The IRR versus NPV debate

Calculating the Internal Rate of Return (IRR) and/or the Net Present Value (NPV) are
two common techniques of finance in Capital Budgeting. Both these methods recognize
the Time Value of Money (TVM). However, their approach to the problem is slightly
different.

The IRR is a rate, i.e. it is expressed as a percentage. Therefore we may hear

expressions such as “the IRR of this project is 17%”. The NPV is a monetary value. It is
expressed in currency units. We may hear expressions such as “the NPV if this project
is 64.11 Crores”. The question arises as to whether evaluating a set of projects by
either the IRR or NPV method results in the same choice? To determine the answer to
this question we first look at the mathematical definition of IRR and NPV.

As explained above, IRR is a rate. Specifically, it is that rate at which the NPV of the
project is zero. So if the IRR of a project is 8%, it means that at 8% rate of return, the
net present value of the project is zero. It is in some sense a “break even” rate of return.
Intuition tells us that if a project has a high IRR then it is a good project. The important
question here is how do we define “high”? It has to be in reference to some benchmark.
This benchmark is the cost of capital. Thus if our cost of capital is 7% and the IRR of a
project is 8%, then it is giving better returns as compared to the opportunity cost.
Mathematically, if a project has cash flows of 𝐶𝑜 , 𝐶1 , 𝐶2 , . . , 𝐶𝑛 .With 𝐶𝑜 being the initial
cash outflow and 𝐶1 , 𝐶2 , . . , 𝐶𝑛 being the periodic cash flows during the life of the project,
𝑛 𝐶𝑖
then the IRR of the project is the solution to the equation 𝑖=1 (1+𝑟)𝑖 -𝐶𝑜 =0, where 𝑟 is
the unknown (IRR) for which the equation is being solved.

NPV is a monetary value. It expresses a series of cash flows that occur during the
lifetime of the project, as if these were received on the date of commencement of the
project, duly discounting the future cash flows for the time value of money, hence the
term: “Net Present Value”. Mathematically, it is not very different from the IRR
calculation. As with IRR, if a project has cash flows of 𝐶𝑜 , 𝐶1 , 𝐶2 , . . , 𝐶𝑛 .With 𝐶𝑜 being the
initial cash outflow and 𝐶1 , 𝐶2 , . . , 𝐶𝑛 being the periodic cash flows during the life of the
𝐶
project, then the NPV of the project is determined as 𝑁𝑃𝑉 = 𝑛𝑖=1 𝑖
𝑖 − 𝐶𝑜 . The
(1+𝑟)
difference between IRR and NPV is that in case of IRR, NPV was set as zero, and the
rate of return calculated. In case of NPV, the rate of return is taken as the cost of capital
and the NPV is calculated.

Now, that the two concepts have been defined, let us look at the relative advantages
and disadvantages of each method.

Ease of calculation is much higher for NPV as compared to IRR. By observing the
nature of the calculation to be carried out for the IRR, we note that it is computationally
more difficult to calculate the IRR of a project. Also we note that the IRR function is a
polynomial of the form 𝑎𝑜 𝑟 + 𝑎1 𝑟 2 + ⋯ 𝑎𝑛 𝑟 𝑛 . Therefore it is possible for a project to have
multiple IRRs. This aspect becomes clear when we come across projects wherein there
are cash outflows not only at the beginning if the project but also during the period of
the project.

The following example was discussed in class:

Year Project A cash flows Project B cash flows

0 (A) -1000000 -2000000
1 (B) 2000000 8200000
2 (C) -1000000 -6000000
3 (D) 500000 200000
Output of IRR(Values,0) 56.52% -9.59%
The plot of y versus x is shown alongside for both cases.

We note that for project A, there is only one value where NPV is zero, while for
project B there are three solutions.

Project A Project B
In the first case, the NPV is zero at only one point which is the IRR of the project i.e.
56.25%. In the second case, the NPV is zero at three points, r=-96.5%, r=-9.59% and
r=216.09%. Thus, as discussed earlier, an intuitive decision rule when comparing
mutually exclusive projects would be to choose the one with the highest IRR. However,
IRR must be compared not across projects but with the cost of capital. The
benchmarking is with cost of capital.

