“[ The stock market] is not a weighing machine, on which the value of each issue is recorded by an exact and impersonal mechanism, in accordance with its specific qualities. Rather should we say that it is a voting machine, whereon countless individuals register choices which are the product partly of reason and partly of emotion (emphasis in the original).”¹ So wrote Benjamin Graham and David Dodd in 1940, in a discussion relating the inherent worth of companies with the prices of their shares.

Graham and Dodd went on to point out the value of systematic analysis to the investment process, and how careful research on companies—both quantitative studies on their financial health and qualitative reviews of their competitive context and strength of management—could aid investors in evaluating whether the prices investors set by their votes offer attractive investment prospects.

This chapter considers the variety of styles and techniques equity managers apply to their investing, and highlights what Epoch considers to be at the core of companies’ economic value: free cash flow.

Manager Style

Among investors, deliberating over whether to pursue active versus passive approaches may be a binary, yes-or-no question. Within the community of active managers of equities, however, there are second-, third-, and fourth-level debates over how to best earn superior risk-adjusted returns from selecting securities. The main divisions are in the dimensions of style (value versus core versus growth) and research process (fundamental versus quantitative).

Each manager likely has its own definition of where its style falls within the different categories. Stated simply, however, growth-oriented managers seek out stocks where revenues and profits will expand more rapidly than the economy or the market as whole, often in industries innovating new technologies or services. In contrast, value managers are less concerned with sheer earnings growth, and seek out companies selling at relatively low multiples of earnings or book value per share (compared either to other stocks or to their own history), typically expecting some catalyst to emerge and trigger a higher price. Core strategies draw candidates from the entire market.

Fundamental managers rely on traditional financial research techniques—in-depth financial statement analysis, supplemented by qualitative evaluations of management skill, and subjective judgments about the future of a particular industry. Quantitative managers focus on a more objective analysis of data, numerically scoring companies on characteristics that they believe are associated with outperformance. With heavy reliance on computer models, they look at financial parameters, but also factor in measures derived from market trends such as price momentum and share volatility, as well as other statistics of their own design. Fundamental managers seek to know a lot about a relatively small number of companies, while quantitative managers seek to know a little about a large number of companies.

The boundaries defining managers’ styles were quite distinct a generation ago, but they have become blurred thanks to the availability of massive computing power and data resources at relatively low cost. Thus, in current practice, most managers apply some sort of quantitative technique in their investment process.

According to one comprehensive database of investment managers, published by eVestment, roughly 1,200 firms were managing strategies in large cap U.S. equities at year-end 2014. Firms pursuing growth, core and value strategies each made up about one-third of the population. Sixty-five percent of managers followed fundamental methods, while 16 percent deemed themselves heavily quantitative, and the remaining 19 percent reported a combined fundamental/quantitative process.

Epoch Investment Partners does not take a traditional view on growth versus value styles, or quantitative versus fundamental research techniques. We don’t find those distinctions especially meaningful. But we do firmly advocate active management. Accordingly, we do not believe the financial markets are “efficient” in the academic sense, where all information available on a security is incorporated in its price. That’s not to say that we think markets are chaotic, and that prices of individual securities are completely out of whack. In fact, the collective judgment of investors usually arrives at pretty reasonable values for most stocks in the fullness of time.

However, investors’ behavior can be highly flawed over shorter periods. On any given trading day, the market is arriving at mistaken judgments of varying sizes on hundreds or thousands of stocks, resulting from the tendencies toward loss aversion, mental accounting, trend extrapolation, and overconfidence described by behavioral finance.

Free Cash Flow Is the Measure of Value

The lifeblood of an operating business is cash. Managers of a business start with cash, then convert it into assets of various kinds, such as plant and equipment and inventories. When products are made and sold they are converted into accounts receivable, which then are collected as cash, and the cycle starts all over again.²

Simple enough. Over time, however, as businesses have become more complex and sophisticated, keeping track of their profits has become rather intricate, and there are many ways that companies can record earnings, in an accounting sense, that are not accompanied by cash flow. Consider the hypothetical example of a construction company with a contract to put up a large building, which is expected to take three years. Payment on the contract will be made when the building is complete. This transaction could be accounted for in a number of ways: one might be to wait to recognize the revenues and profit until the end of the project, when all is done and agreed to (accountants designate this technique the “completed contract” method).

Under that approach, though, the company is not recognizing its efforts as they go along, or matching the expenses of construction with the anticipated revenues. To measure that progress, management might record the expected payments as receivables as the construction moves along, under a technique known as “percentage of completion.”

