a. Calculate the centroids for each group in the analysis sample. What do these values indicate?

b. Use the regression approach to two-group DA to develop a rule for predicting whether or not a bank will enter financial distress. State the classification rule you identify.

c. Use the classification rule that you identify in question 2 to classify each observation in the analysis sample. Create a confusion matrix summarizing the classification results. Overall, how accurate does this classification rule seem to be?

d. Suppose that the FDIC also gave you the data below on six new banks. Based on the classification rule that you developed using regression, which (if any) of the six new banks would you expect to enter financial distress?

e. The FDIC also wants you to analyze the data using the Mahalanobis distance measure approach discussed in this chapter. Carry out this analysis. Which of the two methods would you recommend that the FDIC use?

The number of bankruptcies in the banking industry recently reached a historic high, unparalleled since the Great Depression. In 1984, fewer than 50 bankruptcy cases were filed with the Federal Deposit Insurance Corporation (FDIC), whereas in 1991, more than 400 were filed. To monitor the member banks and to ensure their proper compliance with federal regulations, the FDIC commits substantial efforts to both on-site examinations and off-site surveillance activities. To facilitate this process, the FDIC is developing an early warning system that will alert the agency to banks that might be approaching or entering financial distress. Using the technique of DA, you will help the FDIC develop such an early warning system.
Specifically, the FDIC wants to develop a procedure to help predict if a bank will enter financial distress in the coming year. The following financial ratios are helpful in making such predictions:

The FDIC supplied you with data on these ratios collected one year ago for a sample of banks. It also indicated which banks are currently in financial distress. These data are summarized in the file BankData.xls on your data disk (where Group 1 = financially distressed banks, Group 2 = financially strongbanks).

Transcribed Image Text:

Obs Group R Ratio 2 Ratio 3 3.3 9.2 9.3 3.2 11.4 8.1 0.13 0.08 0.09 0.16 0.09 0.14 0.63 0.26 0.59 0.57 0.69 0.72 Tot Total Captial Ratio 1 = Total Expenses Total Assets Total Loans and Leases Ratio 2 Ratio 3 Tota!Deposits