This one is from my list of “top two problems of the Net Present Value method”. Imagine that you are evaluating a “Company based scenario” of an investment. In order for the problem to be more obvious, let’s keep the scenario very simple.

Usually we evaluate the creation of a new company on a 5 or 7 or 10 year time frame. However for the purpose of simplicity of the example, the duration of this evaluation is going to be for only one Fiscal Year (from Jan 1^{st} of 2011, till Dec 31^{st} of 2011). Also, let’s say that only one incident is budgeted to take place, and that is a sale of a service (i.e. no stocks of goods are involved). Let’s also forget about taxes, and let’s say that it takes place on June 30^{th} of 2011 (in the middle of the Fiscal Year), and it will get paid immediately by the customer (i.e. zero days of sales outstanding).

In that oversimplified example that takes place in the middle of the Fiscal Year, according to NPV theory, we must discount the collection by 6 months. I will not get into the issue of choosing the appropriate discount rate, as this problem has already been discussed. However, in a “Company based scenario”, the reality is that the investor does not receive the money on the collection day (June 30^{th} of 2011), but somewhere in July 2012 or August 2012, when the company distributes dividends to its stockholders. That’s 12-13 more months that have not been accounted for in the calculation.

I really wonder how many people have realized that the “Net Present Value” method, when implemented in a “Company based scenario”, is by its own nature inaccurate by a margin on 6-18 months (or on average 12 months).

But then again, if we depart from the above oversimplified scenario, and move into a more realistic one, we will see that not all discounting calculations are off by 12 months. This is relevant only to the part of the sum that contains a profit or a loss. So, in every collection and payment, we must determine the “Profit or Loss” that is included in its sum, and make the adjusted discount. There are two problems on that. The first problem is that the establishing of such a breakdown is not really feasible or practical. The second problem is that the situation has become so complicated and the volume of work so gargantuan, that the only practical thing to do, is to throw in the towel, and forget about NPV as a whole.

Stick around, and be entertained by a completely illogical result that can be achieved by the “Net Present Value” method.