By Robert E Katz, MComm CA (SA)

he ascendancy of shareholders in most developed countries has led more and more managers to focus on value creation as the most important metric of corporate performance. Both earnings and cash flow valuation techniques are widely used in practice. Whilst neither method is so flawed that they cannot be relied upon, over time it has come to be recognised that cash flow valuation techniques result in more accurate and reliable results than earnings based techniques. Only in very rare situations can earnings be deemed a surrogate for cash, therefore the question of which technique to use will usually apply. The key to effective valuation remains unchanged: choosing the right methods for estimating critical variables and understanding the real nature of value creation.

Discounted cash flow, Free cash flow, Predicting future income, Adjusting for risk, Cost of equity,Value, Price Earnings, Multiple, Operating cycles, Credibility & Estimates.

“There is nothing so dangerous as the pursuit of a rational investment policy in an irrational world”. John Maynard Keynes.

In general, there are two approaches to valuation. The first, being discounted cashflow valuation, relates the value of an asset to the present value of expected future cashflows generated by that asset. The second, the capitalised earnings approach, is where a stream of maintainable earnings is capitalised at a price/earnings multiple to arrive at equity value. In theory, both methods are capable of producing the same estimates of intrinsic value, but in actuality, there can be significant differences in outcomes, depending upon which method is adopted (Francis, Olsson & Oswald, 2000:7).

Both earnings and cash flow valuation techniques are widely used in practice. Whilst neither method is so flawed that they cannot be relied upon, over time it has come to be recognised that cash flow valuation techniques result in more accurate and reliable results than earnings based techniques.

It is useful to acknowledge from the outset that valuation is an imprecise science. All valuations, be they cash or earnings based, involve an educated guess whereby analysts peer into an uncertain future and attempt to predict what a company might be worth (Lee, 2003:4).

The purpose of this essay is to investigate both earnings and cash flow valuation techniques so as to ascertain:

  • If any one technique is so fundamentally flawed in practice so that it should not be relied upon; and
  • If earnings can be deemed a surrogate for cash.

Both earnings and cash flow valuation techniques involve two fundamental tasks:

  • Predicting future income flows; and
  • Adjusting for risk.

These basic observations on value imply that valuations are forward looking i.e. they look to future income flows, only using past income flows as a means of corroborating the forecasts.

It is now recognised that the most relevant form of income to be predicted is the entity’s ability to generate cash, in particular ‘marginal free cash’ (Stewart, 1991:307).

Adjusting for risk requires an assessment to be made of the market price of risk i.e. the compensation required by investors for taking on the risk, rather than investing in risk free investments. The Capital Asset Pricing Model (hereafter CAPM), which calculates the cost of equity, is most commonly used for making this assessment for discounted cashflow (hereafter DCF) valuations. Although CAPM is potentially flawed, no universally acceptable technique has emerged to replace it (Black, Wright & Bachman, 1998:29-32).

3.1 The free cash flow method

A company’s value is driven by its ability to generate cash flow over the long term. Furthermore, a company’s cash-flow generating ability (and hence its value creation ability) is driven by long-term growth and returns that the company earns on invested capital relative to its cost of capital (Copeland, Koller & Murrin, 2000:131).

Free Cash Flow (hereafter FCF) is the net cash generated from operations after taking into account the cash reinvested to fund future growth. It is the primary source of cash available to repay the suppliers of debt and equity and to compensate them for risk, and thus it is the key component of value creation (Copeland,et al. 2000:167-169).

The present value of the expected free cash flows from the asset results in the value of that asset. The discount rate when discounting cash flows is always the cost of capital, which is the weighted average of cost of equity and cost of debt, using target market weightings (Brealey & Meyers, 2003:531).

3.2 Advantages
The FCF method encourages a move from an accounting basis of evaluating an entity’s performance to a more economic one. Management time and effort can be focused on the four actions that create value:

  • Cash flows;
  • The growth rate;
  • The length of the growth period; and
  • The cost of capital (Black, et al. 1998:49).

