This chapter pulls together the theory and practice of markets and investors, and discusses, at a high level, Epoch’s process for selecting individual stocks for investment. To sum up the foundation of the past several chapters: first, we respectfully depart from the theoretical arguments of Modern Portfolio Theory, and substitute the observations of behavioral finance—that real-world investor behavior is complex and leads to inefficiencies in stock prices, creating mispricing in securities and opportunities for active managers. Second, we believe that the reliance by corporate managements, as well as by many investors, on backward-looking financial measures derived from generally accepted accounting principles (GAAP) financial statements distorts the financial picture many companies present. Moreover, accounting-based investment decisions can give rise to their own category of mispricings and inefficiencies. Accordingly, we believe that cash flow is the origin of value in stocks, and that forecasts of cash flows should be the basis for security selection.

In the next few pages we discuss Epoch’s framework for evaluating the cash flows of individual companies. We emphasize the importance of managements’ decisions in allocating their available cash flow—either by adding value to their businesses through capital reinvestment, or returning capital to the owners through share repurchases and dividends.

The Starting Point: Generating Free Cash Flow

Epoch seeks to invest in companies with business models that are transparent, and can be readily understood from the sources and uses of their free cash flow. As noted above, we define free cash flow as the cash provided by the company’s operations, less planned capital expenditures and cash taxes.¹ Management can thus deploy this uncommitted cash at its discretion.

Once a company has reckoned its free cash flow for a given period, how will management allocate it? Epoch outlines five possible destinations: reinvesting capital in the business through corporate acquisitions and capital spending on internal projects, as well as distributing cash to the owners through dividends, buying back outstanding shares, and repaying debt (Figure 10.1).² Of the companies Epoch chooses to research and monitor, we fully investigate their capital allocation process. Some companies turn out to be unappealing investments, for reasons such as a weak competitive position, low profitability, a poor growth outlook, or unwise capital application, and thus we reject them from further consideration. For those companies with attractive strategies and strong operating cash flow, if their share prices fall within Epoch’s estimates of reasonable value, we will add them to our portfolios. And if they are sound businesses but seem too expensive at the current price, we monitor their progress until their share prices move to a point where we think they can be capitalized upon.

Reinvesting capital for further growth and returning excess capital to the firm’s owners are not mutually exclusive. Depending on the economic and competitive environment, most companies are likely to find profitable projects for new investment, but not enough to consume all the free cash flow they create. In such cases a sensible capital allocation policy devotes some resources to expanding the business, and the rest to rewarding its owners.

Choosing to Reinvest

In order to grow, companies must reinvest in their businesses, but simply expanding the size of operations does not ensure an increase in the firm’s value. Basic principles of finance and investment hold that a company should invest only in new opportunities that are expected to earn a return higher than a firm’s marginal cost of capital (that is, the cost of an incremental issue of debt or equity). John Maynard Keynes formalized this notion in 1935, and called it the “marginal efficiency of capital.”³ Keynes was a peerless economic thinker and writer, but his definitions on this topic are roundabout and complex, so we will not quote them here. To paraphrase, however, the marginal efficiency of capital is a breakeven rate of return, which equates the present value of the expected benefits from investing in a new plant or company to the cost of the new investment.

Capital Investment: Returns and Capital Costs

For management of a current-day corporation, the challenge in qualifying new projects is to identify the prospective free cash flow they might generate, anticipating as best it can future changes in revenues, costs, and the competitive environment. This amount, which Keynes called the “prospective yield,” is then compared to the cost of the new investment, including financing. If the project is expected to generate more cash than it consumes, including the marginal (that is, incremental) cost of financing the new project, then undertaking the investment in the project can increase the value of the business.

It is crucial that managements evaluate projects on their cash flow merits, rather than projections of accounting-based earnings. Some investments that generate apparently positive accounting earnings may in fact reduce the value of a firm when considered through the lens of cash flow, primarily due to the accounting mechanism of depreciation. Depreciation defers the cost of an investment over time, rather than recognize the full cash outflow at the start, and thus reduces the cost of the investment in present value terms. Therefore a project may appear to have a positive net present value from an accounting view of the associated costs and revenues, while at the same time its net present value calculated from the expected cash flows could be negative.

