• 응용통계연구
• /
• 제28권5호
• /
• pp.1013-1024
• /
• 2015

본 논문에서는 다변량 DCC(dynamic conditional correlation) GARCH 모형에서 동태적 상관계수를 추정하기 위한 대표적 방법인 쌍별 추정법과 다차원 추정법의 효율성을 비교한다. 이를 위하여 금융 시장의 변동성을 반영하는 다변량 시계열을 생성하고 이에 대한 DCC GARCH 모형을 수립 및 추정하는 시뮬레이션을 실시하였다. 또한 KOSPI 200 섹터지수를 이용하여 포트폴리오를 구성하고 이의 변동성 추정 및 VaR 계산을 통하여 동태적 상관계수 추정에 대한 정확성을 평가하였다. 그 결과로서, 전반적으로 다차원 추정법이 쌍별 추정법보다 우수함을 발견하였다. 특히, 다차원 추정법에서 상대적으로 상관관계가 낮은 시계열을 추가할수록 쌍별 시계열에 대한 동태적 상관계수 추정의 정확성을 높여줌을 발견하였다.

• Communications for Statistical Applications and Methods
• /
• 제21권5호
• /
• pp.445-459
• /
• 2014

This paper investigates the dependence structure of Korean financial markets (stock, foreign exchange (FX) rates and bond) using copula-GARCH and dynamic conditional correlation (DCC) models. We examine GJR-GARCH with skewed elliptical distributions and four copulas (Gaussian, Student's t, Clayton and Gumbel) to model dependence among returns, and then employ DCC model to describe system-wide correlation dynamics. We analyze the daily returns of KOSPI, FX (WON/USD) and KRX bond index (Gross Price Index) from $2^{nd}$ May 2006 to $30^{th}$ June 2014 with 2,063 observations. Empirical result shows that there is significant asymmetry and fat-tail of individual return, and strong tail-dependence among returns, especially between KOSPI and FX returns, during the 2008 Global Financial Crisis period. Focused only on recent 30 months, we find that the correlation between stock and bond markets shows dramatic increase, and system-wide correlation wanders around zero, which possibly indicates market tranquility from a systemic perspective.

• 무역학회지
• /
• 제47권4호
• /
• pp.215-231
• /
• 2022

Analysis and forecasting of the Baltic Capsize Index (BCI) is important for managing an entity's losses and risks from the uncertainty and volatility of the fast-changing maritime transport market in the future. This study conducted volatility transition analysis through the GARCH model, using BCI which is highly related to steel raw materials. As for the data, 2,385 monthly data were used from March 1999 to March 2021. In this study, after basic statistical analysis, unit root and cointegration test, the GARCH, EGARCH, and DCC-GARCH models were used for volatility transition analysis. As the results of GARCH and EGARCH model, we confirmed that all variables had no autocorrelation between the standardized residuals for error terms and the square of residuals, that the variability of all variables at this time was likely to persist in the future, and that the variability of the time-series error term impact according to Iron ore trade (IoT). In addition, through the EGARCH model, the magnitude convenience of all variables except the Iron ore price (IOP) and Capesize bulk fleet (BCF) variables was greater than the positive value (+). As a result of analyzing the DCC-GARCH (1,1) model, partial linear combinations were confirmed over the entire period. Estimating the effect of variability transition on BCF and C5 with statistically significant linear combinations with BCI confirmed that the impact of BCF on BCI was greater than the impact of BCI itself.

• 부동산연구
• /
• 제28권2호
• /
• pp.21-34
• /
• 2018

본 연구의 목적은 거시경제 및 금융변수가 주택시장과 주식시장 사이의 상관관계에 미치는 영향을 분석하는데 있다. 이를 위해 본 연구에서는 표본기간(2004년 1월부터 2017년 11월) 동안의 월별 종합주택 매매가격지수와 KOSPI지수를 이용하였고, 시간가변적 상관관계는 AG-DCC GARCH 모형을 이용하여 추정하였다. 추정결과, 상관계수의 비대칭성을 나타내는 추정계수가 모두 통계적으로 유의한 양(+)의 값을 가지는 것으로 나타났으며, 이는 주택시장과 주식시장의 양(+)의 충격보다는 음(-)의 충격이 주택시장과 주식시장 사이의 상관계수를 더 증가시킨다는 것을 의미한다. 또한, 거시경제 및 금융변수가 상관계수에 미치는 영향을 분석한 결과 기간스프레드는 상관계수에 음(-)의 영향을 미치고, 신용스프레드는 상관계수에 유의한 양(+)의 영향을 미치는 것으로 나타났다. 이러한 결과들은 위험프리미엄 효과로 해석이 가능하다. 기간스프레드가 하락하거나 신용스프레드가 상승하는 경우, 즉, 미래 경기가 하락할 것으로 예상되는 경우 우리나라 주택시장과 주식시장의 변동성이 증가하고 투자자들은 더 높은 위험프리미엄을 요구하게 된다. 이로 인해 주택가격과 주가는 하락하게 되고, 두 시장 사이의 상관계수는 상승하게 된다는 것이다. 또한 이 결과는 두 시장이 서로 헤지의 역할을 하기는 어렵다는 것을 의미한다.

