Engineering Economic Analysis

330 UNCERTAINTY IN FUTURE EVENTS

Summary

Estimating the future is required for economic analysis, and there are several ways to do

this. Precise estimates will not ordinarily be exactly correct, but they are considered to be

the best single values to represent what we think will happen.

A simple way to represent uncertainty is through a range of estimates for each variable,

such as optimistic, most likely,and pessimistic. The full range of prospectiveresults.maybe

examinedby using the optimistic values to solve the problem and then using the pessimistic

values. Solving the problem with the most likely values is a good single value estimate.

However,the extremes with all optimistic values or all pessimistic values are less likely-it

is more likely that a mix of optimistic, most likely, and pessimistic values will occur.

One approach uses weighted values instead of a range of estimates. One set of weights

suggested is:

Estimate

Optimistic

Most likely

Pessimistic

Relative Weight

1

4

1

The most commonly used approach for decision making relies onexpected values.

Here, known or estimated probabilities for future events are used as weights for the corre-

sponding outcomes.

Expected value=OutcomeAx ProbabilityA+ OutcomeBx ProbabilityB+ ...

Expected value is the most useful and the most frequently used technique for estimating a

project's attractiveness..

However, risk as measured by standard deviation and the probability of a loss is alsQ

important in evaluating projects. Since projects with higher expected returns also frequently

have higher risk, evaluating the trade-off's between risk and return is useful in decision

making.

More complicated problems can be summarized and analyzed by using decision trees,.

which allow logical evaluation of problems with sequential chance, decision, and outcome

nodes.

Where the elements of an economic analysis are stated in terms of probability dis-

tributions, a repetitive analysis of a random sample is often done. This simulation-based

approach relies on the premise that a random sampling of increasing size becomes a better

and better estimate of the possible outcomes. The large number of computations means that

simulation is usually computerized.

PROBLEMS

10-1 Telephone poles exemplify items that have varying
useful lives. Telephone poles, once installed in a
location, remain in useful service until one of a variety
of events occur.
(a)Name three reasons why a telephone pole might

be removed from useful service at a particular

location.

(b)You are to estimate the total useful life of tele-

phone poles. If the pole is removed from an orig-

inal location while it is still serviceable, it will