When looking at prospective returns on invested capital (ROIC) versus the cost of capital, it is essential to distinguish between marginal and average measures, and to make an evaluation based on the marginal figures. Weighted average cost of capital (WACC) measures the proportional cost of a company’s total existing base of capital, for both debt and equity combined. The weighted average cost of the debt portion simply measures the interest rates the company has to pay on each form of debt, weighted to reflect the size of the borrowings.

The cost of equity capital, however, is harder to observe directly. A firm is not required to make cash distributions to shareholders, but that does not make equity capital free: even if equity holders do not receive cash dividends from the company, they expect to earn a return on their investment through the appreciation of the stock price. Thus, the cost of equity capital is usually thought of as the expected return on equities in general, adjusted for the volatility of the equity returns of the individual firm (as measured by the beta coefficient in the CAPM—firms with higher betas are thought to have a higher cost of equity capital, and vice versa).

Suppose the capital structure of the mythical Griffin Corporation consists of 80 percent equity and 20 percent debt, and its total enterprise value (equity plus debt) is $1 billion. The debt carries a weighted average interest rate of 5 percent, while the equity cost of capital is assumed to be 10 percent. The firm’s WACC will therefore be 9 percent.⁴ The firm operates one business line, earning a 14 percent return on its invested capital—500 basis points over the WACC.

Suppose further that the firm is considering an investment in a new line of business, estimated to cost $200 million, which is to be funded entirely with newly issued debt. Management believes that the new business will earn an ROIC of 8 percent. On the cost side, interest rates have fallen, and the firm can now borrow at 3 percent, rather than the 5 percent it pays on its debt borrowed in the past. Should Griffin’s management go ahead with the investment?

From the perspective of average cost of capital, the answer would be no. The existing WACC is 9 percent, and an 8 percent return would fall short by 100 basis points. Against a 9 percent hurdle rate, the project would not earn its keep and destroy value.

But the project will not be funded by the firm’s existing capital; instead, it will be funded by an incremental $200 million in debt costing 3 percent. The 8 percent expected return on the project would be 500 basis points higher than the cost of the capital to fund the project. Therefore, on its own merits the new project would be profitable, and the company would create value by taking on the investment, notwithstanding the apparently unfavorable comparison against the averages.

Similarly, on the return side of the calculation, what matters is the marginal ROIC, rather than the average ROIC the firm earns on all of its businesses. Suppose that instead of earning a projected 8 percent, the proposed new business was expected to earn only 2 percent. If the firm goes ahead with the investment, its average ROIC on all businesses would drop from 14 percent to 12 percent. (Equal to $1 billion times a 14 percent average ROIC, plus $200 million times a 2 percent marginal ROIC, divided by total assets of $1,200.) In addition, adding $200 million in debt costing just 3 percent would reduce the firm’s WACC. As a first approximation—ignoring any impact on the cost of equity capital—the lower cost of the new debt would reduce the WACC from 9 percent to 8 percent. However, adding the new debt might make the enterprise look more risky to equity holders, so we assume that the cost of Griffin’s equity capital would rise to 11 percent. All in, the net effect on WACC would then be a reduction to 8.7 percent. (Equal to $800 times an 11 percent cost of equity, plus $200 times a 5 percent average cost on existing debt, plus $200 times a 3 percent marginal cost of debt, divided by total liabilities and equity of $1,200.)

Since the new average ROIC of 12 percent would still be higher than Griffin’s new 8.7 percent WACC, does the investment makes sense? In this case the answer is no because the cost of the new borrowing, 3 percent, would be funding an investment earning just 2 percent. That’s a formula for destroying value.

What matters in evaluating new projects are not the averages of capital costs and investment returns, but instead the marginal costs and returns they generate.

Once More: Cash Flow-Based Measures Are Superior

We’ll take one more look, in depth, at how decision making on the financial measures of free cash flow and ROIC is superior to relying on accounting-based techniques.

Consider the five-year records of two hypothetical companies, Green Enterprises and Red Corporation (Table 10.1).⁵ In the initial year, both companies earn revenues of $1,000, and generate accounting earnings of $100. And for both, revenues and earnings grow 5 percent annually. On the basis of accounting earnings alone, the two enterprises would appear to be of equal value.