Under that method, though, the company would show a profit for one-third of the project each year, but hasn’t collected any cash. This is a stylized example, but it serves to illustrate the gap that can arise between what a company estimates and believes its earnings are and will be, versus what has actually been collected in cash. If no cash has come in, has the company made a genuine profit each year, and if so, how much?

The question is not new. In 1985, Leopold Bernstein of Baruch College and Mostafa Maksy of the University Chicago wrote: “Analysts of financial statements and other users have long complained that the increasing intricacy of the accrual accounting system masks real cash flows from operations and widens their divergence from reported net income. Not only do they point to net operating cash inflows as ultimate validators of profitability, they also emphasize that it is cash, and not ‘net income,’ that must be used to repay loans, to replace and expand the stock of plant and equipment in use, and to pay dividends.”³

Corporate managers have many accounting techniques at their disposal, but there are three crucial areas where accounting rules can drive a wedge between a firm’s reported accounting performance and its true economic value: depreciation, accruals, and the treatment of research and development costs.

Depreciation

The greatest shortcoming of earnings reported under generally accepted accounting principles (GAAP) is probably depreciation expense. Through a well-meaning notion known as the matching principle, accounting rules try to allocate the cost of an asset over its expected productive life through depreciation, deferring the recognition of the full cost by expensing a portion each year. However, depreciation is based on estimates of assets’ useful lives, and if those estimates are not accurate, accounting depreciation can greatly distort the picture of their productive capacity, as well as the company’s earnings. The dollar amounts can be very large relative to revenues, and because the estimates span long periods, faulty estimates can have significant and long-lasting effects.

One classic example is commercial aircraft. Accounting conventions once commonly estimated the useful lives of airplanes at 10 to 15 years, and charged their costs to earnings over that period. In actual practice, however, many planes stay in the air and churn out revenues much longer—say 30 or 40 years. That may seem like a windfall to the company in the later years, as there is no longer any depreciation expense, but it also means that the charges to accounting earnings early on were too high, so that the accounting earnings of the past were understated.

At the other end of the spectrum, consider the case of a manufacturing firm that builds a new plant with an estimated life of 30 years, only to learn a year later that the widgets it was to produce suddenly have become technologically obsolete, and there’s no market for the product. In this case, the plant is not worth much in terms of future earnings, and the depreciation rate should be 100 percent.

Now, we are not faulting the accounting profession for not having perfect foresight: in estimating how long their planes or factories will be productive, companies have to start their accounting somewhere, and the accuracy of those estimates will vary. But our point is that the accumulated results of the accounting process can be highly inaccurate, so that a company’s accounting earnings can present a very different picture from what is relevant to investors—economic earnings, or free cash flow.

Accruals

Accruals are a way to anticipate a transaction—recording revenues before the related cash is received, or expenses before it is paid out. At the end of a month, a quarter, or a year, companies routinely accrue sales transactions and the related accounts receivable, or payrolls and other expenses where services have been rendered but not yet settled. When the cash eventually arrives at or leaves the company, the accrual is reversed and the cash transaction takes its place.

Accruals have been part of accounting for centuries, and are an invaluable tool in the accountant’s kit. However, like any sharp object they must be used with care. While revenues and expenses received and paid in cash are definite, accruals are merely conceptual, and by their nature estimates, so that companies can accrue whatever transactions management’s logic, and their auditors, will allow. If companies are too optimistic with their accruals, accounting earnings will be overstated relative to the actual cash they generate, and will eventually have to be scaled back in some future period.

Reported earnings can include large non-recurring items such as restructuring charges, extraordinary expenses, and gains or losses from discontinued operations, and while all are important events to take into account, they may not be relevant to future results. Consequently, investors typically try to strip out the effects of such special one-time items when assessing the sustainable portion of reported earnings.

A more serious problem arises from the smaller-scale, yet ongoing efforts of a company’s senior executives to manage their earnings. Common tactics include “channel stuffing”—artificially boosting sales to meet announced targets—or aggressive timing of the recognition of gains and losses to raise earnings in one quarter (and penalize future periods). Other examples of smaller-scale aggressive accounting are prematurely recognizing revenues before shipment or customers’ acceptance; recording revenue although customers are not obligated to pay; and capitalizing, that is, recording as inventories or other assets, costs that should be expensed.⁴ Additional problem areas are accounting methods used in acquisitions, assumptions applied to pension accounting, and subsidiary companies not reported as part of the parent company.⁵ To the extent that managements rely on these techniques, a company’s historical accounting earnings—even if they technically meet all required reporting principles—become a less reliable guide to the future.

