Forecasting the Equity Risk Premium in the Korean Stock Market: A Factor Analysis Approach

전성주(Sungju Chun) People&Global Business Association 2021 Global Business and Finance Review Vol.26 No.4

Purpose: This article investigates stock return predictability in the Korean stock market using the methodology of dynamic factor analysis. Design/methodology/approach: This article collects monthly data on the equity risk premium on the KOSPI and twelve financial and macroeconomic variables spanning from October 2000 to December 2020 and evaluates the forecasting performance of the dynamic factor predictive regression model by comparing in-sample and out-of-sample predictability with those of individual predictors. Findings: The article finds that the dynamic factor predictive regression exhibits statistically and economically significant in-sample predictability for the future equity risk premium for the KOSPI, as strongly as the best individual predictor can do. Also, the dynamic factor approach can outperform the benchmark historical average in out-of-sample predictability. The detailed analysis of the diffusion indexes reveals that each factor captures different information from various financial and macroeconomic variables relevant for return prediction and the diffusion indexes can deliver better forecasts of the future equity risk premium. Research limitations/implications: There exist different regression methods to combine forecasts comparable to the dynamic factor predictive model such as the forecast combination method by Rapach et al. (2010) and the bagging method by Inoue and Kilian (2008) and Jordan et al. (2017). The study proposes to compare the performance of these models with that of the dynamic factor predictive model in the Korean stock market as future research. Originality/value: The article is the first attempt to apply the dynamic factor predictive regression model to a large set of financial and macroeconomic data in Korea and evaluate its in-sample and out-of-sample predictability in comparison to those of individual predictive variables.

생명보험산업 보험료 성장률 예측계량모형 비교

전성주 ( Sungju Chun ),조영현 ( Younghyun Cho ) 한국금융연구원 2015 금융연구 Vol.29 No.3

In this article we evaluate the performance of forecasting models to predict the Korean life insurance premium growth rates. Comparisons are made for the Vector Auroregressive (VAR) predictive model, the multivariate leading indicator model, and the diffusion index model proposed by Stock and Watson (2002) against the Univariate Autoregressive (AR) predictive model as a benchmark. We compare each model’s predictability for the total premium incomes and the initial premium incomes of three types in the individual insurance; protection, endowment and annuity. The complete quarterly data spans from the second quarter of 1986 to the first quarter of 2014. The lag selection for AR and VAR forecasting models depends on the Bayesian Information Criteria (BIC) with the maximum number of lags set to 4. For the multivariate leading indicator model, we use 4 leading indicators of GDP growth rates, inflation, education, and age. In order to avoid data mining concerns, we select the variables that have been found to be the determinants of life insurance demands by previous studies. The diffusion index model is an approximate dynamic factor model that relates the future life insurance premium growth rates to a number of factors estimated by principal components using a large number of macroeconomic variables. The set of macroeconomic variables consists of 57 variables representing 6 main categories of macroeconomic time series: demand for final output; balance of payments and international trade; price indexes; money, interest rates and financial markets; labor, production and population; and world variables. We also include in the data set the variables related to the life insurance industry such as the industry’s total asset returns, claims paid and etc. In predicting total premium income growth rates, the AR predictive model produces the smallest mean squared predictive errors (MSPEs). For the premium incomes of protective insurance all the forecasting models have the MSPEs less than 3%. But they become unreliable in predicting the premium incomes of annuity with producing the MSPEs more than 10%. When we test the null hypothesis of no difference in MSPEs, it is rejected at 5% significance level for the VAR and the diffusion index predictive models when we forecast the total premium income growth rates of protective insurance. In predicting initial premium income growth rates, we find that there are no statistically significant differences in the MSPEs of each model. In addition, all the models have MSPEs more than 10% so that we may not be able to depend on any model to forecast in practice. We conclude that it may not be beneficial to take advantage of the information contained in macroeconomic variables for predicting life insurance premium growth rates. It may be due to the fact that most of the insurance contracts in Korea charge monthly premiums, which induces heavy autocorrelations among quarterly insurance premium data and makes an AR forecast very effective. Rho and Shin (1998) also found that macroeconomic variables are not likely to influence on insurance demands as is largely determined by insurer’s push-marketing. Policy holders cannot surrender their insurance contracts without heavy penalties, which makes them not so much responsive to macroeconomic environments. Lastly, life insurance demands are very sensitive to changes in insurance regulation and taxation, which could not be controlled for by our estimation procedure due to the lack of time series observations.

