9781118041581

Sequential Decisions 511

Africa and the Philippines. Instead, it has been humbled by well-established local

competitors catering to local tastes. In 1977, Coca-Cola was the leading soft drink

in India before the company pulled up stakes refusing to divulge its secret formula

to the Indian government. Returning in 1993, the company found itself a distant

second to Thums Up, an imitation cola in a similar glass bottle, that won the alle-

giance of Indian consumers in Coke’s absence. Admitting that its tried and true

business formula was the wrong one for India, Coca-Cola’s management purchased

Thums Up and now aggressively markets that drink alongside “The Real Thing.”

The lesson to take from these companies’ experiences is that international

businesses, if they are to be successful, must be especially vigilant in identifying

myriad risks and capturing them in carefully conceived, bushy decision trees.

SEQUENTIAL DECISIONS

Some of the most interesting and important business and economic problems

call for a sequence of decisions to be made over time. For example, suppose a

chemical firm is considering a large capital investment in a new petrochemical

facility. The profitability of such an investment depends on numerous uncer-

tain factors: future market demand, reactions of close competitors, and so on.

Profits also depend on the future product and pricing decisions of the firm

itself. It is not simply that the firm faces many decisions over time; the more

important point is that the sequence of decisions is interdependent. A correct

investment decision today presupposes that the company will make optimal

(i.e., profit-maximizing) pricing decisions tomorrow if the plant is built. The

following example illustrates this general point about sequential decisions.

An R&D

Decision

Revisited

In Chapter 1, we sketched a decision problem facing a pharmaceutical firm

that must choose between two research and development approaches. Suppose

the profits and probabilities of the competing methods are summarized in the

following table:

Profit

R&D Choice Investment Outcomes (Excluding R&D) Probability

Biochemical $10 million Large success $90 million .7

Small success 50 million .3

Biogenetic $20 million Success $200 million .2

Failure 0 million .8

All profit figures are expressed in terms of present discounted values and thus

are directly comparable to investment figures.

c12DecisionMakingunderUncertainty.qxd 9/29/11 1:34 PM Page 511