When it comes to corporate financial performance, we typically think in absolute terms, measuring ROA in percentage points. We are less accustomed to thinking of corporate performance in relative terms, but knowing a company’s relative performance is essential to setting and achieving performance improvement targets and, eventually, exceptional performance.
Editor’s note: This issue marks the fifth anniversary in a continuing series on superior corporate financial performance. Beginning in January of 2010 with “Survival of the fattest,” the research ultimately led to the publication in 2013 of The Three Rules: How Exceptional Companies Think. This is the twelfth installment of the project’s findings in Deloitte Review.
Our ongoing research into the drivers of superior performance has led us to conclude that exceptional profitability is a function of a corporation’s commitment to following three rules:
View related materials
Deloitte Review, Issue 16
The journey to exceptional performance (interactive video)
Join the conversation
These rules are rules because their validity does not seem to depend on circumstances. Regardless of industry, time period, or competitive context, the companies that follow these rules seem systematically more likely to realize superior long-term profitability. When seeking to create value for customers, if a company cannot be both better and cheaper, choosing its path using Rule No. 1, and opting for better before cheaper, appears to be the better bet. When increasing profitability can no longer be done by increasing revenue and cutting cost, turning to Rule No. 2 and putting revenue before cost tends to yield superior results. And when there are more worthwhile initiatives—innovation, geographic expansion, brand building, acquisitions—than there is money to fund them, use Rule No. 3 to set priorities: Since there are no other rules, choose the initiatives that best support rules No. 1 and No. 2.1
But wait … there’s a catch. However confident we are that the rules define what makes exceptional companies truly exceptional, we cannot assume that following the rules is what allows a company to become exceptional in the first place. Those companies that are not exceptional (by construction that is most companies), yet aspire to exceptional performance, must embark on a journey to this difficult-to-find destination.2
Our research suggests that ultimately, most companies that become and remain exceptional owe their results in material measure to following the rules. However, when and in what ways companies need to follow the rules to become exceptional turns out to be highly dependent on a company’s industry and the magnitude of the performance improvement required to achieve exceptional status. In other words, like every journey, how you get there depends largely on where you’re starting from.
Every journey demands that we define our starting point, our destination, and the route we wish to follow. Today, the task of navigation is almost humdrum. Thanks to accurate maps built on satellite and ground-level photography and GPS technology that places us within feet of our actual position, identifying where we are, where we want to go, and the route to follow to get there is literally at our finger tips.
Navigating the journey to exceptional performance, however, has far more in common with ocean voyages of over 300 years ago than the touch-screen-enabled confidence we all enjoy today. When Barbosa, Columbus, or Drake set sail, they surely understood the importance of fixing their location in terms of latitude and longitude, but likely had the means of measuring only the former. One’s distance north or south of the equator could be calculated based on the angle between the noonday sun and the horizon. These angles are unvarying, determined by the angle of the Earth’s axis and the sun’s seemingly seasonable wobble between the tropics of Cancer and Capricorn. Latitude seems in every practical respect an absolute measure of position.
Ascertaining one’s longitude, however, demanded calculating a relative position: one’s distance east or west of some other point. The only way to determine this reliably is simultaneously to know the time aboard ship and the time at some other place. The difference tells you how many degrees of longitude there are between the two points. With 360 degrees to account for within 24 hours, every 15 minutes’ difference between the two is one degree of longitude, every 3 minutes and 45 seconds after that, an additional minute of longitude. However, the distance accounted for by these degrees and minutes depends on one’s latitude: At the equator, one degree of longitude is just over 69 miles, and it gets shorter as one moves north or south. In other words, longitude seems very much a relative measure of position.
A large part of what made those long-ago ocean voyages so dangerous was the ability to measure only latitude—the absolute measure. The motion of the ship foiled the pendulum-based timekeeping technology of the day, which reduced navigators to relying on dead reckoning. This meant that estimates of longitude—the relative measure—were dangerously and consistently inaccurate. This cost many a sailor his life as ships routinely ran aground on shoals the ship’s captain thought were dozens of miles away, or as starvation and thirst took their toll while ships sailed desperately but vainly back and forth, unsure of over which horizon lay safe harbor.3
When it comes to corporate financial performance we typically think in absolute terms, measuring profitability, in this case return on assets (ROA), in percentage points. No matter the industry, the company, or the time, an ROA of 5 percent is immediately understandable.
We are less accustomed to thinking of corporate performance in relative terms, but knowing a company’s relative performance is essential to setting and achieving performance improvement targets and, eventually, exceptional performance.