데이터 마이닝 효과를 감안한 한국 주식수익률 예측가능성 검정

전성주(Sungju Chun) 한국자료분석학회 2021 Journal of the Korean Data Analysis Society Vol.23 No.1

이 연구에서는 한국 주식시장의 대표적인 지수인 KOSPI 종합지수와 KOSPI 시가총액 규모별 대ㆍ중ㆍ소형주 지수에 대해 다양한 거시ㆍ금융변수들의 미래 주가수익률 예측력을 검증하고자 한다. 특히, 본 연구는 선행 연구에서 다룬 표본 내 예측가능성 검정법과 표본 외 예측가능성 검정법에 대해 Inoue, Kilian(2004)의 계량방법론을 바탕으로 Rapach, Wohar(2006)가 사용한 데이터 마이닝 부트스트랩 방법을 사용하여 많은 수의 예측변수들을 고려할 때 발생하는 대표적인 문제인 데이터 마이닝 효과를 통제한다. 최대검정치(maximal statistics)의 실증적 분포를 사용하여 데이터 마이닝 효과를 통제할 경우 기존 문헌연구 결과와 달리 KOSPI 종합지수와 KOSPI 대형주 지수에 대해서는 미래 수익률 예측력을 지닌 변수는 없는 것으로 나타났다. KOSPI 중형주 지수의 장기 수익률에 대해서는 주가순자산비율이 표본 내 예측가능성 귀무가설을 기각하고 있다. KOSPI 소형주 지수의 경우 배당수익률이 단기 수익률에 대한 표본 내 예측력을 지니고 있으며, 주가순자산비율이 장기수익률에 대한 표본 내 예측력과 표본 외 예측력을 지니고 있는 것으로 나타났다. The study performs an empirical analysis on the predictability of macroeconomic and financial variables for the KOSPI and the size portfolio returns. This study applies the data-mining bootstrap procedure proposed by Inoue, Kilian(2004) and Rapach, Wohar(2006) to control for the data-mining effects accompanying the predictive regressions involving many predictive variables in the in-sample and out-of-sample tests. As the study controls for the data-mining effects using the empirical distribution bootstrapped by maximal values of t-statistics, MSE-F statistics, and ENC-NEW statistics, there is no significant predictor for the returns of KOSPI and large-firm portfolio. For the medium-firm porfolio, price-to-book ratio has a significant in-sample predictability for the long-run returns. For the small-firm portfolio, price-to-book ratio can predict the long-run returns for both in-sample and out-of-sample while dividend-price ratio has only an in-sample predictability for the short-run returns.

경기북부 지역의 Richettsia 질환에 관한 임상적 고찰

전성주(Seoung Ju Chun),이기영(Kee Yeoung Lee),최두혁(Du Hyok Choi),양현억(Hyun Eog Yang) 대한내과학회 1989 대한내과학회지 Vol.36 No.4

N/A Rickettsiosis is one of the most prevailng acute febrile diseases in the northern part of Kyoung Ki Do during harvest-time. We experienced 36 cases of rickettsiosis in this area during 1978. Herein we report the epidemiological and clinical observations with laboratory findings. 1) The disease was prevalent in adult female farmers and the epidemic seasons were fall and early winter. 2) The serologic test revealed a positive rate of 67% (24 cases) in R. tsutsugamushi and 33% (12 cases) in R. typhi and two cases were seropositive to Korean Hemorrhagic Fever. 3) Clinical pictures of Tsutsugamushi disease were characterized by fever (88%), chills (71%), headache (71 %) and abdominal pain (50%) but were less severe in R. typhi as fever in 83% of cases, chills 67% and headache 67%. 4) On physical examination, abdominal tenderness (54 %), fever (38%), CVA tenderness (29%), tachycardia (25 %) and eschar (13%) were found but were less severe in R. typhi as abdominal tenderness in 25% of cases, fever 25%, CVA tenderness 17% and tachycardia 17%. 5) On laboratory examination, a left shift of leukocytes, atypical lymphocytes, anemia, microscopic hematuria and proteinuria were found. Blood chemistry findings were characterized by increased SGOT & SGPT and increased CPK & LDH. The chest X-ray revealed interstitial pneumonia in nine cases (25%).