• The Journal of Asian Finance, Economics and Business
• /
• 제8권2호
• /
• pp.759-770
• /
• 2021

Using a DCC - GARCH model analysis, this paper examines the existence of financial contagion from the U.S. stock market to the Vietnamese and the Philippine stock markets during the global financial crisis and the COVID-19 pandemic crisis. We use daily data from the S&P 500 (U.S.), VN-Index (Vietnam), and the PSEi (the Philippines). As a result, there is no evidence of contagion from the U.S stock market to the Philippine stock market that can be found during global financial crisis, while the Vietnamese market is influenced by this effect. Besides, both these developing stock markets (the Vietnamese and Philippine stock markets) are influenced by the contagion effect in COVID-19 pandemic crisis. Another finding is that the contagion effect during the coronavirus pandemic crisis in Vietnam is smaller than that during the global financial crisis, however, the opposite is the case for the Philippines. It is noticed that the Philippines seems to be more affected by the contagion effect from the COVID-19 pandemic than Vietnam at the time of this study. Because financial contagion is important for monetary policy, asset pricing, risk measurement, and portfolio allocation, the findings in this paper may give some useful information for policymakers and investors.

• 무역학회지
• /
• 제48권4호
• /
• pp.189-208
• /
• 2023

본 논문은 한국 ETS시장, 에너지시장 및 주식시장 간의 동태적 조건부 상관관계를 분석하였다. 본 논문은 2015년 2월 2일부터 2021년 12월 30일까지의 한국 탄소배출권 거래가격, WTI원유 선물가격, 코스피지수의 일별 자료를 이용하여 실증분석하였다. 우선, GARCH 모형을 사용하여 세 시장의 변동성에 대해 분석한 후, 이변량 DCC-GARCH 모형을 사용하여 세 시장 간의 동태적 조건부 상관관계를 연구하였다. 연구결과는 다음과 같다. 첫째, 한국의 ETS시장이 주식시장보다 투자 수익률과 투자 위험도가 높은 것으로 나타났다. 둘째, 한국 ETS시장의 수익률 변동성이 외부 충격의 영향을 가장 많이 받고, 시장 자체의 변동성 정보로부터 받는 영향이 가장 작은 것으로 나타났다. 셋째, 한국 ETS시장이 WTI원유 선물시장보다 주식시장과의 상관관계의 지속성이 더 강한 것으로 나타났다. 본 논문은 한국 ETS시장, 에너지시장 및 주식시장 간의 상관관계를 분석하여 한국 ETS시장의 금융화 수준이 상당히 낮은 것을 확인하였다.

• Journal of Korea Trade
• /
• 제27권5호
• /
• pp.167-198
• /
• 2023

Purpose - This paper aims to compare and analyze the co-movements and information spillover effects between the international commodity futures markets and the South Korean stock markets during the COVID-19 and the 2008 financial crises. Design/methodology - The DCC-GARCH model is used in the co-movements analysis. In contrast, the BEKK-GARCH model is used to evaluate information spillover effects. The statistical data used is from January 1, 2005, to December 31, 2022. It comprises the Korea Composite Stock Price Index data and daily international commodity futures prices of natural gas, West Texas Intermediate crude oil, gold, silver, copper, nickel, soybean, and wheat. Findings - The results of the co-movement analysis were as follows: First, it was shown that the co-movements between the international commodity futures markets and the South Korean stock markets were temporarily strengthened when the COVID-19 and 2008 financial crises occurred. Second, the South Korean stock markets were shown to have high correlations with the copper, nickel, and crude oil futures markets. The results of the information spillover effects analysis are as follows: First, before the 2008 financial crisis, four commodity futures markets (natural gas, gold, copper, and wheat) were shown to be in two-way leading relationships with the South Korean stock markets. In contrast, seven commodity futures markets, except for the natural gas futures market, were shown to be in two-way leading relationships with the South Korean stock markets after the financial crisis. Second, before the COVID-19 crisis, most international commodity futures markets, excluding natural gas and crude oil future markets, were shown to have led the South Korean stock markets in one direction. Third, it was revealed that after the COVID-19 crisis, the connections between the South Korean stock markets and the international commodity futures markets, except for natural gas, crude oil, and gold, were completely severed. Originality/value - Useful information for portfolio strategy establishment can be provided to investors through the results of this study. In addition, it is judged that financial policy authorities can utilize the results as data for efficient regulation of the financial market and policy establishment.