Calculating a company’s relative performance, like calculating longitude, is not straightforward, and for at least two reasons. First, we wish to capture the performance of the company that is a function of those factors most subject to the company’s control. When it comes to assessing a company’s historical performance, we typically wish to separate out the material impact that year, industry, and company size have on profitability. Dealing with unadjusted ROA numbers confounds the quality of management with the good fortune of having operated during expansionary times, the foresight to have chosen a good industry, or the compounding effects of a stroke of good luck early in the journey.
Once we choose to understand a company’s performance contingent on these factors, we run into our second challenge: the vanishingly small set of companies against which we can benchmark another company. After all, if we insist on comparing only like with like, we can easily end up with but a handful of companies to compare with each other. Notions of 50th, 75th, and 90th percentile performances break down quickly under such conditions. In a small sample, the relative standing of a given absolute level of performance can be artificially inflated, leading to complacency, or artificially depressed, leading to exhausting and fruitless efforts to fix problems that don’t exist.
To address this problem, we used quantile regression on the Standard & Poor's Financial Services LLC's Compustat database of more than 25,000 companies between 1966 and 2013.4 This allows us to control for year, industry, and size factors and estimate conditional quantile benchmarks for any company. The figure below shows the estimated quantile cutoffs in the technology industry for 2013 for five size categories. Note the material differences between the adjusted and unadjusted benchmarks. For example, the unadjusted median profitability in 2013 was –1.0 percent, yet for no individual size band was the median ROA below 4.0 percent. Yet these estimates of benchmarks, had they been built simply by rank ordering companies within a size band, would have been equally misleading.
To see the importance of controlling for industry and size, observe in figure 2 that the same levels of absolute performance translate into dramatically different relative performance across industry, size category, and time. For example, an ROA of 2 percent means a company was better than almost 70 percent of a relevant peer group in the power and utilities segment in 2013, but in the bottom third among similarly sized technology companies in 1989.
Determining a company’s relative performance using quantile regression affords us the ability to “tell time at sea” and fix a company’s performance “longitude,” that is, its relative performance. This means that we can determine with a new degree of accuracy from where a company is beginning its journey to exceptional performance.
Few companies that make the journey to exceptional performance reach this destination in one bound. We calculated the frequency with which companies were able to transition from each decile of performance to every other decile of performance in a single year. Not surprisingly, large leaps are quite rare; perhaps more surprisingly, staying right where you are is the most likely outcome of all.
Note that even when beginning from the middle of the distribution—the 50th percentile of performance—a company has barely better than a 10 percent chance of making it into the 70th percentile or higher in the next year, and only a 3 percent chance of making it into the 90th percentile of performance. Worse still, this matrix captures only the likelihood of making it into a given percentile in the following year; it says nothing about the likelihood of staying there should one manage to cross that chasm. The implication is that most companies aspiring to exceptional performance should steel themselves for a several-years-long journey, plodding through the deciles of performance one after the other.
We can combine the benchmarks for given quantiles of performance (figure 1) with the transition probability matrix (figure 3) to create rough approximations of how targets for performance, often set in absolute terms, translate into relative targets. For example, a company in the technology sector that has turned in an ROA of –2 percent and aspires to positive profitability is, in relative terms, looking to jump from the second decile of performance to the fourth decile, a transition that, at the population level, has a probability of 9 percent. Even including the welcome possibility of overshooting this target and landing somewhere higher than the fourth decile, the cumulative probability of success is estimated at less than 25 percent.5
These probabilities do not prescribe or proscribe any particular performance target; as population-level parameters, they should be seen as general guidelines only. What they provide, however, is a perspective on current and desired performance that can often be absent from traditional goal setting. Although few experienced managers would fail to set goals based on an assessment of the likelihood of successfully meeting them, this more-nearly objective empirical assessment reveals that seemingly reasonable improvements in performance can in fact be quite unlikely. Similarly, just as few managers are likely to dismiss the claim that dramatic improvements can take time, this perspective reveals that moving from “good” to “great” can sometimes require a surprising level of patience and persistence.
There are three main levers a company can pull to help improve its profitability: the two components of return on sales (ROS)—gross margin percentage (GM) and other costs percentage (OC)—and total asset turnover (TAT). Not surprisingly, a company hoping to move from the middle of the pack (or worse) into the top echelons of profitability must, in the vast majority of cases, pull hard on all three of these levers.6
Tackling everything at once, however, can be somewhat overwhelming, and in some situations, might not even be possible. Consequently, it would help to know just what a company’s priorities should be as it embarks on its own voyage to exceptional performance.
