October 24, 2019

NPV vs IRR: Which is Best for Comparing Technology Investments?

Image of someone explaining finance concepts to represent NPV vs IRR.
Image by rawpixel from Pixabay

As a business case analyst for technology investments, I’m mindful of how people use and misuse key financial metrics. Metrics like net present value (NPV), internal rate of return (IRR), return on investment (ROI), and payback period provide insight into the potential value of investments and how they compare to other options. Yet, it’s important to understand how these metrics work and when they can be misleading, particularly when it comes to NPV vs IRR.

NPV tells us the value of an investment in today’s currency and we arrive at that value by applying a discount rate to the expected yearly net benefits. Whereas IRR is simply the discount rate that brings NPV to zero. However, today’s technology investment models, like subscriptions and SaaS (aka cloud economics), have basically made IRR irrelevant. Because, mathematically, the IRR becomes inflated and deceiving when you spread cash flows out over time. This makes NPV a more reliable metric.

Let’s take a hard look at NPV vs IRR, and how we use them, so we can understand why this happens and why NPV is the best method for comparing technology investments.

What is NPV?

Net present value (NPV) represented by an hourglass.
Image by nile from Pixabay

Net present value (NPV) converts the multi-year benefits and costs of an investment into today’s dollars (or other currency value). We accomplish this by doing a discounted cash flow analysis, whereby the multi-year benefits minus costs are “discounted” with a “hurdle rate” to account for the time-value of money (i.e., a dollar today is worth more than a dollar in a year).

The hurdle rate is commonly the firm’s risk-adjusted weighted average cost of capital (WACC). I know, it sounds complicated, but it’s really not. It’s simply the least amount of return the company will accept for an investment. Companies establish this required rate of return after considering many factors such as sources of capital, (like common and preferred stock, long-term debt, and working capital) and expected returns on other investments.

Companies calculate NPV to level the playing field when comparing riskier investments, like acquiring technology or implementing a business process change. NPV provides executive management with a better understanding of the potential impact of each project or investment on shareholder value.

How do you interpret NPV? Anytime you have a positive NPV that’s good. Seriously, it’s that simple. It means positive cash flows outweigh your project costs even when they’re adjusted to account for the time value of money. When you’re comparing two mutually exclusive projects, of course, the higher the NPV, the better.

The Net Present Value Formula

The net present value (NPV) formula.

Now, don’t let the complexity of the formula scare you. Excel makes it easy to calculate the present value of cash with built-in formulas. After reviewing IRR, we’ll look at NPV and IRR examples using Excel.

What is IRR?

Internal rate of return (IRR) is directly related to NPV because the definition of IRR is that it is the discount rate that makes NPV equal zero. Therefore, the formula used to derive IRR looks quite similar to the NPV formula.

The Formula for Deriving Internal Rate of Return

The internal rate of return (IRR) formula showing that IRR is derived from NPV.

Again, don’t worry. We’re not going to solve for IRR from this formula. Done manually, it would require trial and error guessing and, once again, Excel makes calculating IRR easy.

For situations where it could make sense to use IRR, like when there’s a heavy upfront capital investment, an internal rate of return greater than the company’s hurdle rate is what makes the investment plausible. The higher the IRR, the better.

Example NPV and IRR Calculations for a Technology Investment (Using Excel)

Earlier, I suggested that IRR is misleading when looking at a subscription or SaaS technology investment because the “math” gets funky when you spread costs over time. This would infer that IRR is also flawed when looking at consumption models or deferred payments. Let me show you what I mean.

Let’s imagine a company is considering a software acquisition and paying annual maintenance. This is an arrangement we refer to as a “perpetual license.” They would like to compare this to subscribing to the same software and service on an annual basis.

To keep it simple, we’ll assume the perpetual scenario involves using on-premises infrastructure. The subscription option would leverage public cloud infrastructure (e.g., Amazon Web Services). We’ll also call the infrastructure a wash from a cost and benefit perspective.

For the perpetual license, we’ll assume a $500K upfront license cost with 20% annual maintenance and support. For the subscription we’ll assume an annual subscription of $240K per year. We will also presume that the technology, under either licensing model, drives $250K in year one business benefits and $500K of value in years two through five. Here’s how we would calculate IRR and NPV.

NPV and IRR Example Calculations

NPV vs IRR calculation example showing a perpetual license compared to a subscription to show the discrepancy between IRR, NPV, and ROI.

