VaR Type Risk Calculations

Suppliers may face several different disruptive events. In our case, hurricanes are assumed to cause the most significant supply disruptions hence, buyers initially want to focus on VaR type risk brought by hurricanes. We present next the steps to calculate VaR type risk values for supplier S19. 

Step 1 Collect historical data about supplier S19 (see Table 7.22). Note that the method does not require any increasing or decreasing pattern in data. 

Step 2 Use Table 7.22 data to fit an appropriate GEVD that is, find the parameter values to be used in the GEVD function. The values are calculated as K = 0.15621, 5 = 197,835, and X = 250,632. (Refer to Section 7.9.5 for estimating the parameters of GEVD.) 

VaR Type Risk Values for Suppliers with 80% Confidence 

Step 3 Estimate the mean parameter of the Poisson distribution to model the occurrence. In this example, we assumed, on the average, only one hurricane impacted the operations at S19 every year. 

---

Values in Table 7.23 can be interpreted similar to the MtT type risk values. For example yearly losses from hurricanes at S19 s region would not exceed 694,308.35 with 80% confidence. The use of small sample size in VaR type risk calculations is justified by the infrequency of these events. PWM estimators and the approximate estimators used in this section were shown to yield bias smaller than other estimators (Fiolsking et al., 1985). 

---