But Green is actually worth far more than Red, as revealed by a comparison of their cash flows. Green has a higher ROIC, and is able to maintain its 5 percent growth in revenues and earnings by investing just 25 percent of its earnings, while Red needs to reinvest 50 percent of its cash flow to keep growth steady. Accounting earnings of the two are equal, but Green generates $75 in free cash flow in year 1, while Red generates only $50 in free cash flow. In years 2 through 5, the respective free cash flow figures then grow at 5 percent per year, in line with revenues and accounting earnings.

In choosing between the two companies as investment candidates, comparing accounting earnings would be a constructive start, but would not go far enough: they provide no insight on the resources required to keep the business going. The additional dimension of cash flow provides a more complete view, showing how much will be claimed by reinvestment, and what will be left over for the owners at the end of each year. In this example, the simple sum of the free cash flow for Red over the five years is $276, while the higher ROIC at Green generates $415. In present-value terms, assuming a growth rate for both companies of 5 percent in perpetuity and a cost of capital of 10 percent, Green is worth $1,500, while Red is worth $1,000.

Table 10.1 conveys two messages. First, a company’s value originates not in its accounting earnings, but instead from its free cash flow. Second, a higher ROIC leads to a higher valuation, holding other factors equal.

Consider what the traditional earnings-based valuation metrics that most investors rely on would say about these two companies. Both companies have the same $100 in earnings, but because of its higher ROIC and lower reinvestment requirements, Green is worth $1,500, while Red is only worth $1,000. That means that if both companies were trading at their fair value, Green would have a P/E of 15, while Red would have a P/E of 10. Investors who rely on P/E as a measure of value would say that Red is more attractively priced than Green, yet both are in fact fairly priced. Now suppose that Green was trading at $1,400 and Red was trading at $1,100. Green’s P/E would be 14, and Red’s would be 11. The P/E investor would still say Red is priced more “cheaply,” yet now Green is actually undervalued and Red is overvalued. A P/E ratio simply doesn’t tell the investor anything about a company’s underlying ROIC, and as a result its usefulness is limited.

Some investors believe that they can overcome this shortcoming of the P/E ratio by looking at what is known as a “PEG ratio,” which divides the P/E ratio by the rate of the company’s expected growth. Companies with higher expected growth will trade at higher P/E ratios, the thinking goes, and a P/E-to-growth ratio will enable an investor to make an apples-to-apples comparison. A company with a P/E of 10 and expected growth of 5 percent will have the same PEG ratio as a company with a P/E of 20 and expected growth of 10 percent, implying that they are similarly valued. But our example shows that the PEG ratio can also be misleading. Both Green and Red are growing at 5 percent per year, so if they were trading at their fair values, Green would have a PEG ratio of 3 (P/E of 15 divided by 5 percent growth) while Red would have a PEG ratio of 2 (P/E of 10, growth of 5 percent). This implies that Red is somehow the better value relative to its growth rate, yet we know that both companies are fairly priced. And in the situation where Green trades at $1,400 and Red is at $1,100, Green would have a PEG ratio of 2.8 (P/E of 14 divided by 5 percent growth) and Red would have a PEG ratio of 2.2. As we saw when we used the P/E ratio alone, Red looks cheaper than Green, yet Red is overvalued and Green is the one that is undervalued, even with its higher P/E and higher PEG ratio.

From the perspective of investors in the stock market, the superior framework for evaluating a company’s worth is a focus on capital allocation—how the company deploys the money it has already made. Accounting-based measures such as earnings per share, book value and the price-earnings ratio aren’t designed to estimate what the company’s business is expected to earn over the life of its investments, and thus offer information far less valuable than the dynamics of return on investment and the cost of capital.

To summarize, company managers maximize returns to shareholders by allocating capital toward its most profitable use. The first step is investing only in opportunities that can earn an ROIC higher than the marginal cost of the capital needed to fund the investment. When projects cannot justify an ROIC higher than the marginal cost of capital, managers should instead return capital to shareholders, who may have a higher-return alternative for the funds.

