Terry Smith And The Big ROCE Fallacy
Why high return on capital isn't the key to high investment performance.
It was a sunny Tuesday last week in Milan, which was refreshing after several dishwater-gray, rainy days. (It’s now rainy again..)
When the sun is up and the sky is open, I always feel more intellectually stimulated and thus a bit more contrarian. It’s probably a side effect of having lived in LA for years. My brain is wired to work when it’s sunny and hot outside.
I was browsing on my Twitter with a brain hungry for some deeper-level thinking, and I saw this post by Thomas Chua:
“Fundsmith is on track for its 5th year of underperformance.”
Terry Smith, one of the most talked-about and popular investors in the world, has been underperforming for 5 years. I didn’t know that!
Well, I normally think it’s normal and even necessary for good investors to underperform in some periods, especially at times of rotations and portfolio reconstruction.
The reason is simple.
Let’s return to the basics for a minute. How would you describe the recipe for success in investing? Buffett has reduced it to three ingredients:
Buy exceptional companies.
Buy at or below fair value.
Hold forever.
The problem is that exceptional companies are rare, and they are generally well known. This is why they are often overpriced. If you manage to buy them at or below fair value, this means that they were in a downtrend for some reason. And a trend, by its nature, tends to continue in the short term.
So, it’s natural that a good investor can underperform after big purchases, rotations, and reconstructions until his exceptional stocks break the downtrend that allowed him to purchase them cheaply.
However, if you are underperforming for 5 years in the middle of one of the greatest stock market rallies of all time, I think there is a problem.
I thought about this and noticed that I have never liked Terry Smith’s portfolios. I have looked at his portfolio many times in the past 10 years, and there wasn’t a single instance I was impressed. None.
I remembered watching his presentations, listening to how he praised his positions for having high Return on Capital (ROC); but when I looked at those positions one by one, I could never find a single interesting issue that I would want to own.
Then I started to think he was putting too much emphasis on ROC as something that signals future growth, and very little on intangible analysis of growth opportunities and qualities to exploit them.
His investment thesis was falling into the ROC fallacy!
ROC is basically like a mini-skirt. It shows a lot of things, but it doesn’t show what you really want to see. It’s nice to observe it, but you’ll likely get lost at some point if you make ROC your north star metric.
Why? Let’s discuss.
Terry Smith’s Way: ROCE Is The North Star
The formula for investing success has been clear, as I have said above. Warren Buffett and Charlie Munger have left an amazing canon of text and speech that explains it.
Their personal success aside, there is also a whole cult of investors implementing their formula and achieving nice results. These are the likes of Bill Ackman, Dave Kantaseria, Chris Hohn, etc.
Their formula stood the test of time and practice, which proves its competence.
Terry Smith is perhaps one of the few investors following the Buffett formula with so much religiousness. His fund has Buffett’s recipe as its motto—buy good companies, don’t overpay, do nothing.
He puts this everywhere, the entrance of his office, every presentation material, etc.
The interesting thing about this formula is that two of the elements don’t require much deliberation. They are pretty straightforward—valuation and patience.
Doing nothing is a simple directive, though it’s hard to implement. It was basically meant to remind us that it takes time for businesses to perform and compound.
Avoiding overpaying is also pretty simple. Valuation can be challenging to perfect as a craft, but once you get it, you’ll be able to say what’s overvalued and what’s not with fair accuracy (that’s the best you can hope for).
The third point, however, is as complex as hell.
There is no one-size-fits-all recipe for a good business; they come in all shapes and sizes, so it’s always been the hardest thing for investors to decide on.
I would say there are way fewer investors who failed because they got the valuation wrong but the company right than those who failed because they got the company wrong. If you get the company right, your chance of success at least quadruples.
If you get the company right, valuation will be a way smaller problem; if you get the company right, it’ll give you relieving signals along the way, so it’ll be easier to wait and do nothing.
An exceptional company solves 90% of the problems in investing.
You could have paid 241x earnings for Altria (Philip-Morris) in 1973, and you would still outperform the MSCI World Index for the next 46 years. That’s something.
Naturally, it’s very hard to do; it’s a million-dollar question. As there is no recipe, many investors come with their own formulas based on experiences, logic, and assumptions.
