文摘
This dissertation explores several theoretical and empirical issues related to the existence of a soft budget constraint (SBC). We first model the effect on the likelihood of SBC of the possibility for lenders to acquire reputation in a setup where ex ante unprofitable projects would be refinanced by lenders after the initial losses become sunk costs. We then empirically show, using a firm-level data set, that severe SBC had existed prior to the 1997 financial crisis in Korea. While SBC has often been conjectured as one of the primary causes of the Korean crisis, no formal testing has been presented in the literature. In addition, we find that the performance of firms that are found to operate under SBC was worse than that of other firms.;In the theoretical essay, a lender can either play soft by refinancing poor projects or play tough in order to build a long-term reputation by liquidating poor projects. The results show that reputation can alleviate SBC, by discouraging entrepreneurs with poor projects from applying for funds in the future. Moreover, in a model with lenders' reputation, SBC is less likely to exist under decentralization than under centralization.;The empirical essay examines whether SBC, known to exist in socialist economies, also existed in Korea prior to 1997. We test for the existence of SBC by examining whether firms that were either reasonably expected to default or actually defaulted in year t, had been able to increase their borrowing in year t-1, compared to other firms. The results show strong evidence of SBC in Korea prior to the crisis in most types of loans tested: bank loans, crony loans, foreign currency loans, corporate bonds and trade-related debts. Also, SBC appears to have existed with respect to short-term loans for all firms, but with respect to long-term loans only for chaebols. In addition, we found that loans to SBC firms had significant adverse effects on their profitability, thus implying that SBC resulted in inefficiencies in the Korean economy.