I really like this two-minute video from the 2021 Berkshire Annual Meeting. It’s Buffett’s response to a question from the audience about tech companies and return on invested capital. He covers a lot of conceptual ground in only a few minutes, giving a whirlwind tour of how to think about ROIC.
To make this a little easier to follow and ensure no wisdom is missed, I’ve provided a transcript of the question below, along with some of my comments for each part of the answer (in brackets).
But first, here are the main ideas:
- The ROIC Denominator: Companies that can achieve profits and growth with fewer assets (ROIC denominator) than companies with the same profit levels are better, because they likely require less investment prospectively to achieve profits and growth, which means more value left over for shareholders. So a lower denominator (invested capital) in the ROIC ratio is always better assuming that it reflects adequate investment in growth. This is generally well understood and these kinds of businesses are often described as “capital light.” Still, there is analytical subtlety required here as one tries to discern if there is the optimal level of growth investment to support the numerator in ROIC over many years into the future.
- The ROIC Numerator: Of course, sales growth and high margins are good for NOPAT (ROIC numerator) and the better they are, the higher the ROIC will be. But what about the durability and reliability of that NOPAT? Here investors should be thinking about recurring revenue (higher is better), sustainable competitive advantage, and anything that impacts the overall “riskiness” and growth of the future NOPAT stream. Remember that the ROIC ratio is only a snapshot of the business’s recent performance, and the less durable the income stream is, the less reliable the snapshot is. I think about it like a discount rate – the more reliable the future income is, the lower the discount rate for the business, all else being equal. I think making this judgment is one of the main places where stock pickers still have an edge over quants, because the quants are using historical data and relationships, while the stock picker is making judgments about what does/doesn’t change prospectively.
- The Price Paid For ROIC Matters: Historically strong ROIC businesses are widely recognized, which makes them trade at higher valuations. Investing success isn’t just about identifying high ROIC. It’s about getting high (future) capital productivity for an attractive valuation. Sometimes this makes lower (but highly reliable) ROIC at the right price a great investment; Buffett discusses a utility owned by Berkshire Hathaway as an example of this in his answer below.
The Question
“You have mentioned that the best investment results come from the companies that require minimal assets to conduct high margin businesses.
In today’s world many of these companies tend to be software-driven businesses. Berkshire has not invested in high-growth technology businesses in the past but this appears to be slowly changing with investments in companies like Apple and Snowflake.
As shareholders should we expect that high-margin businesses will begin to constitute a larger portion of Berkshire’s investment portfolio over time?”
The Answer
“We’ve always known that the great business is the one that takes very little capital and grows a lot.“
[Consider the simple math of the ROIC ratio: if the denominator (invested capital) doesn’t increase much over time, but the numerator (NOPAT) does, you will have increasing ROIC. These businesses make great returns on their assets and then reinvest those returns at similarly high rates – that’s what makes them compound capital. This is what Buffett is talking about here. What kinds of businesses are examples of these? How about a franchisor: in theory, it already has tremendous brand recognition and invests to maintain and grow it. It’s able to sell rights to franchisees to operate stores that use the highly recognized name, and the revenue from these fees grows NOPAT while Invested Capital grows very little. Another example is a business with lots of negative operating working capital. That’s usually the result of accounts payable that greatly exceed accounts receivable and inventory. This happens for businesses that collect revenue immediately via credit card (so no accounts receivable) and have very low or no inventory. If the business also does not require lots of tangible capital investment, the negative operating working capital is essentially financing all or most of the business, meaning negligible invested capital for the owners. A low ROIC ratio denominator sure helps the ROIC.]
“And Apple and Google and Microsoft and Facebook are terrific examples of that.“
[They are. They can sell infinite units of their product/service with low/no marginal cost, and benefit from hefty network effects that give them pricing power. Their products are also very sticky making the revenue highly recurring. This results in high gross and operating margins that improve over time as revenue scales in gigantic and ever-expanding TAMs- all with minimal levels of investment in tangible capital. Increasing numerator, modestly increasing denominator – this means ROIC grows or at least stays steadily high. Of course, this is the beauty of almost any successful software business.]
“Apple has $37 billion in PP&E Berkshire has $170 billion or something like that and they are gonna make a lot more money than we do it’s a much better business than we have. And MS business is way better than what we have Google is a way better business…“
[ROIC ratio math. Buffett is saying Apple has higher NOPAT (numerator) and lower invested capital (denominator, via the PP&E component of invested capital) than Berkshire, thus superior ROIC.]
“So we have known that a long time…we found that out with Sees Candy in 1972. Sees doesn’t require that much capital. It has obviously a couple of plants they call them kitchens- it doesn’t have big inventories, except seasonally for a short period, it doesn’t have big receivables..so those are the kinds of business….“
[There is no shortage of content on the internet about See’s Candy. For years Buffett has used it as a colorful example of a high ROIC or “franchise” business that Berkshire acquired early in its existence. The business benefits from a strong brand that creates repeat purchases from loyal customers and that gives it the ability to raise prices with little tangible or working capital investment as sales grow over time. Buffett gives a great explanation of this as an example of economic goodwill in Berkshire’s 1983 Letter. I run through economic goodwill and the letter in this post.]. Here’s a link to the Berkshire letter.
