Tuesday, January 23, 2007

Lies, Damn Lies, and Net Present Value

When you make a proposal for an investment in the development of something new, you are competing against all of the other potential ways that your funding source might invest their money. This is true whether you are a start-up pitching a new business plan to a venture capital firm, a market manager trying to sell new product or feature development to your upper management, or even a researcher making a proposal for a new research project to a funding agency.

Some of the competitors are obvious: other start-ups, other new products or features under consideration by your company, and other research proposals. But what isn't so obvious to most technologists, but is burned into the mind of every M.B.A., is that you are also competing against the stock market. You have to convince your funding source (or, as I like to think of them, money guys) that they would be better off giving you the money than simply investing it in the stocks or other financial instruments.

The bad news is that if the stock market is doing pretty well, or if the perceived risk of investment in stocks is a lot lower than the risk of investing in your project, you will be fighting an uphill battle. The good news is that there is a way to calculate how you stand compared to your money guys just calling their broker. This calculation is called net present value (NPV). It accounts for the time value of money by calculating an equivalent present value of an investment whose return will be made over a span of of time. It is a way of comparing apples to apples when evaluating several investment options.

Since most technologists are hard-wired for math, this is quite likely going to make a lot of sense. While I am most definitely a technologist (I just took business classes as electives as an undergraduate for fun), I found that understanding just a little bit about NPV gave me a lot of insight into how and why new product decisions are made the way they are, and why certain projections seemed to be unrealistically inflated.

NPV has just three variables. It is the choice of the values for these variables around which all discussion with your funding source will revolve. It is pointless to argue about the use of NPV, or how NPV is calculated, because the money guys will know that this is a standard business tool, and they will likely have been using it for years. But battles will be won or lost by the choices made, and the justifications offered, for these variables. The reliability of the value that NPV computes will be no better than the reliability of the values chosen for these variables.

Duration: this is the total life span of your project and the product that it creates, from the first year you start spending money to develop it until when you finally decide to discontinue the product. Don't kid yourself: your money guys know that all products have a life span. High-tech products tend to have shorter life spans than most others, more like fresh produce than building materials. The duration is in units of years (although I can imagine some high-tech projects in units of months). Let's call this this variable N.

Annual Net Income: this is the amount of income that would be realized by your product year by year over its entire duration after all costs are subtracted out. It is likely not be the same for every year. For example, if it takes two years for your product to hit the market, for the first two years this number will be negative, representing the initial development costs with no income. In subsequent years in which this number is positive, it will represent the income generated by your product minus any costs associated with manufacturing, support or maintenance. And even if your product is successful, its annual net income will typically rise until it saturates the market and than fall. If all of your annual net income numbers are zero or negative, you are already screwed. We'll call this variable A[i] where i is the elapsed years that runs from 0 to N-1.

Cost of Capital: this is effectively an interest rate that captures what interest could be made year by year by investing the funding you are asking for in a financial instrument with a similar risk profile as your proposal and over the same duration. Note that this captures not just how well alternative investments might do, but how their risk compares to yours. You could argue that this may change from year to year, although in all the NPV calculations I have ever seen it is assumed to be constant. Unless you have the amazing secret to wealth, it will be a lot less than one, like 0.1. We'll call this variable C.

Here is the formula for NPV. Note the the caret "^" is a "raised to the power of" operator, as in 3^2 equals 9. (I've seen slightly different formulations of NPV from different sources, so your mileage may vary a little, but they are all equivalent.)

NPV = sum (i = 0 to N-1) ( A[i] / ((1 + C) ^ i) )

Let's look at an example.

You have a great idea for a new product, the flux capacitor. You think it will cost one million dollars spread across two years to develop and productize it.

Since you have some equipment you have to buy and some licenses you have to purchase, you have some up front costs the first year that you don't expect to have in the second year, so the expenditures for those two years of development are a little front loaded.

You expect the life span of the flux capacitor to be eight years before it becomes obsolete. That gives you an N of ten years: two years of development plus eight years of sales. You expect to hit your sales peak around the fourth year after introduction.

Here is your projected net income, or A[i], year by year for the total duration of the flux capacitor product, based on how many units you think you can sell, at what price, and what the unit cost of each flux capacitor is likely to be.

A[0] = -$700,000
A[1] = -$300,000
A[2] = $50,000
A[3] = $100,000
A[4] = $250,000
A[5] = $500,000
A[6] = $250,000
A[7] = $100,000
A[8] = $70,000
A[9] = $20,000

So far this doesn't look too bad. For an outlay of $1,000,000 over two years, you make back a total of $1,340,000. That's a net income of $340,000 or 34% of your initial investment.

You still need a value to use for C, the cost of capital. Here's the thing: you don't provide that number. Your money guys are going to have someone with a degree in accounting or business administration who is going to provide that number. They are going to base it on lots of factors, like the Dow Jones Industrial Average, the return on investment (ROI) of their existing investments, maybe the ROI of one or more stock index funds, or what they read in the Wall Street Journal this morning. The WSJ is actually the most likely source: it regularly publishes numbers just for this purpose. No way are you going to have more credibility than the WSJ. You have no control over the value used for C. Let's say your money guys use a cost of capital of ten percent or 0.1.