경기 북부지역에서 발생한 한국형 출혈열의 임상적 특성

전성주(Seoung Ju Chun),전정수(Jung Su Chun),김재하(Jae Ha Kim),최두혁(Do Hyok Choi) 대한내과학회 1987 대한내과학회지 Vol.33 No.5

N/A Korean hemorrhagic fever which was recognized for the first time in Korea near the Demilitarized Zone between North and South Korea in 1951 during Korea War have spread to the southern part of the Korean peninsula and several hundreds of cases are clinically and serologically diagnosed each year. However the Northern Part of Kyoung Ki Do is still remained as epidemic area. In recent years, we experienced 55 cases of Korean Hemorrhagic Fever in the northern part of Kyoung Ki Do during the period, 1984-1986, and we found different clinical patterns of Korean hemorrhagic fever from those reported earlier. Here we report the epidemiological and clinical observations with laboratory findings. 1) The disease was prevalent in adult male farmers and the epidemic seasons were fall and early winter. The epidemic areas of the disease are mainly Yang Ju, Yeoncheon and Pocheon. 2) Clinical pictures of Korean Hemorrhagic Fever were characterized by headache (76%), abdominal pain (69%), nausea & vomiting (67%), fever (56%) and myalgia & general aching (49%). 3) On physical examination, fever (56%), abdominal tenderness (36%), tachycardia (35%), hypotension (27%), petechia (24%), conjunctival injection (22%) and facial flushing and edema (16%) were found. 4) On laboratory examination, leukocytosis, thrombocytopenia, left shift of leukocyte, toxic granule, proteinuria, microscopic hematuria and increased specific gravity were found. Blood chemistry findings were characterized by increased BUN & creatinine, increased SGOT & SGPT, hyponatremia, increased LDH & FDP and hyperkalemia. 5) Clinical forms of typical hemorrhagic fever were as follows: Typical oliguric type 49%, typical nonoliguric type 21% and atypical type 29%. 6) Double infection to Leptospirosia was found in 8 cases.

금융·거시경제변수들의 한국 주식수익률 예측가능성 검정 - 표본 내 검정과 표본 외 검정 결과 비교를 중심으로 -

전성주 ( Sungju Chun ) 보험연구원 2020 보험금융연구 Vol.31 No.1

본 연구는 한국 주식시장에서 12개의 주요 거시경제변수 및 금융변수들을 이용하여 미래 장·단기 주식수익률을 예측할 수 있는지 실증적으로 검정하였다. 특히, 선행연구에서 많이 사용되었던 표본 내 예측가능성 검정뿐만 아니라 표본 외 예측가능성 검정을 함께 시행함으로써 각 예측변수들의 예측력을 보다 강건하게 검정하였다. 이를 위해 내포모형(Nested model)의 예측치를 검정할 수 있도록 McCracken (2007)이 제안한 MSE-F 검정기법과 Clark and McCracken (2001)이 제안한 ENC-NEW 검정기법을 사용하였다. 이와 함께, 부트스트랩을 통한 재표본추출(Resampling through bootstrapping)을 통해 임계치와 p-value를 산출함으로써 주식수익률 예측에서 일어나는 소표본 편차(Finite-sample bias)와 장기수익률 잔차항의 자기상관성(Autocorrelation) 문제를 해결하였다. 검정 결과, 주가순자산비율(Book-to-market ratio) 변수가 표본 내 검정과 표본 외 검정에서 모두 주식수익률 예측력을 갖고 있는 것으로 나타나 가장 일관성 있는 예측변수로 나타났다. This study evaluates the predictive power of 12 financial and macroeconomic variables for Korean stock market returns of different horizons. Both the return predictability of in-sample and out-of-sample tests are considered to examine each variable’s predictive ability more robustly. For this purpose, this article employs the MSE-F statistic developed by McCracken (2007) and the ENC-NEW statistic developed by Clark and McCracken (2001) to compare nested forecast models. In addition, the bootstrapping procedure is applied for both in-sample and out-of-sample inferences to address the finite-sample bias and the autocorrelated disturbances from overlapping observations. As a result, the book-to-market ratio variable is found to be the most consistent and significant predictor as it rejects the null of no predictability for both in-sample and out-of-sample tests.

