e., neither too high nor too low), while an increase takes place when it is relatively high. We offer an insightful explanation of these contrasting behaviors in relation to the magnet effect and stocks’ inherent volatility.