Basu의 비대칭적 적시성 모형의 추정편의: 한국기업표본에서의 조건적 보수주의 측정

Several researchers point out that the estimation of Basu’s(1997) asymmetric timeliness model suffers from the bias. However, there are not many studies that examine the biases in the estimation of the Basu(1997) model in the Korean environment. Considering the accounting and finance literature that suggests that the informativeness of accounting earnings and stock returns, which are the primary variables of the Basu(1997) model, is influenced by the characteristics of the country, re-examining the problems of the Basu(1997) model with a sample of Korean firms is a necessary process to justify empirical studies on conservatism in Korea. This study finds that a bias exists in the Basu(1997) model when estimated with a sample of Korean firms. However, this bias is resolved for firms listed in the Korean Stock Exchange (KSE) when the correlated omitted variables problem is controlled. On the other hand, such controls do not effectively mitigate the bias for the sample of firms listed in the Korean Securities Dealers Automated Quotation(KOSDAQ) market. This result is determined to be from the difference in the informativeness of stock returns between the KSE market and the KOSDAQ market. As an additional analysis, the influence of the bias of the Basu(1997) model on the Khan and Watts(2009) conservatism measure is examined. In the sample of firms listed on KSE, most of the bias in the Basu(1997) model was effectively controlled in the estimation of the Khan and Watts(2009) measure. However, in the KOSDAQ listed firms sample, the Khan and Watts(2009) measure was significantly influenced by the bias.