Identifying business cycle turning points in Korea
with a new index of aggregate economic activity(Vol.7 No.1)

Joong Shik Lee
: Senior Economist, Monetary Policy Analysis Team, Monetary Policy Department, the Bank of Korea (E‐mail: jslee@bok.or.kr). I would like to thank my colleagues at the Monetary Policy Analysis Team for their constructive suggestions. I have benefited from the helpful comments and encouragements of Jong‐Kun Lee and Hangyong Lee on the earlier version of this paper, and would like to thank two anonymous referees for their comments.

 This paper introduces a new index of aggregate economic activity in the Korean economy. The index, based on principal component analysis in combining a large data set into a single summary measure, was pioneered by the work of Stock and Watson (1999). The advantage of this index is that it can include almost all the variables needed in evaluating macroeconomic conditions, while the weight of each individual variable is determined ex post by the degree of conformity with the common latent movements of the variables. The empirical application of the new index indicates that it well reflects the official business cycle chronology of Korea set by the National Statistical Office.

 At the same time, the threshold values of the new index are explored through a Monte Carlo simulation to address: (i) at what values would the new index identify the business cycle turning points correctly in a certain degree, and (ii) what would be the time lag for each value. The simulation results imply that setting the threshold values for the new index to –0.50 for the peaks and +0.20 for the troughs leads to a 76 to 87 percent success rate in identifying the business cycle turning points within a time lag of three to four months.

JEL Classification Number: E32, E37

Key words: new index of aggregate economic activity; reference date; principal component analysis; threshold values; Monte Carlo simulation