Consider the following example to highlight this aspect.

Time Project A Cash Flows in Crores of Rupees Project B Cash Flows in Crores of Rupees
0 -2000 -2000
1 400 2000
2 2400 625
IRR 20% 25%
NPV at 5% 557.82 471.66
NPV at 11% 308.25 309.07
NPV at 20% 0.00 100.69
IRR of project B is higher that IRR of project A. We may thus be tempted to select
Project B over Project A. However, as explained earlier, IRR has to be compared not
across projects but with the cost of capital. Let us suppose, the industry cost of capital is
5%. Now, using the NPV calculations, Project A is better than Project B. At cost of
capital of about 11%, both projects are almost identical. At cost of capital greater than
11% Project B is superior to project A. We can appreciate this aspect by plotting the
NPV and cost of capital of the two projects as shown below.

Project A is shown by the red line while Project B is shown by the blue line. We note
that at interest rates (cost of capital) of upto 11% (blue shaded portion of the graph),
Project A has higher NPV than project B, while at interest rates greater than 11% and
upto rate of 25%, Project B has higher NPV (magenta shaded portion of graph). Also at
rates greater than 25% neither Project A nor Project B is advisable. Therefore, choosing
Project A or Project B will depend on the cost of capital. Merely because Project B has a
higher IRR than Project A does not mean that it will always be a best choice. NPV on
the other hand has no such problems, since it is calculated on “cost of capital” basis.

A final example will further make the advantages of NPV over IRR. Consider the value
of education. One may study MBA from a prestigious institute like MDI or from any of
the numerous private institutions. MBA from these private institutions is cheaper than
MBA from MDI. However, salaries of graduates from both types of institutions are also
accordingly scaled. An illustrative example is given below. Cost of an MBA from a less
reputed institution may be about 5 lakhs, while that from MDI may be about 11
lakhs. Salary received may be about 5 lakhs in case of the ordinary institution while it
may be about 7 lakhs. It is assumes that there is 5% growth in salary per year. If we
consider the two cases as two “Projects” with a span of 15 years (after which,
experience will matter more than the educational qualification), the IRR of MBA from
ordinary institution works out to 105%, whereas that from MDI works out to only 68%. A
student may be misguided to select an ordinary institution over MDI if he depends upon
IRR. However if we look at the NPVs, the picture is very different. A student who selects
MDI over an ordinary institution would be better off by about 72 lakhs over the period
of 15 years, at a cost of capital of about 5%. It is only at cost of capital of about 38%
that the two options breakeven. Such high rates of returns are extremely difficult to
achieve.

Time MBA from an Ordinary Institution MBA from MDI

0 (500,000.00) (1,100,000.00)
1 500,000.00 700,000.00
2 525,000.00 735,000.00
3 551,250.00 771,750.00
4 578,812.50 810,337.50
5 607,753.13 850,854.38
6 638,140.78 893,397.09
7 670,047.82 938,066.95
8 703,550.21 984,970.30
9 738,727.72 1,034,218.81
10 775,664.11 1,085,929.75
11 814,447.31 1,140,226.24
12 855,169.68 1,197,237.55
13 897,928.16 1,257,099.43
14 942,824.57 1,319,954.40
15 989,965.80 1,385,952.12
IRR 105.00% 68.58%

Cost of NPV of MBA from an NPV of MBA from MDI Excess NPV of MDI
Capital Ordinary Institution MBA over Ordinary
MBA
5% 6,642,857.14 13,900,000.00 7,257,142.86
10% 4,523,211.23 9,448,743.58 4,925,532.35
20% 2,383,553.95 4,955,463.30 2,571,909.35
NPV at38% 990,026.33 986,036.86 (3,989.47)

Hence, in conclusion, we note that NPV is a more reliable method for selecting a project
in a Capital Budgeting exercise as compared to IRR.