본 연구에서는 한국과 미국의 월별 자료를 이용해 금리와 환율, 그리고 물가간의 인과관계를 동태적으로 살펴보았다. 먼저 소규모 개방경제를 반영해 해외변수를 외생적으로 가정한 8변수VAR 모형을 통해 충격반응분석을 실시한 결과 이자율평가 가정이 의미하는 바와 같이 콜금리인상 충격은 원/달러 환율을 상승시키는 반면 연방자금금리 인상 충격은 원/달러 환율을 하락시킨다. 또한 구매력평가 가정이 보여주는 바와 같이 원/달러 환율과 미국 소비자 물가의 상승충격은 국내 소비자 물가를 상승시킨다. 뿐만 아니라 분석대상기간을 10년씩 이동시키면서 이동회귀분석을 실시한 결과에 따르면 최근으로 올수록 금리, 환율, 물가간의 동태적 인과관계가 이들 가정과 부합하는 방향으로 움직이고 있다. 경제이론에 근거한 부호제약조건을 사용한 VAR이나 베이지언 VAR 모형의 추정결과도 위에서의 충격반응결과를 지지한다.

The study examines the dynamic causal relationship between interest rates, exchange rates, and prices, using the monthly data from January 1991 to February 2015. The data considered in the paper are U.S. federal funds rates, U.S. industrial production, U.S. consumer prices, Korean call rates, KOSPI, won/dollar exchange rates, Korean industrial production, and Korean consumer prices. In particular, since the Korean economy is a small open economy in which goods and capital markets completely open, it intensively focuses on whether or not call rate changes move won/dollar exchange rates and consumer prices in the direction which interest rate parity and relative purchasing power parity, the key concepts in international finance field, suggest. According to the empirical results from impulse response analysis based on eight variable VAR models with exogenous foreign variables, positive shocks to Korean call rates increase won/dollar exchange rates, while positive shocks to federal funds rates decrease won/dollar exchange rates, as the interest rate parity assumption shows. Positive shocks to won/dollar exchange rates and U.S. consumer prices also raise domestic consumer prices, as shown in purchasing power parity assumption. In addition, positive shocks to call rates increase domestic consumer prices. According to interest rate parity and purchasing power parity assumptions, it is possible that the rise in domestic interest rates increase won/dollar exchange rates and then leads to the rise in consumer prices. On the other hand, positive shocks to call rates bring down stock price and industrial production. The paper carries out the rolling regression analysis in which the estimation period is recursively moved one hundred sixty nine times with keeping the ten year sample period intact, in order to investigate whether or not the estimation results are distorted and the impulse response is time-varying. The empirical results show that the initial responses of won/dollar exchange rates to call rate shocks change over time. The initial responses of won/dollar exchange rates have negative values during the first half of the 2000s. These periods coincide with the comovement periods in which the value of won to yen is not largely deviated from 0.1. The responses of won/dollar exchange rates to positive federal funds rate shocks decrease after the mid 2000s, as interest rate parity implies. The responses of consumer prices to positive won/dollar exchange rate shocks become weak over time. It means that the shifting effect from exchange rates to consumer prices is reduced. Positive shocks to U.S. consumer prices also raise domestic consumer prices larger in recent years than before. In summary, the dynamic causal relationship between interest rates, exchange rates, and prices moves to the direction coincided with interest rate parity and purchasing power parity assumptions over recent years. Lastly, the paper verifies confidence of the empirical results discussed until now, by restricting the range of impulse response function with sign restrictions based on economic theory. The responses of won/dollar exchange rates to positive call rate shocks generally rise rather than fall. On the other hand, the responses of consumer prices to positive call rate shocks are uncertain, even though the case of rise seems to happen more often than the case of drop. It confirms that the impulse response results discussed above are not wrong. The estimation results of Bayesian VAR models are also similar. These empirical results suggest that the casual relationship between interest rates, exchange rates, and prices come into line with interest rate parity and purchasing power parity assumptions in case of small open economy like Korea. In policy aspect, it proposes that both monetary policy and exchange rate policy must be carried out independently and that it is not ease for the monetary policy making authority to control domestic consumer prices with only policy interest rates. Futhermore, this phenomenon becomes stronger over recent years.