This study provides data, overviews, and analyses useful toward developing a composite business cycle index that is suited to the Korean economic structure.
The ongoing industrialization and globalization of the Korean economy is generating interest in the evolution of domestic business cycles. As part of the first efforts to predict and manage fluctuations in the market environment at home and abroad, the Korea Development Institute (KDI) and the Economic Planning Board (EPB) set out to develop a composite business cycle index that is particularly suited to Korean economic structure, using the latest research results and methodologies. This study provides a detailed description of the process in which the index was developed, and surveys the established literature on business cycles and indicators in general, with an objective of helping Korean policymakers develop more effective policies for managing business cycles.
In his “Business Cycles and Composite Business Cycle Index” (1981), Seo Sang-mok introduces the basic concepts and principles of business cycles, and provides an overview of the recent trends in empirical research on the topic. Analyzing various indices, Seo emphasizes the importance of developing a composite business cycle index suited to its economy. He then briefly describes how his team of policy researchers and decision-makers came to develop a new index for Korea using the National Bureau of Economic Research (NBER) method, and applies the new index to his findings as part of his analysis.
Our study involved developing a composite business cycle index using the latest method and applying it to a systematic analysis of the business cycles that the Korean economy experienced in the past, as well as towards the identification of the current business cycle in Korea. This study also analyzes and predicts successive business cycle fluctuations in Korea using the leading composite index, providing basic data upon which a new business cycle management policy can be developed.
Given the dependency of the Korean economy on global trade, an international comparison of the composite business cycle indices of advanced economies worldwide—developed using consistent methodologies—can provide indispensable information and insight into economic analysis. In order to ensure the improvement of the composite business cycle index, Korean researchers need first to gather statistics that are sensitive to business cycle fluctuations. They also need to develop ways to normalize seasonal fluctuations, as well as new techniques for analyzing minor fluctuations.
Yun Hong-nyeol’s “Adjusting Factors of Seasonal Fluctuations” (1981) provides a detailed explanation of the methods used to adjust and control for the factors of seasonal fluctuations in the business cycle, the first step toward developing business cycle indices. He compares several methods for adjusting for seasonal fluctuations, and describes the characteristics of, and formula used, in the X-I I autoregressive integrated moving average (ARIMA) method used in his study.
Although X-I I ARIMA provides a number of advantages in terms of controlling for seasonal factors in time series analyses, Korean researchers have so far failed to make use of this method. They rely almost exclusively on the auto select function of the ARIMA program in extending time series. When we observe time series data using the auto select mode, we may come across certain cases with special ARIMA models. The majority of them, however, are over relatively short spans of time and show extreme sensitivity to non-economic and arbitrary factors. We can overcome this problem first by switching the amorphous model of ARIMA (p,d,q)(P,D,Q) with a formal model, enlarging the “d” and “D” so as to absorb and eliminate arbitrary factors. This, however, may transform the model into something entirely different when there are no future extreme arbitrary factors. Alternatively, we may revise the original time series, revisiting the prior factors so as to discern a more probably ARIMA model. However, we have yet to find an effective method for identifying these prior factors. Therefore, our next task should be developing a method that can determine the ARIMA model suited to the characteristics of a specific time series and also identify the prior factors that can correct for arbitrary factors.
In “An NBER Program for Business Cycle Analysis” (1981), Yun Hong-yeol and Lee Jae-yong discuss the process of creating a composite index by adjusting for the factors of seasonal fluctuations and identifying transition points in a business cycle.
A composite business cycle index attempts to estimate and predict how the business cycle of an economy will change. The index indicators are used to determine normal changes in the economic statistics time series that represents each economic sector. The two main criteria for determining the time series for each indicator are economic consistency and total evaluation score. In order to bring as many economic sectors as possible under the analysis of the index, we must choose series that represent a different sectors and have a higher evaluation scores from time series showing identical levels of economic consistency. This prevents excessive overlap among the time series involved.
It is thanks to the comprehensiveness and diversity of its scope that a composite index is more reliable than other indices as an indicator of business cycle fluctuations, as well as having a lower error rate and less tendency towards arbitrary fluctuations than individual time series. In addition, indicators that can be used monthly and that are not likely to be revised entirely are appropriate to our purpose, which is to find a composite index that is timely and accurate. In trying to bring all the different individual time series or indicators into a single index, it is also important to prevent sensitive indicators from dominating the resulting index, and ensure that time series with higher evaluation scores exert greater influence on the final index. This is done by standardizing the absolute margin of average change in all the time series. The total evaluation scores are then used as weights. Therefore, this study summarizes each phase of developing the composite business cycle index with mathematical equations.
Lee Kang-cheol and Byeon Ho-seob attempted to utilize the evaluation method regarding the time series for economic indices in their “How to Evaluate Time Series of Economic Statistics” (1981). Different researchers prefer different methods for analyzing time series of economic statistics. However, as the selection of time series and their weights can yield significantly different results in the composite business cycle index, we need to find a method that can more objectively evaluate time series. Lee and Byeon apply the NBER method, but not without modifying it to suit Korean circumstances. Yet, there are a number of difficulties involved in rendering actual evaluations through this method. First, it is impossible to exclude the analyst’s subjective bias from time series analysis, while it is also necessary for the analyst to understand as many types of economic statistics as possible. Second, the quantity statistics regarding business cycles that was available to the authors was inadequate, particularly regarding employment and investment. Third, economic statistics in Korea tend to span over relatively short spans of time, making it difficult to analyze economic consistency and responsiveness, evaluations which require distribution analysis. There were also numerous indicators with tendencies towards extreme or arbitrary fluctuations. The authors analyzed these indicators by applying the MCD moving average method. An accurate and reliable score evaluation system will be crucial to the accurate analysis of fluctuations in time series of economic statistics in Korea, and also to enhance the value of the composite business cycle index.
In their “Business Cycle Index” (1981), Kwon O-sul, Lee Jae-yong, Byeon Hyo-seob and Lee Kang-cheol provide an overview of the statistics on business cycles that the authors have used. The study examines the effect of the business cycle on production and income, distribution and consumption, trade, capital investment, inventory investment, consumer prices, currency and finance, employment and unemployment, and wage and cost in Korea, delineating the adjusted time series of seasonal and trend fluctuations in each indicator.
- 경기종합지수 작성에 관한 연구보고서(A study on the development of a composite business cycle index in Korea)
- 서상목, 편
경기종합지수 작성에 관한 연구보고서(A study on the development of a composite business cycle index in Korea)
|Series Title; No||연구보고서 / 제81-12권|
|Subject||Economy < General|