Harvard Business School 5-187-059
October 28, 1986
Citibank Indonesia
Teaching Note
Purpose of Class
The Citibank Indonesia case (9-185-061) is intended to allow students to discuss some of the
important issues related to the design and operation of planning and budgeting systems. The
case also provides some opportunities for secondary learning about management of a
Suggested Assignment
Watch 28 minute videotape before class.
Question: What should Mehli Mistri do about the budget issue described on page 1 of the case?
Discussion
In the case, Mehli Mistri, Citibanks country manager for Indonesia, is faced with a issue related
to his budget for 1984. He has prepared a bottom-up budget which he feels is very aggressive,
but his superior is now asking him to raise the profit target by between $500,000 to $1,000,000.
Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructors Manual
This request (demand?) for a higher target comes from corporate headquarters. During the
normal budget review process, the managers of the institutional bank came to the conclusion
that the consolidated profit plan was inadequate. They raised the targets for the groups
(geographical regions), and then managers of the groups and the divisions were faced with the
task of allocating the higher targets down to the country level.
In preparing the case, the students are asked to put themselves in Mehli Mistris position and to
Pedagogy
The times I have used this case, I found it useful to start with some introductory comments on
planning and budgeting systems. The need for such comments, of course, will vary depending
on the students in the class and the material that precedes the use of the case.
Here is an outline of the lecture I gave:
1. Definition: A budgeting system is a combination of information flows and administrative
3. There are three planning cycles in well-developed processes in large decentralized firms:
4. The planning and budgeting processes cut across all of the types of controls in my control
framework.
The links with results control are obvious, as budgeting involves setting targets which
are often used later as standards against which to evaluate performance.
221
5. Planning and budgeting processes serve both planning and control ends. Early in the process
6. There are a whole range of variables that can be considered in designing planning and
budgeting processes. One way to look at the choices is in terms of the following variables:
Content forms that need to be filled out. Many formats are standard, such as pro
forma income statements and balance sheets, and cash flow forecasts, but many others
are not.
People who is involved?
One of the important choices that must be made is how fast to narrow the span of strategic
options being considered, as described in Lorange and Vancil (1976).17 The choice that is made
can be implemented by shaping combinations of the above four types of variables.
7. Planning and budgeting systems must differ depending on the companys situation; for
8. The design and operation of planning and budgeting systems is an art, not a science. I like to
point out that my first job after my I received my MBA involved designing the annual
Early in the class I like to develop the diagram of Citibanks budgeting process shown in Figure
TN-1. This makes several important points which can be discussed if desired. First, Citibanks
process is relatively compact, as compared with some other processes that students may have
seen (e.g., the ITT process described in Galvor). Second, it shows that the intent is bottom-up
Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructors Manual
decentralized company, and how an operating budget can be prepared before the sovereign risk
limits have been set.
The concept of sovereign risk is new to many students. I like to view this process as just one of
the key elements in Citibanks resource allocation process. They are asking the question: What
markets do they want to be in? Within the confines of the charter of the institutional bank, a key
Here are some other issues that can be discussed:
1. Citibank has no long-term plan. As mentioned in the case (p. 3), the managers felt that the
numbers looking out 35 years were too soft. That suggests the question as to when long-
term plans are useful. Are they useless in situations of very high uncertainty? Or is that
when they are most needed?
2. Citibank does have long-term goals. Are these helpful or harmful? Should they change over
time, for example, with changes in products offered, markets served, or U.S. inflation rates?
3. Is there a danger that the last-minute, top-down goal-setting process described in the case
will cause a loss of commitment to the goal? In the abstract, one would argue yes, but Mehli
4. Is there a danger of Mehli and other country managers taking actions to meet their short-run
target at the expense of the long-term good of Citibank? Yes, clearly there is. Many critics
have pointed out the management-myopia danger in industrial settings. Here is the same
issue in a bank context. Here are some possibilities for short-term actions:
Eliminate the low-earning loans (e.g., to the government). I pose the issue as follows,
using the terminology we heard in Indonesia: Assume Mehli gets the following call
from the lead bank manager organizing a loan syndication for the government. At the
express invitation of the Republic of Indonesia, I am delighted to invite you to join us in
participating in this loan. The rate will be at 3/8% over LIBOR. What should Mehlis
reaction be? His actual reaction, he said, was Oh no! But he also said that Citibanks
5. Does the top-down goal-setting, which Mr. Mistri says on the homework videotape is
relatively typical, create a culture of sandbagging when submitting budgets? Mehli Mistri
223
6. Only U.S. dollars are important to Citibank. Most of Citibanks shareholders are in the U.S.,
but this dollar focus can make life very difficult (or easy) for a country managers depending
7. How should Mr. Gibson allocate the $4 million increase he has committed to? On basis of
Pedagogy
In a class of eighty minutes, I show the in-class videotape after about 4550 minutes of class
discussion. The first discussion should clarify case facts (e.g., sovereign risk), raise issues and
get students to make the judgment Mehli Mistri had to make. He had four alternatives:
Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructors Manual
224
Wrap-up
At the end of class, I think at least the two following points should be made. The first is that
budgeting systems are asked to play multiple roles, and no one system can do an optimal job at
all of the roles. I develop this point on the blackboard using the format shown in Figure TN-2.
Some of the ideas are adapted from Barrett and Fraser (1977),19 and if desired this article can be
used as part of the student assignment for this class. I suggest that firms use budgeting systems
for five main purposes: motivation, planning, coordination, cost control and evaluation. Then I
pick one important design variabletarget difficultyand list three levels of difficulty we
could include in budgets: optimistic (aggressive), best guess as to what will occur, conservative.
To illustrate the no-one-system-is-best point, I ask students what type of budget target best
serves what budget purpose. The answers are as follows, I think:
Merchant & Van der Stede, Management Control Systems, 3rd edition, Instructors Manual
225
FIGURE TN-1 Budget Process
FIGURE TN-2 Matching Budget Targets with Purposes