Trends in Capital Allocation

Managements of U.S. corporations have shown a significant shift in their approach to capital allocation over the past 20 years. Prior to 2005, free cash flow invested in capital expenditures and acquisitions swamped distributions to shareholders in dividends and share buybacks (Figure 10.2): from 1994 through 2005, corporations in the broad S&P 1500 index carried out $7.3 trillion in new investment, while returning $4.1 trillion to owners. More recently, cash flow allocation has been more balanced: from 2005 through 2013 new investment still claimed a greater share, at an aggregate $7.1 trillion, but share buybacks and dividends were not far behind, at $6.1 trillion.⁶ (Discretionary debt paydowns are not included in distributions to owners, as we have not found an authoritative source for them.)

Dividends

Dividends are the traditional means of rewarding shareholders. They also can provide important insights into the thinking of a company’s managers: dividends typically reflect the ability to earn an economic profit with consistency, and many investors see steady dividend increases as an indicator of current financial health. Dividends also can serve as a signal from management: companies are reluctant to cut their dividends, so that steady increases can speak to insiders’ confidence in the future prospects for the business.⁷

While dividends are a hot topic among investors today, they have gone in and out of favor as a force behind the valuation of share prices. For most of the history of corporate America, investors cherished companies that paid consistent and rising dividends. Going back 20 or 30 years, however, companies slowed the growth of distributions (although the total paid continued to rise).⁸ For the companies in the S&P 500, for example, 94 percent of companies distributed dividends to shareholders in 1980,⁹ paying out about 40 percent of annual earnings.¹⁰ Through the 1980s and 1990s, the share of companies paying dividends fell steadily (reflecting in part that companies increased their share buyback activity). The share of S&P 500 companies paying cash dividends hit bottom at 70 percent in 2002, but has since steadily recovered—with a few years’ break after the 2008 financial crisis—to 83 percent of companies in 2015.¹¹ Compared to a recent low of $47.2 billion in third quarter 2009, dividends to shareholders had doubled to $94.4 billion by second quarter 2015, and forecasts called for a record year for 2015, notwithstanding cutbacks by energy and commodities companies necessitated by sharp drops in commodities prices.¹²

Seeking companies paying attractive dividends can indeed deliver superior returns. S&P Dow Jones Indices publishes a passive index made up of companies in the S&P 500 that have consistently raised their dividends over 25 years; it earned total annual returns through January 2016 of 14.0 percent over five years (vs. 10.9 percent for the S&P 500) and 9.8 percent over 10 years (vs. 6.5 percent for the broad index).¹³ Risk-adjusted returns over those periods were higher as well.

One line of traditional investment thinking asserts that those companies paying out large proportions of their earnings in dividends may limit their future growth by starving the enterprise of needed cash, while those that retain a greater portion of profits can promote faster growth. It’s a logical possibility, but the notion was refuted a few years back in a study by investment managers Robert Arnott and Clifford Asness, motivated by the low payouts for the five years ending in 2001.

Rather than confirming the conventional thinking that dividends and earnings growth tended to move in opposite directions, they found instead a positive relationship:

The historical evidence strongly suggests that expected future earnings growth is fastest when current payout ratios are high and slowest when payout ratios are low. . . . Our evidence thus contradicts the views of many who believe that substantial reinvestment of retained earnings will fuel faster future earnings growth. Rather, it is consistent with anecdotal tales about managers signaling their earnings expectations through dividends or engaging, at times, in inefficient empire building.¹⁴

Since the time of that study, many corporations have in fact increased their distributions, whether through dividends or share repurchases, often in response to prodding from activist shareholders. In 2004, for instance, Microsoft Corporation paid from its large cash hoard an epic special dividend of $32 billion, or about $3 per share, and doubled its usual dividend. (At the time the move was thought to be the largest such distribution on record.¹⁵) Payouts have only become more generous, however: Apple Inc. distributed $56 billion to shareholders in fiscal 2014, through a combination of dividends and share repurchases (Apple’s 2015 fiscal year saw smaller, but still large, dividends and repurchases of $46 billion.¹⁶)

