197 /302
197
臺大管理論叢
第
28
卷第
2
期
Concerning the firm-supplier relationship uncertainty (hereafter denoted as FSRU),
Kumar (1996) and Carr and Pearson (1999) have demonstrated that the firm-supplier
relationship is positively associated with inventory flow performance. A higher FSRU may
engender higher inventory flow fluctuations and consequently higher cash flow variations
(Tsai, 2008). However, replacing existing or part of existing suppliers to lower production
costs or raise bargaining power increases the FSRU. This is consistent with the
phenomenon of the
second supplier
in practice. In this instance, higher FSRU reduces the
degree of dependence on existing suppliers for firms, provides other alternative suppliers,
and further improves firm asset value (Porter, 1979, 1985). On the basis of the preceding
descriptions, the effects of FSRU on firm asset value distributions are twofold: (1) The
FSRU increases a firm’s cash flow uncertainty and its asset value variation (Kumar, 1996;
Carr and Pearson, 1999; Tsai, 2008), (2) the FSRU improves a firm’s bargaining power
and enhances its asset value (Porter, 1979, 1985). According to the structural credit models
by Merton (1974), the aforementioned two effects are opposite; therefore, the FSRU has
indefinite effects on firm credit risk. We thus propose the second hypothesis as follows:
Hypothesis 2: A firm’s FSRU has indefinite impacts on corporate credit risk.
2. Design/Methodology/Approach
To estimate a firm’s FCRU and FSRU, we first follow the procedures described by
Fee and Thomas (2004), Kale and Shahrur (2007), and Chen, Liao, and Kuo (2013) to
identify a firm’s customers and suppliers by employing “COMPUSTAT Industry Segment
Files.” After completing the identification of the firm’s customers and suppliers, we use
the percentage of the firm’s sales to the
j
th customer (
Customer Percentage Sold
j
; CPS
j
)
and that of a firm’s purchases from the
j
th supplier to the firm’s total sales (
Customer
Input Coefficient
j
; CIC
j
) as a measure of the firm-customer relationship and firm-supplier
relationship, respectively. We also employ CPS volatility (hereafter denoted as CPSV) and
CIC volatility (hereafter denoted as KCICV) as the proxies of FCRU and FSRU,
respectively, for expressing the variations in the supply chain relationship. A firm's CPSV
is calculated as the standard deviations of CPS for the previous 5–8 years (CPSV5,
CPSV6, CPSV7, and CPSV8). Similarly, a firm’s KCICV is calculated as the standard
deviations of CIC for the previous 5–8 years (KCICV5, KCICV6, KCICV7, and
KCICV8). The higher the CPSV and KCICV values are, the higher is the supply chain
relationship uncertainty.