Hasil (
Bahasa Indonesia) 1:
[Salinan]Disalin!
All of the variables are scaled by the book value of assets(At, data 89) to account for heteroskedasticity. The dependent variable is the market value of equity (Vt)scaled by book value of assets. The independent variables include earnings (Et), interest expense (It),research and development (RDt), and dividends (Dt), all scaled by At.Et equals net income (data 32) plus interest expense (data 15) plus deferred taxes (data 25);RDt equals research and development expense (data 52);It equalsinterest expense (data 15); and Dt equals the total amount of dividends (data 34).dXtis the difference between Xt and Xt 2 for variable X, and dX t+2 is the difference between X t+2 and Xt. We also include a civil law dummy in the regression sinceLa Porta, Lopez-de-Silanes, Shleifer, and Vishny (2002) show that valuations are lower in civil law counties. The main variable of interest is the interaction term between creditor rights and the change in dividends(CR(dxt+2/A)).The coefficient of this variable,a,is predicted to be positive according to the (competing) tradeoff hypothesis.To estimate regression Model 3, we first delete any firm-years that do not have data in the required fields.This process reduces our sample size from 120,507 firm year observations to 67,331 firm-year observations. We check for biases in the reduced sample by comparing its distributions across years, industries, and countries to thefull sample distributions in Panels A, B, and C of Table 9.Overall, the reduced sample appears to be fairly representative of the full sample across years, industries, andcountries.In Table 10,wereport Fama–MacBeth (1973)estimations for variations on regression Model 3. Due to space limitations, we report only the coefficient estimates for the main variable of interest, the interaction term (a16). All other coefficient estimates are consistent with prior studies
Sedang diterjemahkan, harap tunggu..