IT Capital Budgetting

Indiana University of Pennsylvania

Dr. James Rodger

Fall 2010 Ankur Dhakal

Justin Joseph

Seok Joo Kim Introduction What is the smart data? (Ankur) What is the Capital budgeting? (Justin) What is the difference between IT and traditional capital budgeting? (Kim) Can you tell us something about Situational Model? (Ankur) Would you explain about the overall structure and results of experiments? (Kim) Tell us something more precisely about 4 methods?(Ankur) Would you tell me about the Simulation results of 4 Methods?(Justin) Do you think there is any potential improvements and extensions of this

research?(Kim) Do you think there is any potential improvements and extensions of this

research?(Ankur) Tell the audience something about our conclusion?(Justin) What is the smart data?(Kim ask)

Ankur answer

Smart data is the comprehensive concept to use data smartly to manage the organization more efficiently

In relation to the MIS, the use of smart data concept can be

a great tool for solving the organization problem because

making data smart enables the organization to have certain

superior characteristics

By smart data, the data are interoperable and readily

exchangeable among qualified members of an enterprise

And it also contributes to optimizing enterprise performance

by providing a strategy that will vastly improve enterprise

data resource management

So, we will use this smart data concept to select the

effective methods to solve the ITCB problem What is the Capital budgeting ?(Ankur ask)

Justin answer

The capital budgeting problem is the problem of selecting a set of capital expenditures that

satisfy certain financial and resource

constraints

And, the type of capital budgeting problem

considered in our research is sometimes

called a hard capital rationing problem

Because a decision-maker has to select the

best project mix among several competing

projects under the budget constraint such as What is the difference between IT and traditional

capital budgeting?(Justin ask)

Kim answer

Capital budgeting in IT is slightly different from the traditional capital budgeting problem

Because a certain investments are mandatory, the

investments size is large and only certain investments

can be depreciated

About the depreciation issue, IRS provides a

maximum upper limit on the IT depreciation expense

deduction

So, in this study, we will consider these constraints

about finding the optimal solution of the ITCB problem Can you tell us something about Situational

Model? (Kim ask)

Ankur answer

This method used e-based antilogarithm transformation of revenue and IT budgeting variable to compute the

actual values of revenue and IT budget for analysis

A general ITCB problem can be mathematically

represented as the binary variable knapsack

optimization problem

where objective function is a non-linear function that

maximizes certain managerial criterion such as after-tax

profit of a set of IT investments Would you explain about the overall structure

and results of experiments? (Justin ask)

Kim answer

Using several simulations, this research empirically compared the performance of two SA heuristic

procedures with the performance of two well-known

traditional ranking methods for capital budgeting

For the purpose of benchmarking, this research

used two simple ranking methods that can be used

to solve the ITCB problem

According to the experiment result, the heuristic

approaches was best suited for solving the ITCB

problem Tell us something more precisely about 4

methods?(Kim ask)

Ankur answer

4 Different approches were used A Rank D Rank FRSA (Feasibility Restoring Stimulate Annealing) SSA (Simple Stimulate Annealing) CONTINUED Projects ranked in the acending order of their

investment value ,before their selection using

the budget constrai, then the ranking is called

A-RANK.

Projects ranked in the decending order of

their investment value ,before their selection

using the budget constrai, then the ranking is

called D-RANK. CONTINUED

And, we developed two simulated annealing (SA) procedures for solving the ITCB problem

Our first SA procedure(FRSA) utilizes a feasibility

restoration component. The feasibility restoration always

maintains a feasible solution

In our second procedure(SSA), we eliminate the

feasibility restoration step from the first SA procedure.

In the event when a solution is not feasible, failure occurs

in our second SA procedure and annealing schedule is

omitted

This simulated annealing can deal with arbitrary systems

and cost functions and statistically guarantees finding an

optimal solution. This method is relatively easy to code,

even for complex problems Would you tell me about the Simulation results of 4

Methods?(Ankur ask)

Justin answer

About the After-Tax Profit, FRSA, SSA outperformed A-Rank, D-Rank in the maximization of after-tax

profit. But, difference of profits was not significant

between A-Rank and D-Rank because of t he existence of common projects And, the difference of profits was not significant between FRSA and SSA. It resulted from the weak effects of feasibility restoration continued

About the Number of Projects selected, Ranking methods selected more projects than

2 SA methods. It was caused by the Ranking

methods are biased toward smaller or larger

projects. But, SA methods have preference for

the projects which increase the after-tax profit

And, about the pairwise differences in means

comparison, we couldn’t find no significant

difference in FRSA and SSA, but, found the

significant difference in A-Rank and D-Rank Would you make the summary of simmulation

results then?(Justin ask)

Kim answer

Conclusively, Two Smart Data SA heuristic methods(FRSA, SSA) outperformed traditional

methods(A,D-rank) Given that 0-1 knapsack optimization problem is NPhard, Simulation model for 500 companies proved

these facts

And, we can argue that this research contribute the

optimization of IT capital budgeting Because the use of corporate taxation and

depreciation for selecting the profitable IT investments

can improve the organizational profitability Do you think there is any potential improvements

and extensions of this research?(Kim ask)

Ankur answer

First, The objective function can be changed into the maximization of NPV if the management don’t

want the tax-induced interaction

Second, other heuristic methods such as Genetic

Algorithm and Tabu Search can be used because

single investment budgeting period is not realistic

And, finally, for more practical application, future

research is needed for solving the multi-period

ITCB problem Tell the audience something about our conclusion?(Ankur ask)

Justin answer

This case showed that how Smart Data approach could improve the efficiency and

effectiveness of capital budgeting in IT

industry

The utilization of smart data AI methods such

as SSA and FRSA could outperform the

traditional capital budgeting methods such as

A-Rank and D-Rank methods

Conclusively, the organization can select the

IT projects which lead to the highest after-tax

profits and optimizing enterprise performance,

by applying the Smart Data approach Management