9781118041581

554 Chapter 13 The Value of Information

TABLE 13.2

Assessing Loan Risks

Part (a) lists the fre-

quency of loan cate-

gories by actual

default experience.

Part (b) lists the joint

probabilities of all

outcomes.

Part (c) shows the

conditional

probabilities.

(a) Frequencies of Loan Categories by Actual Default Record

Category Performing Loan Defaulted Loan

A (“zero” risk) .2 .1

B (solid) .4 .2

C (uncertain) .3 .4

D (high risk) .1 .3

Total 1.0 1.0

For example, 10 percent of all defaulted loans were (incorrectly) judged

to be “zero” risk at the time the money was lent.

(b) Joint Probabilities

Category Performing Loan Defaulted Loan Total

A (“zero” risk) .18 .01 .19

B (solid) .36 .02 .38

C (uncertain) .27 .04 .31

D (high risk) .09 .03 .12

Total .90 .10 1.00

(c) Conditional Probabilities

.01/.19.05

.02/.38.05

.04/.31.13

Pr 1 defaultƒD 2 .03/.12.25

Pr 1 defaultƒC 2

Pr 1 defaultƒB 2

Pr 1 defaultƒA 2

Business Behavior and Decision Pitfalls

By now you should be familiar with and practiced in the simple mechanics of

computing probabilities based on new information. Of course, the typical man-

ager (and, to be sure, the average person) does not have Bayes’ theorem on the

tip of his or her tongue; rather, the manager probably uses informal prediction

methods based on personal judgment, experience, and intuition. However,

there are two main problems with informal approaches.

The first difficulty is that the logic underlying the prediction often is

uncheckable, or at least hard to pin down. What factors led the individual to

make that prediction? How would this forecast change under different cir-

cumstances or assumptions? Some sort of logical analysis is necessary to answer

these questions. Even forecasters with track records of accurate predictions

c13TheValueofInformation.qxd 9/26/11 11:02 AM Page 554