To address this question, we looked first at how improvements in GM and OC translated into improvements in ROA.7 Specifically, we calculated the impact on ROA of a one percentage point improvement in each of GM and OC above the sector median. We call this the “efficiency factor.” The difference between the efficiency factor for GM and OC reveals at a glance which of the two is the more efficient, and hence, all else equal, a better investment. For example, a difference in efficiency factors of +0.09 means that a percentage point improvement in GM yields 9 basis points more in ROA improvement than does a percentage point reduction in OC. Since the investments required to improve GM can be very different from those required to reduce OC, this difference in efficiency factors provides some guidance in evaluating the relative attractiveness of those different investments.
We then calculated the differences in efficiency factors for companies in the 10th, 25th, 50th, 75th, and 90th percentiles of performance in each of 18 different industry sectors. This allowed us to see whether the relative importance of increasing GM or reducing OC varied with relative performance.
The patterns revealed by the color coding in figure 4 capture the underlying relationships. The general trend from red or orange to light or dark green indicates that at lower levels of performance, companies should put greater emphasis on reducing OC. As their performance improves, priorities should shift, toward improving GM. Note that this analysis speaks only to relative importance to further increases on ROA at each level of relative performance. Red does not indicate that improving GM reduces ROA, nor does green mean that decreasing OC depresses ROA. This analysis speaks only to the relative efficiency with which increasing GM or reducing OC increases ROA. In other words, this chart reveals the relative importance of these levers of profitability enhancement, not their absolute impacts.
Asset turnover must be analyzed separately because it is measured on a different scale. Figure 5 shows the underlying pattern in the impact on ROA (in percentage points) of an increase of one asset turn annually at different relative levels of performance.
Since this is not a comparison of two levers, the negligible and sometimes negative values in the lower levels of relative performance are especially intriguing. They imply that increasing asset turnover actually reduces ROA. This follows when one considers that the typical company at the 10th or 25th percentile has negative ROA. When a company with negative ROA increases its TAT, all else held constant, it is driving profitability further into the red. The curious case then becomes those companies in the 10th and 25th percentiles that are able to improve their profitability by increasing TAT, suggesting that the knock-on effects of improving TAT may serve to improve margins.
As one might expect, as companies’ profitability improves, asset turnover tends to have an increasingly and often dramatic positive impact on ROA. The implication is that an increase in asset turns should typically not be a priority for a company seeking to set itself on an even keel, but can very much come to the fore as the wind begins to fill one’s sails upon reaching the 50th or 75th percentile of profitability.
Note further that the impact of GM, OC, and TAT on improving profitability is not a function of company size. We control for size when determining companies’ relative performance and, as shown in figure 5, at each quantile of performance, higher-performing companies are either similarly sized or larger than lower-performing companies. In other words, the shift to gross margin and asset turns as the key drivers of improved profitability among higher-performing companies cannot reasonably be attributed to a preponderance of small niche players or companies dedicated to shedding assets at the high end of the profitability distribution.
When evaluating the attractiveness of investments that are focused on either increasing gross margins or decreasing other costs it may be easy to dismiss the seemingly small differences in performance afforded by the efficiency factors. To evaluate the impact of these differences we can examine the impact of the efficiency factors by normalizing the entire population of companies on the basis of both ROA and gross margin for year, sector, and size. We find that there is a 4.8 percentage point difference between the 50th and 75th percentile of ROA, and a 9.9 percentage point difference between the 50th and 75th percentile of gross margin. These differences indicate that if the average trade-off in gross margin and other costs efficiency factors between the 50th and 75th percentile were only 5 basis points, it would equate to greater than 10 percent of the performance extension a company needs to improve from the 50th percentile to the 75th percentile.
Further examination of ROA, gross margin, and other costs on a percentile basis provides us with additional insights. Figure 6 plots all possible percentile performance combinations of ROA, gross margin, and other costs. This allows us to see the expected ROA percentile of a company as a function of its gross margin and other costs percentile. Inspecting the other costs percentile axis we notice that relative improvements to a company’s other costs when low provides a greater expected benefit to a company’s ROA percentile than do improvements to gross margin, thereby supporting our analysis that when a company’s relative profitability is low the company should work toward reducing other costs. This is evident in the “drape” of the wireframe as we see the expected ROA percentile quickly rise when a company’s other costs percentile improves. However, the rapid improvement in ROA percentile when focusing on other costs quickly subsides, and at the extreme, a relentless focus on driving down other costs becomes a hindrance. The oft-quoted phrase, “you can’t cut yourself to greatness” looks to be more than a managerial fable.
