Memo
To: RiskAnal Colleagues
From: Ray Martin
Date: July 25, 1999
Subj: Why NPV is Good
Answer: Because IRR is Bad for one thing. Click on Brealey and Myers, chapter 5: Why Net Present Value Leads to Better Investment Decisions to see why it is bad. Scroll down to the bottom of the page.
However, IRR is not really so bad or difficult. The Relevant Internal Rate of Return (RIRR) is 15.24 percent. The RIRR in red below gives an indication of relative risk if you have something with which to compare it. It is consistent with NPV. I had no problem with it at all. It took one click after loading the data to calculate it.
The other root of -50 percent cited has been rejected in the literature dating back to at least the early 1980s. I am unaware of any challenges to the rejection of a negative root, especially with net positive cash flows. It requires zero lines of code in the DCF program to rule it out. Excluding it is a no brainer. So, other than being a root of a quadratic equation, if anyone can make the case for -50 percent being a relevant "IRR" with the cash flows below, or knows of it having been made, please contact me on or off the list.
The data in this example are realistic by the way. Having an expenditure exceeding revenue at the end of a project could occur, for example, if teardown and disassembly or site restoration were needed or mandated. Oftentimes the data used to make the IRR is Bad case are unrealistic. However, it makes no difference. If IRR is incalculable for any reason, NPV is suspect and very likely nonsensical as well. And discounted cash flow is unnecessary in order to decide between alternatives. The financial implications are clear in the undiscounted cash flows themselves.
D i s c o u n t e d C a s h F l o w ---------------------------------------
Prepared by | R. Martin Date prepared | 07/27/99 09:22:20 AM Description | BAonethirty5.html Study period | 6 years Discount rate | 10.00 percent Discount method| End-of-year ---------------------------------------------------------------------------- Project Expenditures Revenues Net Cash Discount Discounted Year (-) (+) Flow Factor Cash Flow ---------------------------------------------------------------------------- 0 $1,000.00 $0.00 -$1,000.00 1.0000 -$1,000.00 1 $0.00 $800.00 $800.00 0.9091 $727.27 2 $0.00 $150.00 $150.00 0.8264 $123.97 3 $0.00 $150.00 $150.00 0.7513 $112.70 4 $0.00 $150.00 $150.00 0.6830 $102.45 5 $0.00 $150.00 $150.00 0.6209 $93.14 6 $150.00 $0.00 -$150.00 0.5645 -$84.67 ---------------------------------------------------------------------------- Total $1,150.00 $1,400.00 $250.00 5.3553 $74.86 ---------------------------------------------------------------------------- Number of years to payback [discounted] | 3.3 [ 4.4] Uniform annual inflow without terminal value | $13.98 Internal rate of return (IRR)(Sign +/-: 2) | 15.24 percent IRR--discount rate differential | 5.24 percent ----------------------------------------------------------------------------
Disclaimer. I:
POSTED ON RiskAnal DISCUSSION LIST
----- Original Message -----
From: Ray Martin <martintx@flash.net>
To: Risk Analysis Discussion List <riskanal@listserv.pnl.gov>
Sent: Sunday, July 25, 1999 3:53 PM
Subject: Re: Separability of Risk and Return
The Dyer and Jianmin and vice versa papers are quite
useful. I recommend them as well.
However, they work with undiscounted value, i.e., the time value of money is
excluded--ignored. This is a popular way of dealing with it--by not dealing with it.
And that is fine. Discounting itself can be troublesome and introduces its
own complexity that can detract from the thesis.
But somewhere you need to work time value back in. It is not a trifling driver.
The key financial variables are how much and when, not just how much.
If you want a mainstream risk view from the perspective of the finance discipline, I
suggest http://econ161.berkeley.edu/Teaching_Folder/BA_130_F96/BAonethirty.html.
The most relevant chapters are 2, 5, 7, and 12.
The short answer to "Why NPV is Good" in Chapter 5 is that "IRR is
Bad." See http://econ161.berkeley.edu/Teaching_Folder/BA_130_F96/BAonethirty5.html,
near the bottom. If you want to see why IRR is not bad go to https://members.tripod.com/~Ray_Martin/NSF9914/DeLong.html.
NPV and IRR are best used together. They are complementary. But one of the
things IRR gives that NPV lacks is an indication of risk. Only an indication mind
you. Somewhat like the relationship between risk and yield in the bond market.
Thank you for your inputs. Please keep them coming. On or off the list.
Ray_Martin@AltaVista.net or MartinTX@Flash.net
---- snip, snip/chop/RM
----- Original Message -----
From: <(A List Member from Colorado, USA)
Sent: Sunday, July 25, 1999 8:24 AM
> In a message dated 7/24/99 2:54:06 PM Pacific Daylight Time,
> (A List Member from Russia) writes:
>
> Actually, there is a very nice "risk-value" theory that I strongly
recommend
> to folks who care about definitions of risk and who like mathematical approaches.
> -- Tony