“They are the best businesses but they command the best prices…and there aren’t that many of them…We are looking for them all the time. We have got a few that are pretty darn good but we don’t have anything as big as the big guys. That’s what everyone is looking for thats what capitalism is about, people getting a return on capital and the way you get it is having something that doesn’t take too much capital I mean if you have to really put out tons of capital…“
[Strong returns come from owning high ROIC companies at purchase prices that are too low for the growth you are getting via their reinvestment of profits. Put another way: High ROIC is not enough, you have to get the price right too. Buffett also alludes to a challenge that is unique to Berkshire: its size. Anything Berkshire buys has to be big enough to move the needle on Berkshire’s performance. There are even fewer high ROIC, attractively priced businesses that are large enough to matter for Berkshire’s performance.]
“…and they don’t always stay that way.“
[I think this might be the most important aspect of Buffett’s answer to the question, yet he says it in passing, very quickly. Many people can do the ROIC math, but the question is, how dependable is the historical ROIC in giving us a sense of the future performance of the business? For investors, it is the future cash flows that will dictate our returns. So, how do we think about the future of the sector? The reliability of management? Regulatory and legal risks? Commodity price changes? The competitive environment? Macro? These are all things that will impact prospective ROIC, which is what will drive company performance and ultimately dictate the investor’s returns. Simply put, high ROIC is a useful starting point, but investing success is ultimately driven by how successfully one uses that initial read to bet on the future.]
“Look at the utility business. It’s not a super high return business, you just have to put out a lot of capital – you get a return on that capital but you don’t get fabulous returns…you don’t get Google-like returns or anything remotely close to it. We are proposing a return in the transaction with Texas I think it’s 9.3%……but yeah you look at the return for most American businesses, it’s a lot higher than 9.3m but they aren’t utility businesses either…“
[I am not an expert on Berkshire’s utility businesses, but I love Buffett’s point here, which is: sometimes, even if a business has middling ROIC, if it offers a highly dependable income stream that grows (even a little), it can still be an attractive investment if bought at a good price. Utility businesses have often made massive investments in generation and distribution (high denominator) that are extremely difficult for competitors to replicate, so they virtually own the customers in their communities, and those communities usually grow over time with U.S. population growth. Utilities can also increase rates via regulators to maintain statutory levels of returns on capital over the long term (this is what Buffett is referring to with the 9.3%), so there is very little risk to a growing income stream from these businesses. As Buffett says, there just aren’t many other businesses that offer this level of certainty (ie. Apple and Microsoft “aren’t utility businesses either” – they don’t benefit from this customer and income growth lock-in). So the ROIC denominator is high, and the numerator gives the business an OK ROIC. But the numerator grows very dependably. I think Buffett is essentially comparing investments like this to U.S. Treasuries – a reasonably good yield given virtually no risk, plus the yield grows (which he doesn’t get with Treasuries). If the risk (discount rate?) is very low and the price paid implies a good risk-adjusted return, a capital-intensive, lower ROIC business could still make for a good investment.]
Buffett bought Apple not because he suddenly became interested in tech companies, but because he felt he understood the company and its prospects well enough to have a view on the future growth of the business (i.e., ROIC) and the price seemed favorable given that expectation. ROIC helped frame his thinking, but far more than that ratio drove the investment decision. As Buffett says about the concept of margin of safety, ROIC is easy to understand, but using it to successful make strong risk-adjusted returns is hard to do.
Here are some examples of companies I have recently looked at where aspects of these ROIC numerator/denominator considerations came into play:
Pluxee is an employee benefits company in Europe, recently spun off from Sodexo. Its main product is essentially a rewards card for employees that allows them to purchase meals, pay for work travel, and access other benefits in its merchant network that are paid for by the employer. Pluxee has negative working capital because it collects revenue from corporate sponsors in advance of the value being spent by employees, requires little tangible capital investment (negative ROIC denominator!), and has grown nicely with solid margins historically. But there are some bear arguments about what might put the NOPAT (ROIC numerator) at risk, including government legislation and threats from traditional payment products. So, while historical ROIC is high, it is by no means guaranteed for the future, and this debate is what frames the company’s valuation.
Atkore is an electrical conduit (PVC) manufacturer that came out of Tyco in 2012. Atkore has had a great run the last few years, with solid revenue growth and very high ROIC, which is largely the result of tremendous margin expansion:
But the bears will tell you the margins are due to high PVC pricing, and its not sustainable for what is essentially a commodity product. So again, high ROIC, but what about the future? That’s the main debate that dictates the current stock price.
If you need to brush up on the math and some of the thinking around ROIC, go to this post, and this one too.