CHAPTER THIRTEEN
Measuring Business Performance
ONE OF THE BASIC causes of poor performance on the part of analysts, investors, and business managers is the yardstick they use to determine how a business is doing—“earnings per share.”
Performance in a business means applying capital productively and there is only one appropriate yardstick of business performance. This is the return on all assets employed or on all capital invested (the two differ, but not significantly). Whether the assets come from the outside or inside makes no difference. Retained earnings are just as much money as a bank loan or new equity. A business that does not earn the going cost of capital on all the money in the business fails to cover its true costs and has an earnings deficiency, whatever its earnings per share.
The return on assets must include all moneys available to service capital. This includes not only profits from which dividends can be paid, but also all interest charges on all debt. It includes depreciation, which does not figure in earnings per share, but it excludes inventory profits, which frequently do, depending on a company’s accounting practices.
These are the measures that make economic sense. If investment flows to companies with high returns by these yardsticks, the economy’s performance will be optimized. These are the measures of the economic regulating function usually assigned to “profits.”
What then are the “earnings per share” that companies and their accountants grind out so regularly and publicize so mightily? The conventional “earnings per share” figure not only doesn’t measure corporate performance but it rarely even measures true “earnings per share.” The term is a misnomer. What it really represents is “taxable earnings.” It is what is left after all the charges the tax collector accepts as deductible. But this is a purely arbitrary figure that has little or nothing to do with economic performance.
“Earnings per share” as reported are practically never what people think they are, that is “earnings available to the shareholder.” If they were, companies could and would distribute most or all in the form of dividends—and practically no company does. Before one can really know what a company has truly earned on its equity capital, he therefore has to correct the reported figures for those capital charges, which, while genuine costs, are not accepted as tax deductible by the tax collector and are therefore included in reported “earnings per share.” This requires starting out with return on assets or on total capital employed. Only this figure shows what capital charges are needed over and above those which the tax collector accepts.
There are four such genuine capital charges—each a true cost, even if included in the reported earnings figure. The first is deficiency of revenue as measured against the cash needs ahead—the only area, by the way, to which analysts pay much attention. A business that cannot provide out of its revenues for the foreseeable cash needs of operations, including service on its debt, does not earn enough.
A business, secondly, must earn the going cost of capital on all the money in the business. A company which shows high “earnings per share” because, for example, it still enjoys the benefits of the lower interest rates of the past is using up capital and reporting it as earnings. Such a business cannot raise money without depressing its earnings. But sooner or later—and usually sooner—the low-cost capital has to be replaced; and then the going cost of capital has to be paid. A gain resulting from low money costs of the past should go into a reserve rather than into earnings. It does not truly reflect the company’s profitability.
The third adjustment is the provision for known and foreseeable risks. This is a genuine cost, as is any insurance premium. The most common risk is the cyclical one. It is a known and foreseeable hazard and has high probability. One year’s earnings, like any one monthly or quarterly figure, are therefore by themselves misleading unless adjusted for a cyclical period. Another typical risk for which earnings need to be adjusted is the high risk of overexposure and vulnerability after a period of rapid growth. It is prudent to assume that even a slight setback after such a period will reduce sales to where they would have been if the company or the industry had grown throughout the period at a rate somewhat between that of the lowest year in the period and the average year. This formula, though quite unscientific, fits remarkably well with actual experience—that of the mobile-home industry, for instance, in the sixties and early seventies. And then one adjusts earnings to the most probable long-term figure, considering the risk of growth in order to have a reliable measurement of business or industry performance.
Finally, the “earnings per share” figure needs to be adjusted to account for the known and foreseeable needs of the business. One of them is the need to provide for the growth required to maintain a company’s market position in an expanding market or its technological leadership in an expanding and changing technology. A company that fails to do so endangers its very survival. A second such need is protection of capital against the ravages of inflation. Depreciation surely has to be adjusted to inflation rather than be based on historical cost.