Interestingly, the converse doesn’t hold for a company’s gross margins. All else held constant, a greater gross margin percentile equates to a greater ROA percentile. The interesting case that we are left with is that of those companies with both a high gross margin percentile and high other costs percentile. One might assume that this would invariably lead to a high ROA percentile. However, it is incumbent to understand first that we are evaluating gross margins and other costs percentiles in the absence of a major driver of ROA, namely, total asset turnover. Further the number of observations of companies with both a high other costs percentile and a high gross margin percentile is relatively small, which limits our ability to draw conclusions regarding those specific combinations of performance.
The rules apply as guidance for sustaining exceptional performance. When to begin applying them along the way to exceptional performance depends on where you start your journey. The worse your performance is in relative terms, the more significantly one should, all else equal, focus on cost reduction. As this is done effectively, and performance begins to improve, additional performance improvements turn increasingly on improving gross margin or asset turnover.
These findings have an intuitive appeal. Companies that find themselves in money-losing positions, unless suffering the consequences of fundamental strategic errors, are rarely at the cutting edge of operational efficiency. Consequently, when seeking simply to get back to profitability, cost reductions yield the most significant financial improvements.
This result is not what Rule No. 2—revenue before cost—prescribes. But then, a company that seeks merely to get out of the red is not aspiring to exceptional performance, at least not yet, so it is perhaps not too surprising that the rules driving exceptional performance do not apply.
As companies move into the middle of the distribution, however, they will typically have captured the benefits of the more readily implemented improvements. Further increasing profitability will thenceforth rely more heavily on gross margin improvements. Our case study research shows that achieving these improvements, in turn, relies on achieving price premiums through competitive differentiation.
To illustrate this phenomenon, consider the case of Cummins, an industrial goods manufacturer specializing in diesel and natural gas engines. The company has been able to improve its ROA percentile rank over the last 14 years from the 36th percentile to the 84th percentile, reaching as high as the 95th and 97th percentiles in recent years. As we look back to the beginning of the company’s transformation, sales in 2000 were flat at $6.6 billion with gross margins of 22.5 percent, a 1.9 percentage point decrease from the year prior, and other costs were 22.4 percent of sales, a rate it had been hovering near for the last several years. However, the company was committed to business improvement initiatives aimed at restructuring its workforce and consolidating its operations.8
In the short term, restructuring costs only helped to increase costs, but were necessary for Cummins’s long-term financial well-being. The focus on costs continued through the early 2000s as Cummins dealt with the aftermath of the bursting of the dot-com bubble and declining sales, but in 2002 as the effects of the recession began to subside and Cummins returned to growth, it found that it was well positioned to capitalize on its investments in fuel economy and emissions technology.9 Its positioning as a leader in these areas allowed the company to grow revenues rapidly between 2002 and 2007 at a compound annual growth rate of 17 percent.10 Throughout this period the focus on cost reduction continued as other costs declined to 16.1 percent in 2007 while gross margins held steady at about 22 percent. The net effect of these simultaneous changes was an improvement from the 42nd to 74th ROA percentile.
2007 marked an inflection point for Cummins. Between 2007 and 2013 growth slowed to a 5.8 percent compound annual growth rate, while other costs remained constant in the range of 16 percent to 19 percent and gross margins steadily improved from 21.8 percent to 26.7 percent. The net result of these factors was a rise in net income from $739 million in 2007 to $1.5 billion in 2013. This 101 percent improvement in net income in five years brought with it a move from the 74th to the 84th ROA percentile.
Cummins’ performance improvement from the 36th to 84th percentile of ROA follows a common trend over the last 14 years: namely, a focus on cost reductions when profitability is low relative to peers and a continuing shift toward a focus on gross margins as profitability improves. Cummins’ ability to affect this shift in performance can be attributed to the company’s ability to understand its relative positioning versus the competition, pick reasonable short-term goals, and set the right priorities along the way.
There was a time when every ocean voyage was a profoundly risky undertaking. As maps and navigation aids improved, what was once treacherous became merely fraught. Similarly, there is no guarantee that your company will become exceptional. Yet, as with every ocean voyage, the odds of a safe journey are only increased with an understanding of the starting point, identifying achievable ports of call along the way, knowing your destination, and sailing with the winds wherever possible.