One of the most popular criteria among investors to spot exceptional companies is looking at Return on Invested Capital (ROIC), or Return on Capital Employed (ROCE).
There are slight differences between the two:
ROIC is after-tax profit divided by operating assets, i.e., total debt+equity-cash.
ROCE is operating profit divided by all capital employed, i.e., equity+debt-current liabilities.
As a result, ROIC is more indicative of the returns generated by operations, while ROCE is more indicative of the capital allocation performance of the whole business.
However, this distinction isn’t of much consequence for us. Just take the perspective that many investors associate high returns generated on investment with higher business quality. The choice between ROIC and ROCE depends on investor preferences, company type, industry, etc.
Terry Smith looks at ROCE and sees it as the north star metric:
It makes perfect sense on the surface, after all, what’s an exceptional company?
When we say exceptional company, I think we mean a company that grows consistently over time, thus generating massive shareholder value.
If a company generates high returns on capital, it’s well-positioned to grow consistently over time. The surface-level implication is very straightforward:
Let’s assume a business with:
$200 million employed capital
Made $50 million this year (25% after tax return on capital)
Retains 50% of it, and the employed capital reaches $225 million
If it can reinvest with 25% again, it’ll make $56 million next year
If it can sustain this, it’ll keep growing.
Now, there is also a hidden implication.
A company sustaining high return on capital with high growth implies that it has a durable competitive advantage.
It’s economics 101: If there is an excess return in the market, competitors will enter, and prices will decline until marginal cost is equal to marginal revenue.
If the company is able to consistently generate high returns on capital with high growth, it means that competitors aren’t able to make an effective entry or capture market share from the company.
If this is for a few years, it may be due to competitors’ inability; if it’s the case for more than a few years, it indicates the company likely has a moat, which is what makes a company exceptional.
Because of this hidden implication of a moat, return on capital is the North Star metric for quality-focused investors. This is why Terry Smith has always taken pride in the high average ROCE of his portfolio:
This is a compelling strategy, and Terry Smith has been very effective in communicating this strategy. However, I can’t say it has worked very well.
Terry Smith’s fund started in November 2010. He picks the MSCI World Index for his performance benchmark.
The fund has done well in terms of outperforming its benchmark. It returned 614% against 474% of the benchmark since inception, or 14% against 12.4%.
However, when you look at the S&P 500, it already returned 13% in the same period while Berkshire returned 14.4% despite its gargantuan size.
So, I don’t think it’s possible to say Terry Smith has done much better than the market or even than his own benchmark in 14 years. If you have all those theories of what makes an exceptional company, but you generate only 1% alpha against the market, I don’t think it’s a strategy that works well.
You could just put your money in Apple and make 33x. Any other well-known, established, high-quality company would also work. You could buy something ultra-defensive like Walmart and still make 6x your money.
Given that his performance has recently diverged in the last 5 years and he doesn’t hold anything that could massively go up more than the market, or diverge in direction in case of a correction, I think it’s likely that this is going to be a lost decade for him.
Given that this is also happening in one of the greatest bull markets we have seen, I think it means that something is broken with the strategy.
What I think is wrong is his North Star, i.e., return on capital. He fixated his eyes on it, so he was pulled in the wrong direction.
Problems With Return On Capital
The day I saw Thomas Chua’s post about Terry Smith’s performance wasn’t the first time I thought about return on capital.
As an investor, I also had a period when I was a staunch proponent of return on capital as the best possible predictor of quality and future performance. I didn’t get much better results than Terry Smith. I did serious thinking and reading, and left my obsession behind. I have been thinking more intangibly about the quality and growth opportunities of the businesses since then, and it has worked very well for me.
I think there are three main problems with return on capital.
1️⃣ It’s a backward-looking metric.
Wayne Gretzky is the leading career point scorer and assist producer in NHL history and has more assists than any other player has total career points.
He once shared his secret of success in one simple sentence: “Skate to where the puck is going, not where it’s been.”
I have never watched hockey, never been in a hockey game, and I don’t even know how many people are in a hockey team or how long games last. Nothing. I heard this quote only because it was voiced by one of the greatest investors of all time— Stanley Druckenmiller.
It has stuck with me since then. You can expect exceptional results only by looking forward, and not backward.