• 재무관리연구
• /
• 제26권3호
• /
• pp.199-226
• /
• 2009

본 논문은 우리나라 주식시장과 외환시장의 기대 수익률과 조건부 변동성간의 시계열적 관계를 2요인 자본자산가격결정모형(two-factor ICAPM)을 이용하여 실증 분석하였다. 주가와 환율의 조건부 분산은 GARCH 모형과 비대칭성을 반영한 GJR(1993) 모형으로 추정하였으며, 주가와 환율과의 조건부 공분산은 Bollerslev(1990)의 일정 상관관계(CCC) 모형과 Engle(2002)의 동태적 조건부상관관계(DCC) 모형을 이용하여 추정하였다. 실증 분석모형은 MGARCH-M 모형을 사용하였으며, 추정방법은 준최우추정법(QMLE)을 사용하였다. 실증 분석결과 외환위기 이후에 주식시장의 기대 수익률은 주가의 분산에 대해, 그리고 환율과의 공분산에 대해 유의한 음(-)의 관계를 갖는 것으로 나타났다. 그러나 외환시장에서 기대 수익률은 조건부 분산과 조건부 공분산에 대해 유의하지 않은 것으로 나타났다. 조건부 분산의 추정에서는 GJR 모형이 GARCH 모형에 비해 더 적합한 것으로 나타났다. 그리고 DCC 모형이 CCC 모형에 비해 설명력이 더 높은 것으로 나타났다. 본 논문의 분석결과는 주식시장에서 환율 변동이 위험 요인으로 작용하고 있기 때문에 포트폴리오 구성이나 위험 관리 등에서 환율 변동을 고려할 필요가 있고, 변수들간의 상관관계는 시변하는 모형을 사용할 필요가 있음을 시사한다.

• 무역학회지
• /
• 제47권3호
• /
• pp.141-156
• /
• 2022

This study intends to examine the regional blocs of the international crude oil futures market by analyzing the dynamic conditional correlation between the international crude oil futures markets using the DCC-GARCH model. For statistical data, from April 2, 2018 to March 31, 2022, international crude oil futures prices such as Europe, the United States, China, and Dubai were used. To summarize the results of the study, first, the phenomenon of regional blocs in the international crude oil futures market is occurring, and it is found that it is gradually strengthening as time goes by. Second, it was found that the dynamic correlation of the international crude oil futures market is temporarily strengthened when a supply-demand imbalance problem occurs due to a global shock. Third, it was found that the volatility of the Chinese crude oil futures market affects the international crude oil futures market. This study confirmed that the regional blocs phenomenon in the international crude oil futures market is strengthened as time goes by. In particular, it suggested that China's influence in the international oil market would increase.

• 무역학회지
• /
• 제47권1호
• /
• pp.99-114
• /
• 2022

This paper analyzed the dynamic conditional correlation between the carbon emission trading prices of Korea, China, EU, New Zealand. This paper was analyzed using the daily data of carbon emission trading prices of each country from January 12, 2015 to January 13, 2021 using the DCC-GARCH model. Summarizing the research results, first, the dynamic conditional correlation between carbon emission trading prices in the EU, Korea, and China, excluding New Zealand, was strong, indicating that there was a co-movement phenomenon. Second, it was found that carbon emission trading prices in major countries have a stronger tendency to co-movement due to global shocks. Third, it appears that the dynamic conditional correlation between the carbon emission trading prices of Korea and China is gradually strengthening. This study confirmed that the co-movement between carbon emission trading prices in Korea and other countries gradually intensified as time passed. In particular, it is meaningful in suggesting the implication that the phenomenon of co-movement between carbon emission trading prices in Korea and China is gradually intensifying.