The impact of Return on Capital on Shareholder Returns

Last quarter, there were 484 mentions of ROCE in the filings of just the 50 Nifty companies.

ROCE stands for "Return on Capital Employed.": how well a company is generating profits from its capital. It is considered a critical part of the puzzle of finding quality businesses.

ROCE typically helps differentiate between ordinary and high-quality businesses.

Imagine we're looking at two restaurants in the same town, Restaurant A and Restaurant B. Both offer delicious food and have a steady stream of customers. How would you identify the better prospect to invest in?

Restaurant A is a swanky place that spent a lot of money on renovations, high-end kitchen equipment, and a renowned chef. Their food is excellent, and they charge high prices. However, due to the high overhead costs, their net income isn't as high as you might expect from looking at the price of the dishes on their menu. Their ROCE, which measures how much profit they generate for each dollar of capital they've employed, might not be very high. Despite the high prices they charge, they're not turning their capital into profit as efficiently as they could.

Restaurant B is a more modest establishment. They've kept their costs down by leasing cheaper equipment and hiring a less well-known (but still competent) chef. They charge lower prices, but they serve more customers and their net income is surprisingly high. Their ROCE might be quite good, because they're generating a lot of profit for each dollar of capital they've employed. They're using their capital more efficiently.

If both restaurants are profitable and have a loyal customer base, you might think they're equally good businesses. But as an investor, you'd probably prefer Restaurant B. Why? Because a high ROCE indicates that the business is better at generating profit from its capital. It's a more efficient business model, and that's a sign of a high-quality business.

The link between ROC and Returns over the years

We did a longitudinal analysis spanning ten years centred on classifying companies on the continuum from high to low ROC and looking at their shareholder returns. The objective was to see what patterns emerge that might be considered generally acceptable to the way the stock market rewards shareholders.

We started with over 800 companies that were operational 10 years ago. We classified this cohort into 10 deciles (10 - best, 1 - worst) based on their 3-year average ROCE at that point. By taking a three-year average, we improve our chances of getting a better representation of each company's financial reality than any one year. In addition, we also track their three-year average ROC in year T-7, T-4, and T-1, where T is current year.

Then we look at their total returns over the 10-year period to check for any obvious patterns. Whether the high ROC companies 10 years ago continue to be high ROC companies today. And to what extent did high ROC translate to shareholder return?

Here's what we found.

Return on Capital is only mildly sticky

When the 812 companies were split into deciles from 1 (worst) to 10 (best), each decile had about 81 companies. The table shows how those companies break down as of one year ago on the same metric.

Over the last 10 years, those companies either improved their relative decile, got worse, or stayed in the same decile.

Figure 1 below shows that split. The numbers in the table indicate where the companies ended up.

The table should be read row-wise: For example, Top row: Of the 82 companies in the top (10th) decile ten years ago, 34 continue to be in the top decile as of 1 year ago, the remaining 48 slipped to lower deciles - 11 to the 9th decile, 9 to 8th decile, 6 to the 7th decile and so on.

Bottom row (above the total): Of the 80 companies in the bottom decile ten years ago, 8 are now in the top decile, 2 in the 9th decile, 4 in the 8th decile and so on.

Probably an easier way but with less information is to visualise decile movement from 10 years ago to 1 year ago by percentage of companies.

4 out of 10 companies that were top decile 10 years ago were still top decile 1 year ago. This means 6 out of 10 companies from 10 years ago are no longer top decile. Of course, decile 10 companies had nowhere to improve to.
1 in 4 top-decile ROCE companies a decade ago are now in the bottom half of their cohort of 800+ companies
The highest proportion of companies showing improvement were in deciles 1, 2 and 3. Put another way, the companies at the bottom of the pile showed the most improvement
4 in 10 companies across the whole set worsened on their ROCE decile, while the same number also improved on the same metric
Not apparent on this chart: 45% of companies are within +1 or -1 decile of their starting ROC decile a decade ago. The remaining 55% saw larger moves in their relative Return on Capital.

The 34 top-decile companies from a decade ago are still in the top ROC decile

The 34 companies that continue to be top ROC decile

Castrol India ranked #3 on ROC 10 years ago and continues to be in the 10% on the same metric. But look at its Total Return to shareholders over that period. Glaxo Smithkline improved its Return on Capital from 10 years ago and shows a cumulative 38% return. This begs the question how effective is a high relative return on capital at predicting future returns.

A glorious past doesn't necessarily mean glorious shareholder returns

Chart shows median annualised 10-year return from the ten deciles, each with 80 to 82 companies.

Chart shows the aggregate ten-year return from the companies in the ten deciles. Remember the ROCE considered was the 3-year arithmetic mean at the time.

The right-most column shows the decile-wise 10-year return of the top-to-bottom ROC deciles as of 10 years ago. The bottom decile (decile 1) has the worst 10-year return. The median return from the 7th decile turns out to be the best followed by the 6th decile.
The top ROCE decile of stocks delivered median 431% cumulative total return over ten years. That's just over 18% annualised
The highest returns seem to have come from the 7th ROCE decile 10 years ago that were in the top decile recently. Referring back to Fig 1 above shows that's a set of 8 companies. The heatmap shows how the best-returning companies are the ones in the top deciles recently.
Also, the median stock went up 395%, almost 4x over ten years. That is 17.4% annualised. Keep in mind, the cohort of 800+ companies includes many small and micro caps at the time which explains the significantly higher return than the Nifty over the same period.

This makes you wonder, how much did consistency of 3yr Return on Capital affect total return?