There’s a lot going on here, so let’s peel this back:

  • The top (white) section has all the math for the perpetual license option; the bottom (gray) section is for the subscription
  • The upper three rows for each are for highlighting the benefits, costs, and net benefit (benefits – costs).
  • To keep it simple, I set the discount rate (hurdle rate) for calculating NPV at 10% (upper yellow cell in each section).
  • Therefore, to discount the cash flows, we divide the net benefit for each year by its discount rate. For year one we divide by 1.1. In year two we divide cash flow by 1.21 (1.1 x 1.1 = 1.21). In year three by 1.331 (1.1 x 1.1 x 1.1) and so on. All we’re doing is discounting the net benefit for each year by 10% because each year that cash is worth progressively less in today’s dollars. That’s why, as we go into the future, we add another 10% discount each year.
  • That math allows us to arrive at the NPV for each year’s net benefit, which we then sum up to find the total NPV.
  • To find IRR, we guess a percent value, which is then compounded like the discount rate. We keep guessing the rate (second yellow cell in each section) until the NPV in the “NPV check” row totals zero (far right). Remember, IRR is the discount rate that causes NPV to equal zero.

So, what can we observe regarding NPV vs IRR from this simple example?

  • To ground the observations, let’s first recall that the five-year benefits were the same.
  • The five-year perpetual costs are $200K lower than the five-year subscription costs.
  • As a result, the ROI for the perpetual model (125%) is higher than the subscription model ROI (88%).
  • NPV, which factors time value of money, adds context by normalizing the two investments into today’s dollars. The NPV for the perpetual model is still higher than the subscription model, but only by $52.5K. Since the subscription model spreads costs over time, it reduces the negative impact of the higher five-year costs.
  • However, the IRR for the perpetual model (53%) is lower than the subscription model (103%).

What? How can this be?

Why Do NPV and IRR Disagree When You Spread Cash Flow Over Time?

Image of a money to represent cash flow.
Image by pasja1000 from Pixabay

The internal rate of return in the example above is mathematically correct. You can calculate IRR for most any situation as long as there’s an upfront cost. But, as we just witnessed, IRR mathematically favors low upfront costs. Moreover, when payments are deferred and the first payment takes place after receipt of benefits, calculating IRR is impossible (akin to dividing by zero).

That’s why IRR is a flawed metric when evaluating multi-year subscriptions. It’s not reliable. IRR is better suited for capital acquisitions with a substantial initial investment.

On the other hand, net present value (NPV) yields a rational result. Like IRR, it credits the subscription model for distributing costs over time, but only nominally. NPV favors the scenario that yields higher absolute value and a higher return on investment. While IRR gives the subscription model too much credit.

Still not convinced? Take a look at this crazy example of NPV vs IRR:

NPV vs IRR calculation example showing a perpetual license compared to a subscription, with an unusual scenario for costs to show how it impacts IRR.

Here we used the same perpetual model, but with all costs paid upfront. As one might expect, the total net benefit and ROI are the same, but NPV is lower.

Now, look at the subscription model. Here we double the total costs but start with a low upfront payment of $100K. No one would do this, of course. The total net benefit is negative, so ROI and NPV are also negative. However, you’ll notice that IRR is not only positive, but it’s also higher than the prior example!

Do you need any more reason to choose NPV vs IRR for your analysis?

What is the Difference Between NPV and IRR?

As a reminder, the NPV method of calculating value takes the net benefits (benefits minus costs) and closely considers when they occur. Then it converts them into today’s currency values by discounting them by the company’s required rate of return (aka hurdle rate). IRR is just the discount rate that causes NPV to equal zero.

The calculation for NPV is complex because it combines the time value of money and a company’s minimum expected return. But the resulting metric provides great insight when comparing investment options and making go-no-go decisions. NPV only needs to be greater than zero to make an investment a viable option.

IRR is a percent, so it doesn’t give any indication for the magnitude of the investment or return. As long as IRR is greater than the hurdle rate, the investment is considered viable. However, as demonstrated above, it greatly favors a low upfront cost, so unnaturally overinflates the value of cloud economics, subscriptions, and deferred payment terms.

Both IRR and NPV are dependent upon the time duration for the given analysis. This is why we often state the result as a three-year NPV or five-year IRR.

NPV vs IRR: Why Do They Matter in B2B Sales?

Image of business people shaking hands to represent B2B sales.
Image by rawpixel from Pixabay

When selling B2B solutions, working with the prospect to develop a business case analysis that includes NPV, ROI, and payback is a great way to demonstrate the value of your solution and differentiate yourself from the competition.

The NPV metric converts the quantifiable impact of a purchase into today’s dollars and makes it easier for financial buyers to make comparisons to other investments. If your solution is subscription or SaaS-based, you could also calculate IRR to show the business benefits. But I wouldn’t recommend it because it can be misleading. Payback is much more useful and is a widely accepted metric for presenting the intrinsic value of SaaS solutions. Most SaaS solutions that are easy to deploy and deliver business value will have a payback in the first year.

Regarding developing a business case, B2B sales representatives will want to allow time for this process, pulling in their value engineer or finding third-party resources that provide ROI Analysis Services.

As always, feel free to contact TFP with your financial metric related questions.


Related Reading:

TCO vs ROI: Which is Best for Making Investment Decisions

How to Do an ROI Analysis in 4 Steps