건강상태가 가계 금융자산 포트폴리오 결정에 미치는 영향 연구

전성주(Sungju Chun),이창우(Changwoo Lee) 한국경영학회 2015 한국경영학회 통합학술발표논문집 Vol.2015 No.08

본 연구는 가구의 외생적 건강상태가 가계 금융자산 포트폴리오 구성에 어떤 영향을 주는지 알아보기 위하여 실증분석을 실시하였다. 나쁜 건강상태는 개인의 소비에 대한 한계효용이나 위험회피도, 근로소득 등 자산 포트폴리오 결정요인들에 간접적인 영향을 줄 뿐만 아니라, 의료비용을 증가시켜 저축여력을 줄이고 자산 유동화 수요를 유발시킬 수 있기 때문에 가구의 건강상태는 가계 금융자산 포트폴리오를 결정하는 중요한 요소가 될 수 있다. 본 연구에서는 입원치료비를 건강상태에 대한 대리변수로 사용하여 건강상태가 위험자산 보유여부와 가계 금융자산 내 위험자산의 지분비율에 미치는 영향에 대해 실증분석을 실시하였다. 그 결과, 나쁜 건강상태로 인해 상대적으로 높은 입원치료비를 지출한 가구가 위험자산을 보유하지 않을 가능성이 높은 것으로 나타났으며 입원비 지출금액이 가계 금융자산 내 위험자산 지분비율에 음의 영향을 줄 가능성이 통계적으로 유의하게 나타났다. 이와 같은 연구결과는 우리나라 국민들이 국민건강보험제도를 통해 동일한 건강보험에 가입하고 있음에도 불구하고 가계 금융자산 포트폴리오를 구성함에 있어 외생적 건강상태에 여전히 영향 받고 있음을 보여 주고 있다.

건강상태가 가계 금융자산 포트폴리오 결정에 미치는 영향 연구

이창우 ( Changwoo Lee ),전성주 ( Sungju Chun ) 보험연구원 2016 보험금융연구 Vol.27 No.4

미국의 많은 선행연구들은 건강상태가 가계 금융자산 포트폴리오 결정에 어떤 영향을 주는지 알아보기 위하여 실증연구를 수행하였다. 특히, 건강상태와 포트폴리오결정은 개인의 위험회피성향이나 동기, 정보와 같은 잠재적인 개인의 특성에 의해 동시에 영향을 받을 수 있기 때문에 실증분석에서 이러한 잠재적인 개인특성을 통제하는 것이 중요할 수 있다. 본 연구에서는 미국과 다른 보건의료제도를 시행하고 있는 우리나라의 가구를 대상으로 건강상태가 위험자산 보유여부와 가계 금융자산 내 위험자산의 지분비율에 미치는 영향에 대해 실증분석을 실시하였다. 먼저, 가구의 건강상태를 대리하는 변수로 가계 총의료비에서 입원치료비가 차지하는 비중을 사용하여 건강상태가 가계 금융자산 포트폴리오 결정과 유의미한 상관관계를 가지고 있음을 발견하였다. 그러나 상관임의효과(Correlated Random Effects) 모형을 이용하여 잠재적인 개인특성을 통제한 결과 건강상태가 독립적인 외생변수로 포트폴리오 결정에 유의미한 영향을 미치지 못하는 것을 발견하였다. There have been many empirical studies in the United States to find a linkage between health status and portfolio choice of the individual households. In particular, it is important to account for the existence of unobserved characteristics such as risk attitudes, motivation, and information in the analysis because both health status and portfolio choice can be influenced by the unobserved heterogeneity. In this empirical study, we analyze how much the household portfolio decision is correlated with health status and whether this correlation is causal in Korea, where they have an obligatory national health care system different from the one in the United States. Most korean household panel data do not survey the subjective health status or, if ever, they do not survey the risky asset holdings of the households. We use the 2009-2013 waves of the National Survey of Tax and Benefit panel containing detailed information about the financial asset holdings of the households. As a proxy for the household`s underlying health status, we construct a ratio of inpatient hospital costs over total medical costs of a household that can reflect its “poor” health status. Employing the random-effect probit and tobit models, we find that there is a statistically significant correlation between health status and portfolio choice. However, once we control for unobserved heterogeneity by applying the correlated random-effects models, we find that health status no longer serves as a significant independent variable for the individual household to decide whether to hold a risky asset or how much financial assets to be allocated in a risky asset.

