금융자산 가격들의 경기예측력 연구

Throughout the media, it is often said that the steepening of the yield curves has indicated confidence in economic growth and that there has been a relationship between equities and the yield curve. In the early 1990s, usually in the US, forecasts of future economic activity highlighted the usefulness of financial market price variables. Previous researches have also shown that the financial market variables lead the real economic activity by a few quarters and can therefore help predicting future business cycle. Because liberalization of the Korea financial market started in the 1990s, it is believed that not many studies have been conducted to analyze the predictive power of financial market variables. This study aims to enhance the understanding of financial market variables as predictors of Korea’s real economic activity. To this end, we regress the annualized cumulative and marginal percentage change in the monthly Industrial Production Index on the financial market variables in the stock and bond market by adopting the econometric method of Estrella and Hardouvelis (1991). The financial market variables consist of two parts. First is composed of Korea’s term spreads, credit spreads and KOSPI return as domestic variables. Second is US term spread as a foreign variable. We can think of the local business cycles in the countries as having two components; a part affected by movements in the world business cycle, and a country specific component (Campbell, 1991). The main results are summarized as follows. First, on the annualized cumulative percentage change in the monthly Korea’s Industrial Production Index, KOSPI return or credit spreads are likely to provide short-term forecast for less than six months. It turns out that the term spread has long-term predictability over the past year. Second, there was a shift in the predictive power of the term spread when we divided the entire period before and after 2007. Compared with the previous period, it was found that the predicted power of the period spread was relatively better after 2007. Third, even if the call rate is included in the model as a monetary policy variable, the predictive power of the term spread has not been decreased. Finally, we considered the US term spread to see if the foreign financial market variables have further information on changes in the future Korean economy. It was revealed that the forecast power did not exist before 2007, but the prediction power was most widely held about the future economic situation in Korea for about 18 months since 2007. To sum up the results, The first and third results are thought to be similar to the results of previous studies, such as Estrella and Hardouvelis (1991). The second is judged to be that the quality of information distributed by the financial market has improved as Korea’s financial market has become more liberalized. The fourth is believed to be in line with the earlier domestic studies showing that the integration between Korea’s real economy and the United States has grown since 2007.