It might be tempting to think that the accruals even themselves out over time, but they don’t, according to George Christy, a seasoned corporate banker and treasurer, in his 2009 book Free Cash Flow. He explains:

A viable company has no “end.” Every company is continually booking new accruals and modifying existing accruals and reserves. Even if there were no new accrual accounts put on the books, the world in which a company is operating is always changing as is the company itself, so the company must continually reassess and modify its accrual assumptions.⁶

Investors are not easily fooled, however, and are able to distinguish between the higher-quality portion of companies’ earnings transacted in cash, and the lower quality accrual portion. The topic has been rigorously studied, but in 1996 Richard Sloan, a professor at the University of California, Berkeley, conducted the first academic research demonstrating that the accrual portion of earnings, as measured by increases in companies’ noncash working capital (such as accounts receivable), is a less reliable indicator of future earnings than the cash flow portion. Investors in his samples didn’t detect the difference instantly, but they did catch on eventually. Sloan wrote that:

[S]tock prices act as if investors “fixate” on [accounting] earnings, failing to distinguish fully between the different properties of the accrual and cash flow components of earnings. Consequently, firms with relatively high (low) levels of accruals experience negative (positive) future abnormal stock returns that are concentrated around future earnings announcements.⁷

That is, share prices react to reported accounting earnings at first, but as the accruals are revealed, prices more often than not gravitate to levels that reflect their cash earning power.

That said, there are many companies in the market trading at nonzero prices even though they do not generate positive cash flow, as well as companies that produce little cash flow that trade at very high prices. As often as not these tend to be “concept stocks”—in the current market, Tesla Motors. Inc. is an example of the first, and Amazon.com, Inc. illustrates the second. In such cases, it seems investors look through the companies’ current financial position to some long-term future year when the companies’ business concepts, and free cash flow and dividends, are eventually to be realized.

In 2004, David Hirshleifer, a professor of finance at the University of California, Irvine, headed a team that undertook further research on the distinctions between cash profitability and accounting earnings. They found that when companies reached too far, and overstated their economic earnings through accruals, future profits suffered: “[W]hen cumulative net operating income (accounting value added) outstrips cumulative free cash flow (cash value added), subsequent earnings growth is weak.”⁸ And like Sloan, Hirshleifer’s team asserted that firms with high net operating accruals are overvalued by those investors with limited attention who naively concentrate on earnings-based valuation, and ignore mounting evidence of “relative lack of success in generating cash flows in excess of investment needs.”⁹

In other words, aggressive accounting may impress some investors for short periods, but ultimately the market sees through the artifice and values the company on its ability to generate free cash flow. Thus, the research of Sloan and Hirshleifer reinforce the investment aphorism, emphasized by banker George Christy, that “[Accounting] profit is an opinion; cash is a fact.”¹⁰

Research and Development Costs

For some companies, research and development (R&D) costs pose the same challenge as depreciation expense. For the most part, generally accepted accounting principles for U.S. companies (U.S. GAAP) require that companies treat their spending on research and development as a current expense. Costs of internally developed software and web site development are handled differently, however: they are deemed to have some ongoing value for the future, and are capitalized and amortized over the useful lives that the company estimates.¹¹ For many old-line companies these costs typically are minor, but for the growing cohort of technology companies software development may represent a large portion of productive assets. Like the depreciation of fixed assets, amortization of capitalized software incorporates subjective estimates of the useful life of internally developed software, which result in equally subjective charges to earnings. However, companies are required to disclose such amortization: investors that base their valuation methods on cash flow can readily make the needed adjustments, but investors working from unrevised GAAP earnings are dealing with softer numbers.

Differences in accounting standards and methods can complicate the comparison of companies based in the U.S. to those overseas. One significant variation is the treatment of research and development expense under International Financial Reporting Standards (IFRS) versus U.S. GAAP. IFRS requires companies to capitalize their successful R&D efforts in some circumstances, deferring the recognition of expense, with a wide variety of results. PricewaterhouseCoopers published a study in 2010, based on financial statements from 2008, which examined the R&D capitalization practices of a variety of European technology companies.¹² It found that computer and network companies capitalized relatively little—from 0 percent to 14 percent of their R&D spending—while software and internet companies capitalized from 0 percent to 61 percent, and alternative energy companies capitalized from 23 percent to 94 percent. When more spending is capitalized, current earnings appear to be higher—less of the cost is charged against them, and future periods are left to bear the burden.

