Let’s start to run through the math/accounting of calculating ROIC. There are lots of different versions, but basically, the ratio is:
| Return On Invested Capital = NOPAT/Invested Capital |
Note: A good breakdown of the basic ROIC formula can be found in Aswath Damodaran’s Return On Capital (2007) beginning on page 7. This piece is excellent, should be read in its entirety, and will be referenced repeatedly in future posts.
Right off the bat, let’s hit some simple FAQs. I will do this in a few separate posts.
Which time period(s)?
The capital was invested, and then the return came in, in the form of operating income after tax (NOPAT). So we look at Invested Capital (see below) at the beginning of the year (balance sheet from the end of the previous year), and the operating income from the year being analyzed – that’s what the business used to generate the return. Many people use the average of the book value of invested capital during the year, i.e., the balance sheet from before the year of income and the balance sheet at the end, the idea being that an average is a closer estimate of the capital that generated the income during that year. This is especially relevant if the balance sheet changes significantly during the year (e.g., a lot of capital is raised or generated internally, or an M&A transaction). Sure, if you want to. You have the numbers right there on the balance sheet, your call.
People will usually calculate ROIC for each year for a company to create a time series, say of the last 3 -10 years, and do the same for comparable companies, so we are shooting for some kind of normalized level either way. Don’t forget, we are using ROIC for a general sense of capital productivity, this is not a statistic that is about extreme precision, regardless of how you do it.
Taxes?
NOPAT = Operating Income x (1 – tax rate). Where does the tax rate number come from? Look at the company’s effective tax rate historically, and the effective tax rate of comps, and ballpark it with some kind of average that seems reasonable. You might be tempted to use actual taxes paid – maybe. Damodaran (page 7) points out a common mistake people make when doing this: they ignore the impact that tax-deductible interest (i.e., the tax shield) has on the effective rate – this should be unwound to get to the rate to use here as we are looking at a pre-tax number. The Federal corporate tax rate is about 21% and then you have state taxes. I believe CapitalIQ uses a fixed number of 25-30% for its ROIC calculations, you can check it out to confirm. Something in that range is probably the right number to use; say 25%. If you think you have a good handle on the actual effective rate and its generally representative of what the company pays, use that, but remember that then you need to do the same for any companies you are comparing your target company to and that can be a bit time consuming – this is probably why Greenblatt (see below) uses a pre-tax operating income number and calls it a day. I just use 25% unless I see a very good reason not to. Here is a lot of data on effective tax rates by industry that may influence your thinking for the particular company you are looking at.
Why/Which Income Numerator & How To Think About It?
The idea is to approximate normalized, unlevered, pre-tax cash earnings of the business as the numerator and then tax effect it (see below). That’s the number that most likely reflects the underlying economics of the business. D&A is not a cash expense but it is included as an expense in getting to operating income for ROIC purposes. Why isn’t it backed out? Because the assumption is that it approximates cap ex and the two cancel each other out so that you are getting something like a cash earnings number for the ROIC numerator by using operating income.
In The Little Book That Beats The Market, for the numerator, Joel Greenblatt uses:
(Earnings Before Interest & Taxes + Depreciation – CapEx)
So in this articulation, you can plug in different numbers for Depreciation & Cap Ex if you don’t want to assume they offset one another 1:1 for whatever reason. Note also that Greenblatt does NOT tax effect the numerator, while most people do. He is probably of the view that ROIC is being used for comparative purposes anyway (in fact, that’s how the Magic Formula works) not in an absolute sense, so why add the variable of taxes which may create some data noise in this process. Every company being compared is pre-tax, period.
One-timers, Non-Operating Items, and Normalized Numbers
Don’t be confused by some other versions of tax effected operating income for the numerator.
EBIT and tax effected Operating Income technically are not the exact same thing because of the inclusion of non-operating items in EBIT (see simple wiki explanation here). Some examples: restructuring charges are non-cash and one-off and probably should be pulled out. If R&D and operating leases are capitalized, R&D expense should be pulled out of NOPAT, which should also be adjusted for the lease expense (it may not be as simple as pulling out the lease expense). More on this in future posts. But the bottom line is that we want the most representative and dependable level of NOPAT that we can get.
In Michael Mauboussin’s Calculating Return On Invested Capital (2014) (also strongly recommended) he calls the numerator of the ROIC ratio Net Operating Profit After Tax (NOPAT). It’s the same as the tax-effected operating income number you are getting using any of the versions above. So if your EBIT number is clean of non-operating income and other funky stuff, it’s the same as NOPAT for practical purposes. Remember that the goal of the ROIC calculation is to get to some kind of back-of-the-envelope estimate of capital productivity, and to do this you can adjust numbers for a given year, use averages from several years, or make any other changes you think bring you closer to a truer estimate of this productivity level. There is no right answer.
