Thoughts on ROIC

I started reviewing Constellation Software’s (CNSWF/CSU) 2018 annual report recently and found its calculation of return on invested capital (“ROIC”) interesting. They use an adjusted net income figure for the numerator and “Average Invested Capital” for the denominator. The company defines “Average Invested Capital” as such:

“Average Invested Capital” represents the average equity capital of Constellation, and is based on the company’s estimate of the amount of money that its common shareholders had invested in Constellation. Subsequent to that estimate, each period the company has kept a running tally, adding Adjusted net income, subtracting any dividends, adding any amounts related to share issuances and making some minor adjustments, including adjustments relating to our use of certain incentive programs and the amortization of impaired intangibles. The company believes that Average Invested Capital is a useful measure as it approximates the retained earnings of the company prior to taking into consideration amortization of intangible assets, deferred income taxes, and certain other non-cash expenses (income) incurred or recognized by the company from time to time.

I disagree with this calculation of invested capital for calculating ROIC. This definition only accounts for the equity invested in the business, which is more akin to return on equity (“ROE”). ROIC should also include funding through debt financing. This got me thinking more about ROIC and the ways that it can and should be calculated. I calculate it two ways – the first is an asset-based approach and the second is a funding-based approach.

For the asset-based approach, I focus on what assets are necessary to operate the business, including a portion of cash and investments, accounts receivable, inventory and property, plant and equipment. Things like deferred tax assets, goodwill, intangibles and unbilled revenue aren’t included because they are not necessary to operate the business. In this case, it does not matter how the assets are funded, whether that be through accounts payable, debt, or equity. This answers the question – what kind of return can I get for a given amount of assets necessary to operate the business?

The second, funding-based approach focuses on the debt and equity financing necessary to operate a business and doesn’t include things like accounts payable and deferred tax liabilities. I look at both debt and equity in the calculation of ROIC because only using equity financing can be misleading as to the relative merits of a company and industry. If you finance a business mostly with debt, it may have a high return on equity but carry with it significant balance sheet risk. Some companies even go into negative equity territory, funding the deficit with additional debt. AutoZone (AZO) and Moody’s (MCO) are two good examples of profitable companies that run equity deficits, although Moody’s has recently moved back into positive equity territory.

So which method of calculating ROIC is best? The answer is that it depends, as neither the funding-based nor the asset-based approach is correct for all companies or industries. The funding-based calculation is good because it tells you how much capital a company needs from debt and equity to operate its business, it’s a straightforward calculation and it puts companies with different capital structures on an even footing. In fact, I prefer the funding-based approach for calculating invested capital, but it can be distorted by the decisions that companies make. That is why I sometimes opt for using the asset-based approach.

There are specific reasons why I choose to use the asset-based calculation, with one primary reason – company’s that make a lot of acquisitions tend to have distorted ROICs when taking a funding-based approach because purchases are typically made at prices above carrying costs on the acquired company’s balance sheet. The asset-based approach disregards those intangibles when calculating invested capital. But why would you want to disregard those assets?

Goodwill and intangibles tell you a lot more about the capital allocation decisions of the managers than the economics of the business purchased. All else being equal, a business that can earn $1 for every $5 of capital invested is better than a business that can earn $1 for every $10 of capital invested. The nuance here is when a company acquires another company for $10 and gets $1 of earnings and $5 of invested capital. The underlying business did not change, but the acquiring company paid a premium for the assets and earnings. That premium will be reflected as goodwill and intangible assets on the acquiring company’s balance sheet and has to be financed with debt and/or equity by the acquiring company. To understand the underlying economics of the acquired business and the return on capital that it earns, the intangible assets must be excluded. If you’re asking yourself a different question – how is management’s capital allocation track record – then goodwill and intangible assets should be scrutinized.

Constellation Software is a good example of a company that has made a lot of acquisitions, with half of their assets – approximately $1.55 billion as of December 31, 2018 – included on the balance sheet as intangible assets. The asset-based approach to calculating ROIC tells a lot more about the economics of the businesses they own, providing an indication of competitive advantages in those businesses. The funding-based approach, on the other hand, will mask the underlying economics of the business because the price paid in an acquisition doesn’t necessarily reflect the returns earned in that business. 

Let’s take a look at the ROIC of various companies and see how the asset-based and funding-based ROICs vary dependent on how the businesses operate.