거시경제 환경변화에 따른 생명보험 수요의 구조변화 분석

전용식(Jeon, Yongsik),전성주(Chun, Sungiu) 글로벌경영학회 2021 글로벌경영학회지 Vol.18 No.5

본 연구는 생명보험 수요를 나타내는 생명보험 신계약가입금액과 KOSPI 종합주가지수, 실질 금리, 실업률, 산업생산지수 등 생명보험 수요에 영향을 미치는 다양한 거시경제변수들의 동태적인 관계를 추정하고 각 변수들과 생명보험 신계약가입금액의 장기 균형 관계를 실증적으로 분석하고자 한다. 특히, 국내 총소득 증가율이 정체되기 시작하는 2017년을 전후로 분석 기간을 나누어 위험자산인 주가지수와 대체 투자자산인 보험 수요 간 장기 상관관계를 집중적으로 분석하고 이들 관계에 어떤 변화가 발생하였는지를 중점적으로 살펴본다. 본 연구는 2003년부터 2020년 11월까지 월별 자료를 Pesaran and Shin(1999)이 제안한 ARDL 모형에 기반한 오차수정모형으로 추정하여 이들 변수 간의 장·단기 동적 관계를 분석한다. 또한, Pesaran et al.(2001)이 제안한 Bounds 검정을 통해 변수들의 I(0) 혹은 I(1) 여부와 관계없이 공적분 관계를 강건하게 검정하고 모형의 잔차에 대한 누적합 검정(CUSUM 검정과 CUSUMQ 검정)을 실시하여 모형의 안정성을 검정한다. 실증분석 결과는 다음과 같다. 첫째, 거시경제변수들과 생명보험 수요 사이에는 공적분 관계가 존재한다. 둘째, 전체 기간과 2017년 이전 자료의 경우 종합주가지수는 생명보험 수요와 정(+)의 관계가 있으며, 산업생산지수는 부(-)의 관계가 있다. 셋째, 2017년 이후에는 실질 금리가 생명보험 수요와 매우 뚜렷한 정(+)의 관계를 보이며 종합주가지수는 부(-)의 관계를 나타내는 것으로 바뀌었다. 본 연구의 검증 결과 생명보험 수요와 거시경제변수들 사이에는 장기 균형 관계가 존재하며, 2017년을 기점으로 국내 총소득 증가율이 정체되면서 소득 효과는 약해지고 위험자산인 주식과 대체 투자자산인 보험 사이에 대체 효과가 강해지고 있는 것으로 나타났다. 그리고 저금리 현상이 심화함에 따라 보험 소비자들의 실질 금리 변화에 대한 민감도가 높아진 것으로 분석되었다. 본 연구는 이러한 결과를 토대로 향후 연구과제로 기간별로 구조 변화를 추정할 수 있는 계량모형을 적용하여 이러한 변화를 고려한 장기 균형 관계를 추정할 수 있는 연구를 제안한다. This study performs an empirical analysis on the long-term relationship between the life insurance demand and the macroeconomic variables such as KOSPI index, real interest rate, unemployment rate, and industrial production by estimating the dynamic ARDL model. In particular, we split the data into two subperiods before and after 2017, the year after which the rate of domestic income growth became sluggish and investigate whether there is any change in the long-term relationship. The study uses the monthly data from January 2003 to November 2020 to estimate the ARDL model proposed by Pesaran and Shin(1999) and explore the dynamic long-term and short-term relationships among the macroeconomic variables. We apply the Bounds testing procedure proposed by Pesaran et al.(2001) to perform a robust test for the co-integration regardless of whether the variables are I(0) or I(1), and apply the CUSUM and CUSUMQ tests to examine the stability of the estimated models. The main results are following. First, we find that there exists a cointegrating relationship among the variables. Second, the analyses based on the first subperiod and the entire data indicate that the insurance demand is positively related with the KOSPI index while it is negatively related with the industrial production. Thrid, after 2017 the long-term relationship has been changed so that the insurance demand becomes negatively related with the KOSPI index and positively related with the real interest rate that had previously not been significantly related with the insurance demand. Our results indicate that there occurred an important change in the long-term relationship between the life insurance demand after 2017 in which diminishing domestic income growth rate entails a substitution effect between the stock and the insurance while the consumers become sensitive to the change in the interest rate in the “low interest rate” environment. The study proposes a future research that applies an econometric model to account for structural changes in the long-term relationships.

중년기 기혼 성인의 가족건강 구조모형

김희숙(Kim Hee Sook),신동수(Shin Dong-Soo),전성주(Chun Sung Joo),이성희(Lee Sung Hee) 한국성인간호학회 2008 성인간호학회지 Vol.20 No.4