하도급거래에서 고객업체와 공급업체의 위험분담에 관한 연구

This paper aims to examine risk sharing in the Korean supplier relations. Risk sharing in supplier relations can be conceptually framed contrasting two opposite hypotheses : the risk shifting and the risk absorption hypothesis. Under the risk-shifting hypothesis, customers seek to minimize purchasing costs transferring to suppliers the negative effects of business risk. Customers try to use them as a buffer against business fluctuations. Under the risk absorption hypothesis, customers are concerned not only with short- term reductions of purchasing costs but also with building and maintaining long-term relations with reliable suppliers (Kawasaki and McMillan, 1987; Asanuma and Kikutani, 1992). Within this framework, customers have an interest in absorbing at least part of the risk deriving from unpredictable cost or demand fluctuations. There is a common interest of customers and suppliers in pursuing supply chain global optimization and in designing effective supply contracts, able to allocate profit and risk to each partner in a way that no partner can improve at the expense of the other. Customers can offer to suppliers a sort of ‘insurance' against unexpected variations of production costs, and get the advantages associated with ‘relational quasi-rents’ (Aoki, 1988). The results of previous empirical studies (Kawasaki and McMillan, 1987; Asanuma and Kikutani, 1992; Yun, 1999; Tabeta and Rahman, 1999; Camuffo et al., 2007) seem to support this idea of risk absorption by large customers. However, their empirical results cannot be readily accepted because of serious problems in the econometric method. In the theoretical discussion, Kawasaki and McMillan(1987) assumed the quantity ordered from the customers for each period is constant. From this assumption, risk sharing parameter(α) deduced from Kawasaki and McMillan(1987) agency model can not be used when the output of suppliers fluctuates during estimation period. Besides, as Asanuma and Kikutani(1992) admitted, the effect of demand fluctuation experienced over time cannot be analyzed in the Kawasaki and McMillan(1987) model. Since they do not consider the risk of demand fluctuation at all, this may be at least as essential for parts suppliers as that of cost changes. They fail to provide enough evidence to reject the traditional view of risk shifting. Okamuro(2001) presented an alternative way to examine the risk sharing in the supplier relations. He adopted an econometric method to test if the relative stability of the profit rate of the suppliers, which is regarded as the measure of risk absorption by their customers, differs significantly according to the intensity of business relations. Okamuro(2001) appropriately takes into account of risk from demand fluctuation. It can test risk sharing between customers and suppliers, regardless of its particular risk types. An advantage of Okamuro(2001) model is that it enables us to examine the risk sharing as a whole. Moreover, the problems in estimating the risk absorption parameter(α) can be avoided with this method. But his model has some limits. Although it includes proxy variables for the risk aversion of the suppliers (λ), proxy variable for moral hazard (δ) is omitted. Moreover, his model does not distinguish supplier’s risk sharing from cost fluctuation and demand fluctuation. Therefore, his model does not fit for testing the risk sharing arising from cost fluctuation on which the Kawasaki and McMillan(1987) principal-agent model discussed. This paper examines risk sharing in the supplier relationship, based on an alternative method presented by Okamura(2001), using a unique data set of Korean automotive and electronic parts suppliers. It was tested if the relative stability of the profit rate of the suppliers is significantly influenced by business relationship intensity(MCR), the degree of supplier's risk aversion(λ) and possibility of moral hazard(δ). Firm size(LNLAB), financial solidity(EQTR), large group affiliations(GROUP) were used for the proxy variables for supplier's risk aversion(λ). Labor productivity(VAL) was used for the proxy variable for supplier’s moral hazard(δ). Standard deviation of annual sales increase rate(SDQ) and standard deviation of the profit rate of the main customer(MCSDP) were used as controlling variable to measure the effect of business fluctuation on the stability of supplier's profit rate. The main estimation results are as follows;1. The relative stability of the profit rate of the suppliers is significantly influenced by the intensity of business relations with the main customer. In regard to the variable of the business intensity(MCR), estimated coefficients have always expected negative significant signs. It means that customer absorbs the larger part of the risk of its suppliers, the higher their dependence on it in sales is, which is consistent with Okamura(2001). 2. As for the proxies for risk aversion, some coefficients have expected signs, but other has not. Those of EQTR and GROUP have expected positive signs. It shows that customer absorb supplier's risk, partly, according to the degree of risk aversion. But those of firm size(LNLAB) have wrong signs, like that of Okamura(2001). This shows that large suppliers can stabilize their profit rate not by risk sharing, but by themselves. 3. Variable for moral hazard(VAL) has significant effects on the stability of the profit rate. It shows that the more capable the supplier is to reduce production cost, the less is the customer’s incentive to absorb supplier's risk. These findings support the risk absorption hypothesis that the risk sharing by the main customer depends on the intensity of the business relationship, the degree of supplier's risk aversion and possibility of moral hazard. 4. The coefficient of MCSDP(standard deviation of the profit rate of the main customer) shows somewhat different risk sharing between automobile and electronic industry. It has insignificant values in the automobile industry, significant positive signs in the electronic industry. This shows that instability of the profit rate of the suppliers is positively correlated with that of main customer in the electronic industry. The profit stability of the suppliers is significantly influenced both by the profit level and instability of the main customer in the electronic firms. This implies that customer's risk absorption does not exclude risk shifting in electronic firms. These results suggest as a whole that the Korean main customers absorb a part of the business risk of their suppliers depending on the intensity of the relationship, the possibility of suppliers’ moral hazard, and, partly, the degree of suppliers’ risk aversion. However, these factors explain only a small part of the profit stability of the suppliers, as the values of adjusted R2 indicate. Moreover, these provide the evidence from electronic firms that customer shift a part of their own risk to suppliers.