Here is our calculation for NPV. (Note that I'm just showing two decimal places,but my calculator keeps more.)

NPV = -700000/(1.1^0) - 300000/(1.1^1) + 50000/(1.1^2) + 100000/(1.1^3) + 250000/(1.1^4) + 500000/(1.1^5) + 250000/(1.1^6) + 100000/(1.1^7) + 70000/(1.1^8) + 20000/(1.1^9)

NPV = -700000/1.00 - 300000/1.10 + 50000/1.21 + 100000/1.33 + 250000/1.46 + 500000/1.61 + 250000/1.77 + 100000/1.95 + 70000/2.14 + 20000/2.36

NPV = -700000 - 272727 + 41322 + 75131 + 170753 + 310460 + 141118 + 51316 + 32656 + 8482

NPV = -141489

Yeah, you are interpreting that number correctly. Even though you think your project will make $340,000 over its ten year run, your money guys would have been far better off just investing the money in the stock market. If everything goes your way, they stand to lose $141,489 by betting $1,000,000 on you. You are in a battle against compound interest, and that is a hard battle to win. Just be glad your 401(k) is subject to the same math.

If you have any brains at all, you will have done this calculation long before you ever appear in front of the money guys, and you already know you are in trouble. So what can you do? You only have a small number of variables that you can affect. See if any of these strategies sound familiar based on your own experience in product development.

You can ask for a smaller initial investment. We can hire just half as many developers, and have them work twice as many hours. We can buy half as many workstations and tell our developers: it's pair programming! We can ask for less money up front in the hopes that once the money guys commit, we can get more funding later.

You can get the project done in less time so that you start making money sooner and for a longer period of time. We can hire twice as many developers to get it done in half the time. We can skip those time-consuming requirements and design phases. We can claim that we can get it done in half the time in the hopes that once the money guys commit, we can get more funding later.

You can extend the sales life span of the flux capacitor. We're pretty sure that even after the oil economy collapses, there will still be a market for flux capacitors. It's unlikely that our flux capacitor design will become obsolete in less than fifteen years; after all, we're designing for expandability!

You can reduce the on-going cost associated with each flux capacitor you sell. This on-going cost is known as the cost of sales (COS) or sometimes cost of good sold (COGS). We won't need technical support. We can offshore technical support. We don't need a marketing or sales department, these babies will sell themselves! We can use cheaper materials; after all, if it breaks, then maybe we'll sell another one.

You can increase the sales price of each flux capacitor. Of course, this is tricky. There's this little issue of reducing sales by increasing the price.

You can decrease the sales price of each flux capacitor. Everyone loves a sale! We can offer a rebate. How do we do it? Volume!

You can increase the projected annual sales. Once the oil economy collapses, we're positive that everyone is going to need two things: a 9mm Glock and a flux capacitor! We'll be lucky if we can keep up with demand.

And probably some other strategies I haven't thought of, or haven't experienced first hand.

I can see the light bulb going on over your head. This explains a lot. Once I started really thinking about NPV, I began to wonder how any new products were ever developed at all. Of course, if no new products were ever developed, there would be a lot fewer stocks competing for investment dollars, because all the manufacturing companies would eventually go out of business. So there is a kind of dynamic balance going on that impacts the entire global economy. A Nobel prize or two has probably been awarded over the analysis of this very thing.

I've heard engineers claim that an analysis like NPV is short sighted. That's easy to say when it's not their money. They might feel differently if they realized that they might be talking about the money in their own 401(k)s. At least for publicly traded companies, I also think it's important to remember that the money guys have a fiduciary duty to their shareholders, a legal responsibility to make the best use of the shareholders money. The shareholders are literally the owners of the company. They would much rather get that money themselves than have the officers with whom they entrusted their company fritter it away making flux capacitors. Failure to live up to fiduciary duty can result in, best case, shareholder lawsuits, or, worst case, in FBI and SEC investigators pulling up in black SUVs. Somewhere in the middle is where the officers of the company are fired.

As a young technologist, I found it hard to understand the business process. It was a lot easier, and more fun, to read about that new fangled language, C. But as I got more and more experience, and probably as I got more and more money in my 401(k), I began to appreciate the kinds of tough decisions managers of companies have to make on a daily basis about what to make, how to make it, and why.


Daniel P. Golman, et al., "Calculating Net Present Value", Business, Perseus Publishing, 2002

Barry Karafin, "Business Management for Technologists", (course slides), September 2005

Randy C. Perry, David W. Bacon, "The Business Case for Design for Six Sigma", (e-book), Prentice-Hall, 2007


Anonymous said...

I appreciate this article on NPV, since I am putting together figures to raise funds for a film, no less.

It's an extremely high risk ball game where hard cold facts and figures are very difficult to come by for a first timer.

Now, I can throw into the mix this lovely equation! It could be discouraging, but no, it's simply reality. Thanks for helping me prepare. I go to talk to a money guy tomorrow!

Chip Overclock said...

I've sometimes wondered if art would be better if it were unfettered by the realities of economics. But I suspect that nothing is really unfettered by the realities of economics.

Good luck!