Share Repurchases

During the past 10 years, distributions through dividends have been surpassed by corporations’ repurchases of their shares (Figure 10.3). Notwithstanding their popularity, however, financial thinkers, both academics and investment practitioners, are split on the merits of share repurchases. Critics assert that repurchases divert crucial capital from new investment; that companies funding repurchases with debt are adding unnecessary leverage; and that some managements undertake repurchases to drive up earnings per share in order to artificially hit their compensation targets. These criticisms notwithstanding, repurchases now are in nearly every chief financial officer’s toolbox, and during third quarter 2015 more than a fifth of companies in the S&P 500 reduced their share counts by at least 4 percent from a year earlier.¹⁷ For the year ended September 2015, S&P 500 corporations repurchased $559 billion of their own shares, a year-over-year increase of 1.6 percent.¹⁸

Epoch is squarely in favor of share repurchases as a capital allocation tool—assuming they are carried out under the right conditions. First, a share repurchase is reasonable only in cases where the company cannot find sufficient profitable investment opportunities. Second, to be effective a buyback must represent a true reduction in the share base, and any contemporaneous share issuances must not be so large as to offset the effect of the repurchase. When these requirements are met, share repurchases are functionally equivalent to cash dividend payouts of the same magnitude, and thus can bring the same value to shareholders as a cash dividend.¹⁹

Now, repurchasing shares by itself does not increase a company’s free cash flow, and thus does not itself create a higher value for the enterprise as a whole. (To be precise, a company would likely see an increase in cash flow from the dividends it wouldn’t have to distribute on the repurchased shares, but such amounts would typically be minor.) Instead, it’s a question of proportion. From the perspective of the remaining shareholders, each share of stock becomes more valuable—with fewer shares outstanding each owner has a claim on a greater share of the company’s future cash flows. The box “Share Repurchase Example” provides a simple but representative example.

Going back 30 years and more, share repurchases were uncommon, and dividends were the primary means for distributing cash to shareholders. Then in 1982, the U.S. Securities and Exchange Commission enacted Rule 10b-18 of the Securities Exchange Act of 1934, which provided companies trading in their own securities a safe harbor from charges of manipulating their stocks’ prices. With this protection, U.S. corporate buybacks surged for a few years in the late 1980s, but really took off around 1995, and in 1997 they surpassed dividends as means of distributions to shareholders (Figure 10.3, above). Among all publicly traded U.S. companies, the proportion paying dividends fell off from 78 percent in 1978 to 40 percent in 2013, while those repurchasing shares increased from 28 percent to 43 percent.²⁰ (However, the trend among U.S. companies in the aggregate toward not paying dividends runs counter to the rising proportion of S&P 500 companies paying dividends cited above.)

Academic studies reach varying conclusions on whether stock prices benefit from share repurchases, but one comprehensive paper, authored in 2008 by Jeffrey Pontiff and Artezima Woodgate, respectively of Boston College and the University of Washington and Russell Investment Group, made several favorable findings.²¹ Their work examined both share issuance and repurchases over very long periods, and shows that stocks’ future returns are strongly related to changes in shares outstanding. Overall, it’s not a large effect, but it does indicate that repurchases are associated with positive future returns. Pontiff, working with R. David McLean and Akiko Watanabe of the University of Alberta, extended the research to international markets in 2009, and found a similar strong effect in countries with developed and active markets for share issuance.²²

Recent market results confirm the academic work. S&P Dow Jones Indices presents an index focused on the 100 companies in the S&P 500 most active in repurchasing their shares (the S&P 500 Buyback Index), and it has outperformed the broad market. For the five years ended January 2016, the buyback index returned 13.1 percent annually versus 10.9 percent for the S&P 500, and for the 10 years then ended, the buyback subset earned 9.0 percent against 6.5 percent.²³

To be sure, not all share repurchases are constructive, and some of the objections expressed by market observers are valid. A well-researched 2015 paper from investment manager Research Affiliates, “Are Buybacks an Oasis or a Mirage?,” contends that some of the largest recent buybacks were not productive because the companies issued new shares that overwhelmed the effect of the share repurchase, particularly in fulfilling employee stock option compensation programs. Other significant new issuances were made to pay for mergers and acquisitions. The authors also criticize repurchases paid for with newly issued debt: “Dilution of earnings can also occur because a company issues debt, funneling earnings away from dividend payments to shareholders and toward principal and interest payments to the company’s lenders. . . . Five of the largest debt issuers [in 2014] were also among the largest repurchasers of equity.” They conclude that in 2014, notwithstanding all the headlines about buybacks, managements of U.S. corporations in the aggregate diluted their equity base by 1.8 percent.²⁴

Along similar lines, Epoch does not regard all share repurchases favorably. A critical part of our approach to identifying candidates for investment calls for regular visits with management teams—to gain a sense of their capital allocation philosophies, of which share repurchases are a key component, and how their views compare to the actual record in different market and economic environments.

