CHAPTER 10 Risk and Refinements in Capital Budgeting 429ble values for a given variable, such as cash inflows, to assess that variable’s
impact on the firm’s return, measured here by NPV. This technique is often use-
ful in getting a feel for the variability of return in response to changes in a key
variable. In capital budgeting, one of the most common sensitivity approaches
is to estimate the NPVs associated with pessimistic (worst), most likely
(expected), and optimistic (best) estimates of cash inflow. Therangecan be
determined by subtracting the pessimistic-outcome NPV from the optimistic-
outcome NPV.EXAMPLE Continuing with Treadwell Tire Company, assume that the financial manager
made pessimistic, most likely, and optimistic estimates of the cash inflows for
each project. The cash inflow estimates and resulting NPVs in each case are sum-
marized in Table 10.2. Comparing the ranges of cash inflows ($1,000 for project
A and $4,000 for B) and, more important, the ranges of NPVs ($7,606 for proj-
ect A and $30,424 for B) makes it clear that project A is less risky than project B.
Given that both projects have the same most likely NPV of $5,212, the assumed
risk-averse decision maker will take project A because it has less risk and no pos-
sibility of loss.In Practice
Ever since the economy faltered
in late 2001, information technol-
ogy (IT) managers have faced
increased pressure to measure
returns on technology invest-
ments and to show higher ROIs
and faster project implementation.
Managers must justify projects,
proving that they support strate-
gic business goals, and then track
progress against expectations.
Another key trend: Companies are
moving IT approvals to more
senior levels of management in
order to evaluate better the proj-
ects’ overall impact on the com-
pany’s business.
In a poll ofComputerworld’s
“Premier 100” IT companies,
almost half of the respondents
said they do not perform ROI
analysis on proposed IT projects.
For the 43 percent who calculate
potential paybacks, nonfinancial,
“soft” factors are an important
part of the analysis. The chief
information officer (CIO) may con-
sider certain projects—for exam-
ple, business-to-business (B2B)
commerce—essential to the com-
pany’s future.
Methods and metrics to
assess ROI vary among companies.
Illinois communications equipment
makerTellabs Inc. established a
stringent proposal-and-approval
process for IT projects. This formal
analysis now includes project com-
parisons. Another important
change is accountability. “In the
past, we haven’t gone back and
done measurements after a project
went live to see how much we did
save or how much we didn’t,” says
Cathie Kozik, CIO and senior vice
president.
Tyco Capital, a New Jersey
financial services company, takes
a different approach. To reduce
risk and boost returns, CIO Robert
Plante divides large projects into
smaller phases and measures ROI
along the way, not just on the total
project. This “plan, do, test, react”process enables the company to
test the waters to make sure that
new projects will be successful.
“We’re not going in with guns blaz-
ing, but reducing scale to reduce
risk and size out [IT] investments
appropriately,” Plante says. For
example, installation of a customer
relationship management (CRM)
application took 18 months. Before
the company started each new
phase, previous phases had to
show positive ROIs.Sources: Adapted from Gary H. Anthes, “Pre-
mier 100: ROI for IT Projects Necessary, But
Not Easy,” Computerworld(May 23, 2001),
downloaded from http://www.computerworld.com;
Julia King, “ROI: Make It Bigger, Better,
Faster,” Computerworld(January 1, 2002),
downloaded from http://www.computerworld.com;
Thornton A. May, “Return on Rebellion,”
Computerworld(May 14, 2001), downloaded
from http://www.computerworld.com.FOCUS ONe-FINANCE Putting the “R” Back into ROI