Essays on business cycles in Korea

Abstract

Over the last two decades, a large amount of macroeconomic research has been directed towards the business cycle theory. Until now, however, most literature has focused on the business cycle research in the developed economies like the United States and European countries. On the other hand, the studies on emerging markets are very limited. From this perspective, this thesis attempts to contribute to the current literature by investigating the main characteristics of Korean business cycles, comparing its findings with the previous findings in the developed countries, and then drawing out the economic or policy implications. Another motivation of this thesis arises from the experience of the Korean financial crisis in the later half of 1997, which had a catastrophic impact on the Korean economy. Although it is not fully agreed on what caused the Korean financial crisis, there are a large amount of literature on this subject. However, what have not been dealt in literature is whether the Korean financial crisis brought out any noticeable changes on the fluctuations of key macroeconomic variables such as output, consumption, worked hours and asset wealth in Korea. One possible excuse for this scarcity is the limitation of data span, namely too short to analyze the difference between the pre- and post-crisis periods. It appears that as a decade passed, this is the right time to investigate this issue. Therefore, across three essays, this issue will be discussed with a high priority. With these two purposes in mind, this thesis starts by investigating the relationship among consumption, financial wealth and labor income with Korean data. This issue is discussed in the framework of the vector error correction model by applying the full information maximum likelihood approach suggested by Johansen (1988, 1995). The main finding from this analysis is that only financial wealth shows the sizable and statistically significant error correction behavior. This finding is also confirmed by the permanent and transitory component decomposition, which shows that only fluctuations of financial wealth are largely associated with transitory components while consumption and labor income are mainly governed by the permanent shocks during the examined period. By comparing the pre- and post-crisis periods, we also find that although there were several policy and institutional changes during the crisis, most adjustment to the long run relation has been done by financial wealth across the two sample periods. The second and last essays explore Korean business cycles by estimating the micro-founded dynamic stochastic general equilibrium models, using maximum likelihood and Bayesian approaches, respectively. Based on the estimation, the second essay finds that the estimated DSGE model out performs VAR models in predicting hours worked but it has difficulty in predicting other key macrovariables. In addition, although the volatility of the economy has decreased in the recent, the Korean financial crisis seemed not to change the deep parameters in the model. Finally; by comparing the second moments from the HP filtered data with those from the simulated data, this essay finds that the estimated model successfully reproduces the relative volatility of consumption and hours worked as well as the pattern of contemporaneous correlations of output with consumption, investment and hours worked. The last essay extends the baseline model in the second essay by introducing money through cash-in-advance constraints, that is, firms should borrow cash to pay wages in advance while households have to hold cash to purchase consumption goods. After estimating two versions of cash-in-advance models, namely a baseline cash-in-advance model and a limited participation model, this essay shows that the limited participation model is better to match up the stylized facts in Korean business cycles. In particular, it successfully captures the rise of output in response to an expansionary money shock. Finally, the comparison between the pre- and post crisis periods shows a sharp decline of money shock and a slight decline of productivity shock.