This is one of the problems with return on capital. It tells you where the puck is and has been, but it doesn’t tell you where it’s going.
Imagine that you are an investor who wants to invest in businesses consistently generating a high return on capital. You believe that this signals a durable competitive advantage and thus durable growth.
You won’t simply go and invest in businesses that generated high return on capital last year; you would like to see proof of consistency, you would like to see, let’s say, at least 5-10 years of high return on capital.
Well, the problem is that when you find these businesses, their runway for generating high returns on capital will already be shrunk substantially.
Michael Mauboissin’s research clearly shows this:
Companies reach their peak ROIC between growth and maturity. If you look for a consistently high ROIC, or ROCE, let’s say for 5-10 years, you’ll likely miss all those young growth companies and invest in those that are transitioning to maturity.
Once companies reach maturity, their growth stagnates and then eventually declines.
The best time to invest in a company is just before it enters the high-growth phase. This is where you can find promising companies still trading at a discount as they aren’t obvious yet. Once they are in the high-growth phase, the market quickly notices them and bumps up the multiples. They get too expensive too fast.
If you obsess over return on capital, you’ll always miss the best time to invest because it’ll look too low. In the worst-case scenario, you’ll end up investing just before they decline.
What’s even worse is that dominant and mature companies generally arrange their capital structures to minimize burn and maximize shareholder returns. Thus, they’ll keep having a high return on capital, but growth will be sluggish. This brings me to the second point.
2️⃣ High ROC doesn’t mean large capital deployment.
Some businesses intrinsically have little capital requirement.
Think about Coca-Cola 50 years ago.
If it did nothing, it would need more or less the same capital requirement to run the business a year later, after inflation adjustments. And 50 years ago, it was still not possible to find Coca-Cola in most of the emerging economies.
Thus, when Coca-Cola spent a little more money to enter a new country, it would generate very high returns on that incremental capital deployed. The more capital it deployed, the more countries it would enter, and the faster it would grow.
However, as the markets got saturated, returns from incremental capital have diminished.
Coca-Cola is now accessible everywhere in the world. How can Coca-Cola deploy $50 billion next year to the soda business and expect high returns? It won’t happen.
If it ventures out of its domain, return on capital will also be substantially lower as it’ll know nothing about the new business, new competition, new market, etc.
Thus, mature and dominant businesses cut their capital deployment as much as possible, since additional capital deployment won’t bring much benefit. As a result, they tend to have high return on capital, but the growth remains sluggish.
Waters Corporation is a magnificent example.
It’s one of Terry Smith’s biggest positions with 6% portfolio share. It has a 10-year median ROCE of 22%, way above average. However, its revenues grew by only 4% annually since 2015.
This happened because it hasn’t deployed incremental capital. Its capital deployed was $3.7 billion in 2015, and it’s still pretty much the same.
Though the stock still performed well and more than tripled, as earnings grew faster thanks to efficiencies and buybacks. But… The market has also tripled. So, no alpha was generated.
Let’s compare this to Amazon.
Amazon’s total deployed capital increased from $31 billion in 2015 to $522 billion this year, and revenue grew from $100 billion to $700 billion levels, while the stock made more than 12x.
Yet, if you look at Amazon’s ROCE, you wouldn’t expect it, as its 10-year median ROCE is just 10.5%. However, this is only because Amazon is still trying to grow aggressively and thus deploying large amounts of capital, which drastically reduces operating profit due to D&A and investments hidden in SG&A expenses.
Thus, what makes an exceptional quality isn’t simply high returns on capital; it’s generating high returns on capital while deploying large amounts of capital.
Buffett explained this clearly in Berkshire’s 2003 annual meeting:
Buffett accepts that some of their businesses, like See’s Candies, don’t have many opportunities to generate high returns on incremental capital.
What separates Buffett from Terry Smith is that he can take the cash out of those businesses and reinvest elsewhere to generate higher returns, thanks to Berkshire’s holding structure. Terry Smith doesn’t have such a vehicle.
3️⃣ It ignores superlinear returns.
Most businesses in the world, even most of the best ones, are just like See’s Candies.
They can generate high returns on capital in their own market, but once that market saturates, their return on incremental capital drops, leading to small capital deployment and thus sluggish growth.