Company Asset-Based ROIC Funding-Based ROIC
Constellation Software (CSU) 39.2% 30.4%
O’Reilly Automotive (ORLY) 19.2% 34.8%
Diamond Hill (DHIL) 165.5% 43.5%
Roper Technologies (ROP) 56.2% 7.6%

Constellation Software’s invested capital is best looked at from an asset-based perspective because the capital allocation decisions of the managers will distort the underlying economics of the businesses the company has purchased. As I mentioned before, I disagree with Constellation’s method for calculating ROIC in its annual report. It turns out, however, that my calculation results in a higher ROIC. Regardless, Constellation’s asset-based ROIC is higher than its funding-based ROIC because of the intangible assets reported on its balance sheet.

Roper Technologies, a similar company to Constellation with an acquisition-based business model, has a very high asset-based ROIC but very low funding-based ROIC. Roper’s underlying businesses have really strong economics, but the company has paid up to get those returns as indicated by the difference between the asset-based ROIC and the funding-based ROIC. 

O’Reilly Automotive is a different story. As I previously mentioned, I would prefer the funding-based ROIC calculation – that’s exactly what I do with O’Reilly. It should be noted that O’Reilly has made some acquisitions, but it’s not a major part of its strategy and intangibles are a much smaller portion – though not insignificant – part of its balance sheet. O’Reilly also benefits from extended terms for its accounts payable, with accounts payable more than funding their inventory and providing them with a competitive advantage. Taking an asset-based approach to calculating ROIC for O’Reilly would miss this unique advantage that it has.

I calculate Diamond Hill’s ROIC, which is quite high, based on assets with some necessary adjustments. This is because Diamond Hill has a lot of cash and investments on its balance sheet that are not necessary for it to run its core business. I include some cash as part of its invested capital along with accounts receivable and fixed assets. The funding-based ROIC, while still high, does not reflect the underlying economics of the business because of the idle cash and investments on the balance sheet. Big tech companies like Alphabet (GOOG/GOOGL), Facebook (FB), and Apple (AAPL) all have massive cash piles on their balance sheets, most of which isn’t necessary for their business operations. Calculating ROIC for them requires a similar adjustment.

Bottom Line

The bottom line is that ROIC can and should be calculated in a different way based on the specific company and industry you’re looking at. For ROIC’s to be comparable within an industry, you should typically look at them the same way. There are, however, circumstances that would warrant calculating ROIC for two different companies in the same industry a different way. One company may be highly acquisitive, while the other focuses on organic growth. Either way, I would not calculate ROIC as Constellation Software does, which is good reminder to always understand how a company calculates metrics that are “non-GAAP.” 

Thanks for reading and please reach out if you have any comments or questions.

Thoughts on Writing and a Commitment

Writing helps clarify my thoughts – I know this. Yet, I rarely write with the intent of others reading it, leaving the things I put on paper a jumbled mess. You can see that my posts on this blog have been few and far between. For a number of reasons, I’m committing to myself to write on this blog at least once per week.

The first reason is that I want to improve my writing. It’s always been a struggle for me. Growing up, I was told that excelling at math was good enough; that I could be a math nerd who sat in an office all by myself and didn’t have to communicate much. I earned my undergrad degree in civil engineering, well on my way to my secluded office where I could wear out the buttons on my calculator.

That couldn’t have been further from reality. Both verbal and written communication have been a critical component to my career and something that I continue to struggle with. Improving my writing connects well with the second reason I am committing to writing regularly on this blog – clarifying my thoughts.

I spend a lot of time researching businesses with the end result being a long list of notes and a spreadsheet full of numbers. I understand why I make an investment decision based on that information, yet it’s difficult to clearly and concisely explain my theses to clients and other investors.

Finally, I’d like to communicate with other like-minded investors and provide potential clients with a body of work so that they can better understand my investment process. A consistent, solid research process communicated publicly, followed by meaningful investments in my best ideas, will achieve both of those goals.

That’s it. I’ve got three main goals in mind with my commitment to consistently post on this blog:

  1. Improve my writing.
  2. Clarify my thoughts.
  3. Communicate with other investors and clients.

If you find what I’m writing interesting, please feel free to reach out. Even if you dislike it, let me know why. And, if you live in the Pittsburgh area, I’d be glad to meet in person, especially with others who are interested in value investing.

Bare with me as I hone my writing skills. I’ve re-read some of the letters that I distributed a few years back and I can without a doubt say that there’s a marked improvement today. Having said that, I still have a long way to go and I’m looking forward to the journey.