요소가산법을 활용한 비상장 중소기업 자기자본비용 측정에 관한 실증연구

Cost of capital is a vital decision-making factor in almost every aspect of corporate finance and accounting fields such as firm valuation and appraisal of new investment projects. Yet, most of the studies on cost of capital examine listed companies in theoretical perspectives, and analyses on unlisted SMEs, which comprise the majority of the domestic firms, are virtually non-existent. Moreover, many of the previous researches are focused on analyzing the relevance between the cost of capital and stock price or other information, rather than specific methodologies that calculate the cost of capital of each corporation. Such analyses may yield academic implications but cannot assist the investors or staff at financial institutions that conduct valuation. This study has examined the usefulness of cost of capital for unlisted SMEs used by financial industries by analyzing its relationship with default risks. The hypothesis testing expected that the cost of capital of insolvent enterprises will be higher than that of sound businesses. The costs of capital tested include the cost of debt(Kd Model), cost of equity capital using the build-up method (Ke Model), and the weighted average cost of capital of the two previous models (wacc Model), which are used for unlisted SME valuation at KODIT(Korea Credit Guarantee Fund). For a comparison test against Ke Model, Hamada model, which is widely used in financial industries to calculate the cost of capital of unlisted firms, and the Technology_Evaluation Model, which is adopted as cost of capital by KOTEC(Korea Technology Credit Guarantee Fund) for technology valuation , were further analyzed. Data is extracted from the database of unlisted SMEs maintained for and used in rendering guarantees by KODIT. Quasi-population of this study is 290,853 observations for the years 2003 through 2011 to sufficiently reflect the effects of default rates caused by macro-environmental changes of each year. The result showed strong statistical correlations between cost of capital and default probability for all three testing models. In particular, the wacc model, while exhibiting a similar discriminant power as the Ke Model, demonstrated more accurate predictions in terms of the stability of the model including controlled variables, and is recommended for actual practice. On the other hand, Hamada Model and Technology_Evaluation Model fail to have any additional explanatory power even when included in the controlled model. The outcome of this study has implications as follows. First, it proposes a cost of capital calculation methodology applicable for unlisted SMEs, which have been neglected in previous studies. Next, it suggests that the cost of capital of unlisted SMEs is significant only when inherent credit risk premiums reflecting the idiosyncratic risk factor as well as systematic risk are considered. Finally, the methodology that uses beta of listed corporations of each industry as a proxy for the cost of capital of unlisted SMEs does not accurately reflect the idiosyncratic risk and therefore is not suitable for cost of capital calculation.