From there on, fast growth requires entering new markets by deploying large capital. Return on capital naturally declines here as the business ventures out of its domain of expertise. It needs to figure things out all over again.
This necessarily requires experimentation, and most experiments naturally fail. Failures drag down the return on capital. If you look at a business at that point, you would think it’s not worth it.
However, a successful experiment pays for dozens of unsuccessful ones, establishes you in a new business, and opens up a whole new market ahead of you where you can deploy capital with returns for years.
Amazon is the goat in this.
It now has 6 businesses, each generating over $20 billion in annual revenue.
If you looked obsessively at surface-level figures, you would never see this coming.
These are the superlinear returns when you hit something big. Such cases were rarer in the old economy, mostly due to physical constraints. However, the new economy, the digital one, almost exclusively runs on superlinear returns.
Companies don’t invest in things that can bump their growth by 1-2%; they invest in things that turn into big winners, as Amazon did with Prime, FBA, and AWS.
If you looked at just the ROCE figure, you would never see them coming, and if you looked at Amazon’s ROCE for the last 10 years and nothing else, there is no chance you would say this is a $3 trillion business.
Remember the mini-skirt thing. It shows a lot, but it doesn’t show what you really want to see.
🏁 Final Words
I have been thinking about writing thoughts on return on capital since I read “What I Learned About Investing From Darwin” by Pulak Prasad earlier this year.
It’s a great book drawing many useful parallels between biological evolution and the corporate lifecycle, and Prasad is a very successful investor. His main idea in the book is that consistent return on capital is the hallmark of exceptional companies.
I agree with Prasad, consistent high return on capital implies some sort of moat, as I explained above. However, it doesn’t automatically translate to investing success because what matters is being able to deploy high amounts of capital at high returns.
Yet, Prasad did very well. Why was that?
When I looked at the companies he owned, I thought there was no way I would own them. He owned several electrical equipment manufacturers, air conditioner manufacturers, underwear manufacturers, etc.
An alarm went off in my mind—how come these companies manufacturing commodities could have a high return on capital and deploy a lot of capital to maintain growth?
That’s when it struck me. Prasad invests in India.
India is an emerging economy, and it’s one of the fastest-growing markets in the world. Many people are just getting out of poverty, fast fashion is just spreading, and many homes still don’t have AC, etc.
When there is a huge untapped market, the leading companies can just deploy larger and larger amounts of capital, produce more, and sell more to more people. It just makes sense!
So, what allowed Prasad’s companies to generate high returns on capital while deploying larger amounts of capital was the immense untapped market, not their ability to create new markets to deploy massive capital like Amazon, Microsoft, and Google have done.
This is why Prasad’s using high return on capital as the North Star worked.
Try investing in electronic equipment manufacturers in the US, and it won’t work. You’ll instantly see two problems:
High return on capital will be very rare.
Even when return on capital is high, capital deployment will be low.
So, investing in those companies won’t work.
In short, high return on capital implies a moat, but it won’t be important without the ability to deploy high amounts of capital. Even if the company has a high return on capital and deploys large amounts of capital, it doesn’t automatically imply Amazon-like exceptionality. It may be well because of the big untapped demand.
Truly exceptional companies that can deploy large amounts of capital at high returns are very rare, as Buffett says. They aren’t one in a hundred; they are one in a thousand.
The natural implication of this is that if you are investing in developed markets, like Terry Smith, the companies you are attracted to because of high return on capital won’t have many opportunities left to deploy large amounts of capital. As a result, their growth will be stagnant.
If you obsess over high return on capital, you’ll miss most of the opportunity, as return on capital isn’t that high when the companies are in their fast growth phase.
Thus, there is no shortcut or a North Star metric that can replace fundamental understanding and decision quality.
If you looked at return on capital for SoFi, RobinHood, Nebius, and Lemonade, there is no way you would invest in them. Only a strong fundamental understanding of those companies would urge you to invest.
So, my message is that investing is hard. Don’t believe those who are oversimplifying it. Many great investors try to reduce their processes into formulas and provide guidelines to educate others. They do it in good faith, but those processes involve nuances that aren’t apparent to third parties right away.
Return on capital is one of the most popular guidelines, but it has many nuances. I know no one who amassed a massive amount of wealth from the ground just by investing in high-return on capital companies.
