기업재무 데이터특성을 고려한 자본구조조정속도 추정에 관한 연구

초록

The speed of adjustment toward target leverage (SOA) is the key estimate to measure how well the trade-off theory explains the observed capital structure in data. It is well known that an unbalanced panel structure, censored dependent variables and a limited number of cross-sectional observations in corporate panel data potentially cause substantial bias in the SOA estimates. Recent simulation evidence suggests the adoption of Bruno (2005) and Elsas and Florysiak (2014) models to resolve such inconsistency problems in corporate panel data. We newly estimate the SOA for Korean corporations by using the above methods and compare our new estimates with the ones from existing models. Our main results are as follows. First, both of the methods estimate the SOA for the target market leverage as 2.5 years, which supports the explanatory power of the trade-off theory in leverage dynamics. On the other hand, the models of Bruno (2005) and Elsas and Florysiak (2014) obtain the SOAs as 4.3 and 4.9 years for the target book leverage, respectively, arguing against the explanatory power of the trade-off theory. Our contrasting results for the models using market and book leverage is not observed in the estimates from prior models using the ordinary least square method, Fama and MacBeth (1973) and Arellano and Bond (1991). This finding provides substantial implications on the explanatory power of the trade-off theory in Korean corporations.

키워드