All this was known seventy-five years ago. “I never listen to the securities analysts—I listen to the credit analysts,” Bernard Baruch is reputed to have said when asked to explain his performance as Wall Street’s biggest and boldest speculator in the early decades of this century. Fifty years ago the DuPont Company codified this knowledge in its famous “return on investment” charts, around which the DuPont management was organized.
But the definitive work on earnings, on costs of capital, and on measuring economic performance has been done in the last thirty years by a whole generation of business economists.
As a result of this work, we know that “assets employed” or “capital invested” mean precisely what the words say. Whether the money has been put into buildings or into receivables is irrelevant; there are different uses for money, but no different money. Whether the money is equity or loans also makes no difference, nor does it matter that the money invested is “our own” rather than raised on the outside. Earning power and, above all, “earnings per share” are seriously damaged—and very soon—if retained earnings are invested at a lower rate than the full cost of outside capital.
We further know what to include in “return on capital” or “return on assets.” Since the purpose is to measure the economic performance of a business, all moneys available to service capital are part of “return.” This includes all interest charges on all debt, depreciation (which in economic terms is essentially nontaxable return), as well as what is traditionally considered “earnings on equity capital.” It does not, however, include profits which do not reflect the earning power of the business. Inventory losses are genuine losses. But inventory profits do not belong in “return on assets” or in “return on capital employed.” Other nonrecurring gains incidental to the business, such as a gain from selling a plant, are also not part of “return on capital employed”—except in a company whose business it is to buy and sell plants.
Return on all assets or on all capital investment is not the only yardstick available in measuring the performance of a business. Indeed, every business might well use a second one. In a manufacturing business, for instance, return on “value added in manufacturing”—that is on the difference between revenue on goods sold and money paid out for supplies and materials—is an important measurement. It is very sensitive and a “leading indicator,” which tends to go up or down before returns on total capital invested or on assets employed show significant changes. For a retail business, return on selling space is similarly an important indicator of performance.
But these additional measurements have to fit the individual business. The “value added” yardstick makes sense only for businesses that are truly manufacturing businesses. Also, this second yardstick is usually available only to the insider and rarely to an outsider.
Though many business executives are skeptical of the accuracy of “earnings per share” as a business measurement, they often say, “What choice do we have? The market uses this measure whether it makes sense or not.” But this is a half-truth at best. Witness the enormous differences in “price-earnings ratios.” The main reason is that the stock market tends to value a stock primarily on the basis of a rough guess at total return rather than on the basis of the highly publicized and visible “earnings per share” figure. Despite its follies, foibles, and fashions, the stock market is a good deal more rational than the “experts,” at least over any extended period of time.
The best examples of this were the stock prices during the acquisition binge of the late sixties. A good many of the “conglomerateurs” played the “earnings per share” game and structured the purchases of an acquisition in such a way as to show increased “per share earnings” even though total returns did not go up and often went down. The people who sold their businesses against what eventually came to be known as “Chinese money” all lost heavily in the end—unless they immediately sold the securities they got. The stock market speedily adjusted the share price to the actual returns on capital rather than to reported per share earnings.
Thus there is something to the old-fashioned belief that the stock price reflects the discounted value of future dividends, that is, the present value of future earnings actually available for distribution.
It is possible, with a little work and ingenuity, to determine with adequate probability the return performance of publicly owned companies. But it might not be a bad idea for companies to provide the necessary information themselves. The demand for “full disclosure” increasingly focuses on the economic prospects of a company and on its economic performance—that is, on the extent to which it actually produces wealth out of the resources entrusted to it.
The key figure for this is return on all assets (or on capital employed) related to cash needs, to cost of capital, to risks and needs. This rather than the meaningless “earnings per share” figure is what the public, the SEC, the analysts, and, above all, the stockholders should expect—and might demand—from a company’s published accounts and annual statements.
And it is this figure that should provide the link between business performance and executive compensation. It is legitimate and desirable to relate executive rewards to company performance—but it had better be true performance. To tie compensation to reported “earnings per share” subordinates performance to appearances. It may even reward executives for milking rather than building a company.
(1976)