I will not describe the theory of IRR, as I assume that you already are familiar with it, and if not, you can easily get. To begin with, it is not really another method, but simply a variation on the theme of the “Net Present Value” method. So, everything that has already been said about NPV applies equally to IRR.
What I would like to point out is that on a cerebral level, IRR acts as a surrogate to the notion that we only invest in projects that deliver huge sales, coupled with enormous margins, as the existence of that pair of circumstances, tends to have a forgiving effect on errors and miscalculations.
For example, a project with forecasted sales of 1,000 mil USD and a forecasted margin of 50%, even if in reality it manages to make actual sales of 800 mil USD and an actual margin of 40%, it’s still going to create earnings of 320 mil USD (instead of the forecasted 500 mil USD), and thus still yield a very good return.
If anyone lives in a world in which that magical couple of huge sales and enormous margins exists in abundance, I would like to ask that person to send me an invitation to visit that fantasy world. In the world that I live in, the norm is for razor thin margins and cutthroat competition for every penny. Ask anyone who has worked in a super market chain to tell you stories on that.
This is the business reality, and the only serious way to go about it, is by accurate Financial Analysis work. Now you might wonder how is it possible to create serious and accurate work, when the currently existing Financial Analysis methods have so many built-in flaws and inaccuracies. Well, as you have probably guessed by now, the answer is: You can’t. At least, not with those methods which are currently being employed.
Stick around, and we are going to see the problems in the way “Sensitivity Analysis” is currently being performed.