Because Returns on Capital fluctuate absolutely, not just relatively

The first table considered how much companies shift on relative scale of comparison. This one looks at how consistent companies are on return on capital with their own past.

We measured consistency as the Standard Deviation of ROCE across 10Y, 7Y, 4Y and 1Y ago divided by the Average ROCE over that time frame. This measure is called the Coefficient of variation, which looks to only measure how much a metric changed over time. Lower the CV, the more consistent the company was on its 3-year average ROC over 10 years.

The top decile companies on consistency had the lowest CV (most consistent) while the bottom decile had the highest variation (least consistent).

The bottom ROC-decile companies 10-years ago were the least consistent since then, which is a probably a good thing for them
The more consistent companies are in the top few ROC deciles (top left portion of the heatmap), which makes you wonder if they also delivered significant return

Chart shows consistency decile-wise returns

Chart below maps ROC consistency with starting ROC decile on aggregate 10-year return.

The more consistent the return on capital, the better the long-term returns. The bottom row shows overall returns by consistency; there's a significant drop as you move from left to right.
The highest returns came from companies in the top-left quadrant i.e. decent and consistent ROC.
15 companies had with highest ROC 10 years ago and were consistent with those healthy ROCs over the last 10 years. They delivered healthy returns but interestingly not the highest returns of their cohort.

So far, the analysis looked at relative ranks of the universe.

But, what if its the trajectory that matters more than where a company starts from?

Trajectory of ROC matters (a lot)

The columns in the table bear explanation. We had 4 ROCE metrics, T-10, T-7, T-4 and T-1. Each ROCE number is the average ROCE of the last 3 years, which means the four ROCE metrics cover 13 years of yearly ROCE.

The scoring for ROCE trajectory is simple. Starting from the ROCE 10y ago, for each subsequent measurement, the company gets +1 if its ROCE improved over the previous and -1 if it worsens. So, any company can score maximum +3 i.e. improvements in each subsequent measurement, +1: improvements in 2 measurements and a decline in one, -1: improvement in 1 measurement and declines in two measurements or -3: consistently worsening ROCE over ten years.

Note how a company at a subpar 4% starting ROCE improving to 6%, then 8%, then 11% scores +3, while a company at 30% 10 years ago declining to 28%, then 26% and then 24% scores -3. We deliberately stayed away from more elaborate methods that take the degree of change into account.

16% companies saw continued decline in ROCE over the last 10y. This includes 29 top-decile companies. Unsurprising top-decile names that have seen consistent decline : BAJAJ-AUTO, BHEL, IL&FS, SUNPHARMA, HINDCOPPER, OPTOCIRCUITS. Some surprising names: TITAN, ASIANPAINTS

1 top decile ROCE company from 10y ago continued to improve ROCE over the last 10y! P&G
30 companies out of 813 consistently improved ROCE from 10Y ago. Notable mentions: KPRMILLS, DEEPAKNITRITE, SAREGAMA, INDIANHOTELS, TANLA PLATFORMS

80% of companies were the in the middle. Seeing increases and decreases. 48% saw one improvement and two declines, and 33% saw two improvements and one decline.

The table now shows the median returns of companies in each category.

Overall, there is a stark difference between median returns of companies that improved ROCE versus those that did not
Highest returns as a group came from 5 companies that were middle of the pack 10y ago: Deepak Nitrite, KPR Mill, Lincoln Pharma, Mold-Tek Tech, Gufic Biosciences. The 7,000%+ return came solely from Balaji Amines.

Interestingly, from the top-decile cohort, the highest returns came from the companies that saw 1 decline over a 3y period and 2 increases. List includes Navin Fluorine, Abbott india, Coforge, JB Chems & Pharma, Havells India, Page Industries. Emami Ltd is surprising in only having gone up 94% in 10Y inspite of what seem to be healthy ROCEs
The 29 companies top-decile that saw declines in ROCE unsurprisingly did only about as well as the universe, but not too shabby. Outperformers exist even in this list: TITAN, LUMAX AUTO, SUPRAJIT ENGG, CERA SANITARYWARE

Conclusion: ROC is necessary but not sufficient

Expecting just one metric to explain returns for a broad universe of stocks over 10 years is overly simplistic. Factors like growth and valuations play critical roles. However Return on Capital is an important indicator of how commodity or non-commodity a business is.

Knowing the high-ROCE companies today doesn't necessarily tell us where they will be in the future. Common sense tells us that a business earning inordinately high return on capital will see others enter the space thus increasing competition and reducing returns for all involved.

The exceptions are those companies that have advantages (moat) that cannot be replicated. Most often we seem to conflate a temporarily high return on capital with the presence of a moat that will allow those returns to continue.

Most companies see their high returns on capital erode over time, either because of increased competition, or an advantage that was short-lived to begin with. Hence, ROC ranks tend to be fluid.

Then there are the aberrations where the ROC lens does not explain the shareholder returns or lack thereof.

Historically, only a small set of companies (Figure 3) have been able to maintain superior ROC over a decent stretch of time. But even that hasn't ensured higher shareholder return. Just look at Castrol.

On the other hand, companies with distinctly ordinary early Returns on capital have seen strong shareholder returns. The catch being they were able to consistently improve their ROC over time signalling an improving business. (Figure 11) Being able to identify businesses that are currently earning moderate but not non-existent return on their capital with a roadmap to improve over time would be the sweet spot of companies most likely to deliver excess return.

The impact of Return on Capital on Shareholder ReturnsThe impact of Return on Capital on Shareholder Returns