Debt Buydowns

A corporation’s buying back of its own debt is another application of free cash flow that can increase the value of shareholder equity. To illustrate this idea we look to the pioneering research, dating back to 1958, of financial economists Franco Modigliani and Merton Miller, then professor and associate professor, respectively, at the Carnegie Institute of Technology (now Carnegie Mellon University). Each was ultimately awarded a Nobel Prize in Economics recognizing their many contributions to the field (Modigliani in 1985, and Miller in 1990).

Prior to that time, there had been little systematic academic study of corporate finance, and Modigliani and Miller set out to measure enterprises’ cost of capital and risk in a world of uncertain returns on their investment. In their initial paper, they theorized that what determined the value of a firm was not its capital structure—that is, the proportions of debt and equity that funded a company’s operations—but rather the cash flows generated by its assets. In turn, they asserted, changing a company’s capital structure—shifting debt to equity, or vice versa—would not change its value, unless the capital change altered the cash flows as well. Thus, a company’s value is not determined by whether management raises funds as debt or equity; instead, value comes from the quality of managers’ decisions on investment, and the resulting cash flows.²⁵

Observers are not unanimous on the merits of the early repurchase of debt. “Assuming there is no need to pay down debt to target levels, managers should probably consider share repurchases or extraordinary dividends, since these send a favorable signal to capital markets,” according to the authors of Valuation: Measuring and Managing the Value of Companies: “Voluntary debt repayments do not represent a positive signal, unless the company is close to or in financial distress.”²⁶ In their view, increasing the proportion of cash flow paid as dividends sends the strongest signal to investors.

We at Epoch respectfully disagree, and see debt repayments as a constructive application of free cash flow—but as before, assuming the company is unable to identify business investment opportunities to advance growth in value.

The practical implication for equity investors is this: when a company applies free cash flow to repay debt, the transaction transfers proportional wealth from bondholders to stockholders. Assuming free cash flows from the company’s underlying business hold steady, the total value of the firm is unchanged. Accordingly, when claims of bondholders are reduced, the future free cash flow attributable to stockholders is increased, as is their proportionate value in the firm.

Capital Allocation: What’s the Right Mix?

Assuming they have earned the luxury of free cash flow, corporate managers have many levers to pull—investing in their businesses when that investment is expected to be profitable, and distributing capital to shareholders when it’s not. For the outside observer, and the practical matter of selecting stocks for portfolios, the analysis has many dimensions.

Understanding the starting point—how a company generates its free cash flow, and its likely growth rate—is the most straightforward. It can be understood, measured, and forecast through fundamental analysis, including the microeconomic dynamics of industries and the interpretation of companies’ financial statements.²⁷

Grasping a company’s capital allocation policy is more complicated and subjective. A critical initial step in Epoch’s investment approach is to interview senior management teams on their capital allocation philosophy (and return to the topic frequently, to learn how the story may have changed). From there we validate their thinking, examining the historical record to see that the company’s free cash flow has been put to its highest and best use—that they have identified investment opportunities that enhance the value of the firm for the long run, and that distributions to owners have been productive (rather than focused on short-term manipulation of accounting earnings).²⁸

We have found no universal constant for the best mix of investment and return of capital: Each company’s investment set, capital structure, and opportunity costs are unique. Among the five potential uses of free cash flow, we believe all to be effective, although each may add more or less value at different companies, and at varying points in an economic cycle.

The portfolio management and analyst team at Epoch has considerable training, practical knowledge and market experience. But the investment world grows increasingly complex and gains velocity each year. In the final section of our book, we look at the role of technology in informing and expanding the investment world generally, and how Epoch has pushed an evolution in the firm’s investment process through the infusion of technology.