초록 열기/닫기 버튼
본 연구는 한국의 Shadow rate(잠재금리)를 추정하고 이것이 금융시장에 미치는 영향을 파악하고자한시도이다. 명목 금리는 0%이하로 내려갈 수 없으나, 기관투자자를 비롯한 여러 투자자들이 시장에서 인지하고있는 금리는 이보다 훨씬 낮을 수 있다. 시장이 인식하고 있는 금리는 체감금리로써 시장의 여러 지표 및상황에 선행될 가능성이 높으며 이러한 금리를 Black(1995)은 Shadow rate로 정의한바 있다. Shadow rate를 시장이 인식하는 금리수준으로 표현하여 추정하기 위해서는 당연히 시장자료를 이용하여야할 것이다. 이에 본 연구에서는 한국의 Shadow rate를 추정하기 위해 시장 상황을 나타내는 대표적인 지표인채권 수익률의 금리 기간구조를 이용하였다. 또한 Shadow rate는 실제로 관찰되는 이자율이 아니기 때문에이를 추정하는 방법은 기존의 모형으로는 불가능하다. 이에 본 연구에서는 관찰 불가능한 변수를 추정할 수있는 칼만필터(Kalman Filter)방식을 이용하여 하여 추정하였다. 그리고 추정한 Shadow rate가 한국의금융시장에는 어떠한 영향을 주는지 금융위기 이전, 금융위기 이후, 전체기간으로 나누어 분석하였다. 본 연구의 주요 결과를 요약하면 다음과 같다. Shadow rate를 추정한 결과 2016년 7월 기준으로 국고채이자율보다 1% 가량 더 낮은 0.355%의 값을 얻을 수 있었고 이렇게 추정한 Shadow rate가 금융시장에 어떠한영향을 미치는지 파악하기 위해 CD91, CP91, 원/달러환율, KOSPI지수, VKOSPI를 이용하여 금융시계열로분석하였다. 그 결과 그랜저인과관계 검정에서 Shadow rate가 거의 대부분 변수들을 인과하는 양상을 보였다. 또한 VECM 검정과 충격반응함수에서 CD91과 CP91은 전부 양(+)의 결과를 얻을 수 있었는데, 이는 Shadow rate가 단기금리의 이자율 상승에 큰 영향을 미친다는 사실을 의미한다. 그리고 VKOSPI에 음(-)으로 영향을주고 KOSPI지수는 대부분 양(+)의 반응을 보였는데, 이는 Shadow rate가 주식시장을 안정적으로 유지시키는기능을 지니고 있음을 보여준다.
This study is an attempt to estimate the shadow rate of Korea and its effect on financial markets. Real interest rates cannot be below 0%, but many market participants including institutional investors may perceive hypothetical negative interest rates below a zero-lower bound in the market. If the negative shadow rate is observed in the market, asset management strategies depending on the market interest rates are expected differently as they are seen in an ordinary situation. Recently, low interest rates have been observed in various countries where governments try to stimulate their economic conditions. Even though the U.S. Federal Reserve is expected to raise interest rates in the near future, the U.S. interest rates are yet 0% until recently. Korea has also lowered its base rate to encourage the economy, so lowering the market interest rate. Then, according to Black (1995), a shadow rate is defined as the level of interest rates that the market may perceive when policy (nominal) interest rates are low. The interest rates recognized by the market are likely to be preceded by various economic indicators and conditions of the market. Therefore, the shadow rate would have a close relationship with investment in financial assets as sell well as other economic variables. In the United States, the shadow rate is estimated as follows. First, Bernanke, Boivin, and Eliasz (2005) analyzed the market situation by adding economic variables to the VAR (Vector Autoregression) model. The shadow rate is argued as a key indicator to predict the future movements of economic variables. Therefore, it is possible to implement monetary policy with a shadow rate. Wu and Xia (2014) noted that investment should be made with the effects of the stimulus caused by the shadow rate. In order to estimate the shadow rate as a market-recognized interest rate level, it would be necessary to use market data. In this study, we use the interest rate structure of the bond yield, which is a representative indicator of market situation, to estimate the shadow rate in Korea. Also, since the shadow rate is not actually observed in the market, it seems not to be able to estimate the shadow rate by implementing existing models with observed variables. Therefore, we employ the Kalman filter method to estimate the unobservable variables. And, we discusse on how interest rates are recognized in the market and how this shadow rate affects the financial market. The main results of this study are summarized as follows. As a result of estimating the Shadow Rate, the value of 0.355%, which is 1% lower than the Treasury bond interest rate, was obtained as of July 2016. In the United States, the Shadow Rate is estimated as a negative number, but Korea still has positive results because the US interest rate is significantly lower than that of Korea. In order to determine the impact of the estimated shadow rate on the financial market, we analyze the financial time series using CD91, CP91, KRW/USD, KOSPI, and VKOSPI. The shadow rate in the Granger Causality test causes almost the most economic variables. The tests with a VECM model and impulse responses present that CD91 and CP91 were all positive indicating that the shadow rate has a significant effect on the rise of short-term interest rates. Forecasts of shadow rates are also expected to be very important indicators of asset pricing, risk management, and economic outlooks. There are some limitations on this study. This study estimates the shadow rate based on two factors, level and slope. However, Krippner (2015) estimated the shadow rate with three factors by adding bow, level and slope. Therefore, if we add the bow to the model and estimate a shadow rate, new results may be presented. And, this study estimates the parameters of a shadow rate based on the 1-year, 3-year, 5-year, 10-year and 20-year maturities of Korean bond data. The estimation based on the shorter maturity bonds such as 3-month and 6-month Korean bonds could be implemented by subsequent studies.
키워드열기/닫기 버튼
Shadow Rate, Bonds, Kalman Filter, Term Structure, Financial Time Series