The brutal truth is that decision-making quality is a real thing.
What you buy, at what price, and what size cannot be reduced to a process. If it were, all the business school graduates would have similar track records.
The good news is that you can improve it by deliberately trying to understand what went wrong and what went right in your decisions. But unfortunately, you can’t reduce it to a formula.
High return on capital is good, but it’s not the North Star many believe it to be.
I’ve read your article and I think it’s quite good, but I have a different opinion. It’s clear that Terry Smith has been underperforming for five years, but I don’t think the main reasons for this underperformance are the ones you mention. Part of it has to do with his portfolio composition. The rally of the last few years hasn’t come from the outperformance of all companies, but from just a few. The S&P equal weight, or the index without the major tech names, hasn’t done particularly well. A large portion of Fundsmith’s portfolio is made up of consumer staples and healthcare, two sectors that have been heavily hit in recent years. And still, his Sortino ratio is significantly better than the market average. This is very relevant, because as Warren Buffett said in his early letters, the key to his success wasn’t so much outperforming every year, but rather falling less than the rest in bad times, and Terry Smith continues to deliver on that.
Moreover, if AI really is a financial bubble, it’s possible that this underperformance could completely reverse.
Another element to consider is that the factor he invests in the quality factor, has underperformed the market, but that relates to what I said earlier. This factor typically does worse in bull markets but much better in bear markets. Also, if we compare him with the most similar class of funds (Seilern, for example), Terry Smith hasn’t done worse, quite the opposite.
There’s another part of your argument that I don’t buy: the idea of putting all your money into Apple and having made a 33x return in that time. Doing that would have meant taking on much more risk than investing in a portfolio of 20+ companies across different sectors and revenue sources. Berkshire is more comparable because it can be considered a pseudo–investment fund, but arguments similar to the ones made about Terry Smith were also made about Warren Buffett for many years just look at what people were saying about him in the years leading up to the dot-com bubble.
I do agree, though, on the importance of understanding the future, the ROIIC, and the reinvestment rate. That said, I believe he can continue delivering double-digit returns over the long term with very low volatility and low fees (although if someone prefers to invest in an MSCI Quality ETF, it might well be a better alternative, to be honest). Many of the companies he owns are over a century old, still have room to grow, and remain leaders in their current markets. Their slower growth compared to other companies is due to supply-chain issues and affordability crises. They have low debt, lots of cash, and strong returns characteristics that help them fall less during difficult economic periods.
P.S.: Another reason for this year’s large underperformance is the undervaluation of the dollar. Terry Smith doesn’t hedge currency risk, which makes sense for a long-term investor. Although yes, he also would have underperformed even without the dollar depreciation. Anyway, great article, lots of interesting ideas, and I’m looking forward to reading your thoughts!
I completely agree—high returns on capital mean very little if a company has no opportunities to reinvest that capital. And, of course, it is extremely favourable when a business can reinvest substantially more. Still, I believe the Terry Smith investing approach works over the long run. As someone mentioned in the comment section, AI drove a large share of market gains over the last five years, so using that period to claim the Terry Smith approach doesn’t work isn’t a strong argument. If this underperformance persists through a market downturn or another five-year stretch, then perhaps criticism will be justified—but it’s too early to draw that conclusion now.
I continue to believe investors should look for companies that earn high returns on capital and have decent reinvestment opportunities. Consider two companies: one earns a 10% return on capital, and the other earns 20%. Both maintain these returns going forward. Company A can reinvest 50% of its earnings; Company B can reinvest 25%. The compounding rate for both ends up roughly the same: 5%. However, Company B—with fewer reinvestment opportunities—returns 75% of its earnings to shareholders, while Company A returns only 50%. And this is before considering scenarios where returns on capital rise well beyond 20%.
Of course, this framework doesn’t always hold. Ideally, you would have invested in an Amazon—lower returns on capital early on, but the ability to reinvest nearly 100% of earnings back into the business. The challenge is that analysing such companies is much harder and often riskier, because reinvestment frequently requires expansion far beyond the original business model. Businesses able to reinvest at high rates within their core operations tend to be smaller, faster-growing firms. Identifying those future winners is far more difficult and, in most cases, riskier. With the Terry Smith approach, you are not trying to find the next winners—you are buying the companies that have already won.