Estimation of Current Account Benchmarks via Bayesian Model Averaging (in Korean)

The aim of this paper lies in estimating the current account benchmarks for Korea to be employed for assessing imbalances and sustainability of the current account balance. As the model uncertainty involved in the estimation is considerable, Bayesian model averaging(BMA) is chosen as a estimation method. The estimated current account benchmarks have shown a stable path while varying within a narrow interval between 2.2% and 2.6%. The current account imbalances, assessed by the estimation results, tend to expand on the arrival of forceful negative shocks from abroad and then contract with the global and domestic economies improving in tandem. This analysis implies that it is desirable to induce the imbalances to dwindle naturally through implementing policies to contribute to a stable growth of macroeconomy in the medium-term horizon rather than the ones targeting only curtailment of current account surplus in the short-term. Comparing the current account benchmark estimated via BMA and the one via classical inference, the former proves to be more consistent with the concept of benchmark and existing estimation results in the literature in that it shows a more stable path.