What’s Invested Capital, Exactly?
It’s the denominator in the ROIC ratio. What we are trying to get at here is the assets used by the operating business (aka asset intensity) in getting that income as a return, so that we get a sense of the overall productivity of the business in using its assets to create value each year. So in the same way that we don’t want non-operating items in the numerator, we don’t want investments in other businesses or securities, cash that is not needed to operate the business (see below), or anything else (e.g., a fleet of corporate jets used mostly for personal purposes?) that might sit on the balance sheet but that is not deployed by the operating business itself. Going back to the Greenblatt story, it’s all the assets needed to make and sell gum, nothing more and nothing less.
You should go item-by-item on the balance sheet to make sure you are getting this right. Look at how Mauboussin walks through it in Calculcating Return On Invested Capital and ROIC: How To Handle Common Issues (2022). Specifically, he talks about the Operating vs. Finance approaches to getting invested capital and the details of them. They are also discussed in this blog here.
This is a great exercise because you should know these balance sheet items well anyway if you are a serious investment analyst. I will go through many examples over time.
Here’s some good advice on thinking about the denominator from Mauboussin:
“If you are wondering about whether to include an item in invested capital, simply ask if the company could generate the same level of NOPAT without the item. If not, include it. If so, exclude it.”
Well said.
Note that Greenblatt uses Net Working Capital + Net Fixed Assets in the denominator, which is simply a rephrasing of Book Value of Operating Assets above using the left side of the balance sheet instead of the right side (equity +debt), and Moubassin just calls it Invested Capital. This is confusing but they are all essentially the same thing (assuming you are using only tangible assets in both cases), except to the extent that an analyst has good reasons to include/exclude certain items based on their own interpretation of what the business’ productive operating assets are or aren’t. Let’s break down the two components of Greenblatt’s version:
Net Working Capital & Cash
For the unfamiliar, NWC is just Current Assets minus Current Liabilities. The idea is that some or maybe all of current liabilities are an infinite, “free” source of financing for current assets, so they should be backed out, because the only real investment for the business is the net amount. Mauboussin, in Calculating Return On Invested Capital refers to “Non-interest-bearing current liabilities” (NIBCLs) as what should be netted against Current Assets i.e., if there is an interest charge it’s not free so it should not be netted. So NIBCLs are a subset of Current Liabilities. You should probably go item-by-item in CLs so that you make sure you are only deducting NIBCLs, but if you are doing a fast and dirty, just netting out CLs is fine.
Mauboussin (page 7) also deducts non-operating cash from Current Assets to get to NWC, and this is a common practice in financial modeling. This is accomplished by assuming that only a percentage of the cash on the balance sheet (e.g., 5%) is actually needed in operations and including only that amount in Current Assets. I wouldn’t be too mechanical about this – you should only do it if you have a good idea of how much cash the business actually does need to operate, based on historic working capital levels, operating cash flow, and what you know about the company’s plans for the future and what cash that might require.
There is a lot out there about the virtues of a negative net working capital business. This is beyond the scope of this post and probably something I will discuss in a later post. But big picture: businesses that collect sales immediately from customer credit cards have almost no accounts receivable, and no inventory if they are software or service businesses, and may be able to delay paying vendors (current liabilities), creating a negative working capital business. This is great because the business is partially being financed by these vendors, indefinitely. Contrast this with a more traditional business that holds lots of inventory (more and more as its business grows) and waits say 60 days + to actually be paid (account receivable). This business will have positive working capital which means that as it grows it must invest more and more capital in current accounts – this dilutes capital productivity, and is usually modeled by assuming some percentage of sales must be invested in NWC indefinitely. It’s harder and more risky to grow a business when it constantly requires more incremental investment in current assets, not to mention long-term assets.
Intangible & Other Assets
A long time ago PP&E was the main non-working capital asset used to create income, so that’s what analysts focused on in the ROIC equation. But this is not the case anymore and grows less true every day. Intangibles play a very big role in how companies generate income, especially the more highly valued ones, and being thoughtful about them is critical to getting reliable ROIC numbers. Should goodwill and other intangibles be included in the ROIC denominator? This is discussed here. Short answer: probably not. Intangibles aside, you need to scan the balance sheet to make sure you aren’t missing important assets of any kind being used by the business even if they are not directly addressed in the ROIC formula you are using. Here’s one of many examples on this blog of recasting intangible investment for a company to get a better handle on ROIC.
More discussion of ROIC calculation items in the next post.