in one basket. Firms operating in “dirty” industries must continuously assess
the risks posed by changing environmental regulations. In the wake of the mon-
umental losses associated with Hurricane Andrew, disaster insurers have been
taking a microscope to their risk portfolios. Using computer models, they sift
through decades of data on storm patterns and earthquakes to estimate risk
probabilities. While looking out for excessive geographic concentration of
insurance coverage, the insurers are also reassessing shoreline properties, scru-
tinizing building codes, raising premiums, dropping policies, reinsuring por-
tions of their risks, and even offloading risks by selling “catastrophic” (CAT)
bonds to investors. (With CAT bonds, an investor obtains a high-interest return
in “normal” circumstances but loses a portion, or all, of the principal if yearly
hurricane damage claims exceed specified thresholds.)
An important insight offered by risk management is that many risks are
interdependent. Decisions made in one area create (or mitigate) risks in
another. Alerted to the risks of mass tort litigation for repetitive stress injury,
Microsoft incorporated this cost ($2.82 per unit) when setting the licensing
fee for its new innovative keyboards, thereby providing a prudent monetary
reserve for this risk.Expected Utility
How can a manager formulate a criterion, reflecting the firm’s attitude toward
risk, to guide his or her decisions? The formal answer to this question was devel-
oped more than 50 years ago by mathematical economists John Von Neumann
and Oscar Morgenstern, and is called the expected-utility rule. (At the same
time, Von Neumann and Morgenstern developed the field of game theory,
which we encountered in Chapter 10.)
The use of expected utility proceeds in two steps. First, the decision maker
must think carefully about the firm’s preferences concerning risks: what risks
it is willing to accept and how to value those risks. In the process, the manager
constructs a utility scale that describes this risk tolerance. Second, the manager
analyzes the decision problem in much the same way as before, that is, he con-
structs a decision tree showing relevant probabilities and possible monetary
outcomes and then evaluates the tree. However, there is one crucial difference:
In contrast to the risk-neutral manager, who averages monetary valuesat each
step, the risk-averse decision maker averages the utilitiesassociated with mone-
tary values. At each point of decision, the manager selects the alternative that
supplies the maximum expected utility. With this summary in hand, let’s see
exactly how the method works.A RISK-AVERSE WILDCATTER Once again, let’s consider the wildcatter’s basic
decision problem, reproduced in Figure 12.7. Now suppose the wildcatter is
risk averse; he is unwilling to rely on expected profits as his choice criterion.522 Chapter 12 Decision Making under Uncertaintyc12DecisionMakingunderUncertainty.qxd 9/29/11 1:34 PM Page 522