European companies in other industries capitalize their R&D as well, including a high proportion of technology spending of the automotive sector. Pharmaceuticals, on the other hand, capitalize relatively little because regulatory approval of new drugs, which determines the ultimate value of the investment in research, tends to come late in the R&D cycle.¹³ Careful cash flow-oriented investors have financial disclosures at their disposal to put the results of U.S. and European companies on equal footing by looking through the accounting earnings to cash flow, but relying on unadjusted accounting earnings alone can lead to misleading comparisons of companies’ economic earning power.

The CFO Perspective

Companies themselves appear to place greater weight on earnings-based measures than on cash flow. F. W. Cook, a firm of compensation consultants, conducts an annual survey of executive compensation methodology among large companies, and for its 2015 report, said that 51 percent of companies relied on profits to measure performance, while just 11 percent looked to cash flow metrics. (However, most companies combine multiple statistics in setting compensation.) The leading gauge of performance, however, has become total shareholder return, cited by 55 percent of companies in 2015, up from 48 percent in 2012.¹⁶

A revealing survey conducted in 2006, by academics John Graham, Campbell Harvey, and Shiva Rajgopal, polled chief financial officers of about 350 U.S. companies—public and private, large and small.¹⁷ “We asked CFOs to rank-order the perceived importance of several competing measures of value: earnings, pro forma earnings, revenues, operating cash flows, free cash flows, and economic value added,” the authors reported: “Earnings are king. . . . Nearly two-thirds of the respondents ranked earnings as the #1 metric; fewer than 22 percent chose revenues or cash flow from operations.”

Executives’ preferences differed by company characteristics: private companies placed greater emphasis on cash flows, “which suggests that perhaps capital market motivations drive the focus on earnings.” The authors went on to say that CFOs in the study emphasized the importance of reporting steady, predictable patterns in earnings, and how crucial meeting market expectations of quarterly earnings is to share prices, as well as to CFOs’ current and futures career prospects: “The common belief is that a well-run and stable company should be able to ‘produce the dollars’ necessary to hit the earnings target, even in a year that is otherwise somewhat down. In other words, because of the common belief that everyone [manages accounting earnings], missing earnings indicates that a company has no available slack to deliver earnings . . . and must have already used up its cushion.”

It’s curious and indirect, but such a finding suggests that those companies that are forthright, and don’t manage their earnings, could place themselves at a disadvantage with investors. Accounting earnings are by their nature smoothed, while cash flows are more erratic: if executives believe investors will mark down their companies’ share prices for volatility in their results, a steady path of earnings is a natural choice, especially when condoned by generally accepted accounting principles.

Graham, Harvey, and Rajgopal, joined by Ilia Dichev, extended their survey research on earnings quality in 2013, polling 375 chief financial officers of U.S. companies. “CFOs endorse the sustainability notion of earnings above all else,” they concluded. “[This result] dovetails nicely with the well-documented importance of earnings for valuation, because most valuation models view a company as a stream of earnings and cash flows, and sustainability in profits is the key to projecting such variables.”¹⁸

CFOs were asked to provide a list of red flags for identifying poor earnings quality: “Lack of correlation between earnings and cash flows was the top choice. . . . The presence of lots of accruals and one-time charges and consistently beating analyst forecasts also scored highly.” One CFO in the survey said: “I think if earnings are not backed by actual cash flows, then they are not good earnings.”

The researchers also found that CFOs as a group believe that as many as 20 percent of companies intentionally misrepresent their earnings in any given year, through discretionary application of generally accepted accounting principles. Moreover, the CFOs surveyed believed that the misrepresentations were quite large—about 10 cents on every dollar. Two-thirds of the misstatements were thought to overstate actual earnings, while the balance underreported profits.

Although earnings management may be fairly common, detecting it from public sources is challenging, but worth the effort, in the researchers’ view. “The takeaway for analysts is that the identification of poor earnings quality has potential for significant rewards,” they wrote. “Companies with deteriorating earnings quality incur substantial price declines, lower price multiples, and higher costs of capital.”

To their credit, many companies are now reporting pro forma free cash flow measures when they release accounting earnings. And while it would be extremely helpful, in our opinion, for companies to publish free cash flow figures on a per share basis in their financial statements, it’s actually forbidden under GAAP:

Financial statements shall not report an amount of cash flow per share. Neither cash flow nor any component of it is an alternative to net income as an indicator of an enterprise’s performance, as reporting per share amounts might imply.^{19, 20}

We disagree. Cash flow may not be easy to interpret, but it’s not all that complicated, and the effort is worthwhile as it provides far better measures of companies’ successes in creating value. However, we are not eager to see the abandonment of GAAP earnings reports: the system has to have rules, but to the extent that accounting profits exaggerate true economic profits, they create inefficiencies in share prices, and open opportunities for earning excess returns through active portfolio management.