3.3 Derivations of the Free Cash Flow Methodology
A number of adaptations of the basic FCF methodology theme have been developed. These models feature similar advantages and limitations. The main advantage is that they are conceptually sound and can be applied to most companies. The disadvantage is that they require explicit forecasting of future cash flows and discount rates (Lee, 2003:5). Some of these variations are:

  • Economic Value Added (EVA);
  • Cash Flow Return on Investment (CFROI); and
  • Cash Value Added (CVA)

The CVA approach was popularised by Professors Eric Ottosson and Fredrik Weissenrieder of the Department of Economics, Gothenburg University, Sweden. Ottoson & Weissenrieder (1996) describe the approach as one which makes a distinction between the uses to which cash is put, being either strategic or non-strategic. The authors believe that a focus on strategic investments will give managers a better chance to form the future of their companies.

3.4 Practical Considerations
3.4.1 Nominal versus Real Valuation

One of the first decisions in DCF analysis, especially for valuations of companies outside the United States, is the basic decision of whether to do valuations in nominal or in real terms. In theory, using nominal cash flows and nominal discount rates will produce the same value, but when inflation rises above 10 percent, DCF valuations start falling apart in nominal terms because they become extraordinarily sensitive to small changes in assumptions. The solution is to use real valuation or perform the valuation in a more stable currency, for example the US Dollar (Damodaran, 2000:5).

3.4.2 Terminal value
A company’s value has two separate components. The first is the PV of future cash flows during an explicit forecast period, and the second is the PV of future cash flows beyond the explicit forecast period. This second component is called the company’s terminal, or continuing value.

The relative importance that a company’s terminal value assumes in a company’s total value is largely determined by the length of the forecast period contained in the DCF model. The forecast period is dependent on the forecast risk inherent to the industry. Recently, in an 8 year forecast period for a typical company in the tobacco industry, the continuing value typically represented 56 percent of the total value of the company; for sporting goods, the continuing value represented 81 percent. For a typical high-technology company, continuing value typically represented 125 percent of total value, which means that the PV of the FCF for the first 8 years of the company’s life would have been negative. All the value of the company would derive from the expected cash flow beyond the first 8 years (Lee, 2003:5).

3.4.3 Adjusting for Risk – The Cost of Equity:
Estimating cost of equity requires three inputs: the risk-free rate, the equity risk premium, and beta. For the risk-free rate, the common practice is to use short-term government security rates. For setting a risk premium, the standard practice is to use a historical risk premium. For beta, the most common approach is to regress stock returns against market returns. Unfortunately, all three of these practices can provide misleading answers.

Short-term government security rates

Since investors are looking ahead, they want to know what the ‘risk-free’ rate is over the length of time used for the forecasted cash flows.

For an investment to be risk free, the exact return over the life of the investment must be known. A 6 month treasury bill would not be risk free for the purposes of considering an 8 year cash flow. It would not be risk free because the investor would have to reinvest it at the end of each 6 months period at a currently unknown rate. In this case the appropriate risk free rate would be an 8 year zero coupon government bond rate (Damodaran, 2000:5).

Historical premiums

Many people, who use the CAPM to come up with the cost of equity, prefer to use historical premiums. They use historical data on share and bond returns over a past period; determine the average return made on shares and on a risk free asset over the period, and take the difference, which is the historical risk premium.

How the historical risk premium is computed matters. The result depends on several factors: what historical period is used and what type of investment is defined as the risk-free rate.

Historically, using as long a time period as possible is desirable, which is a little counterintuitive. Since the market risk premium is a random variable, a longer time frame (which encompasses a stock market crash, expansions, recessions, two world wars and stagflation) is a better estimate of the future than a short, more recent time frame (Copeland, et al. 2000:217).

A different way of thinking about the risk premium is to estimate an implied equity premium, i.e. letting the market decide what the premium is. If analysts have ready access to the following:

  • The current level of the ALSI;
  • The expected dividends next year for the entire index;
  • The expected growth rate in earnings and dividends over the next five years; and

If the growth rate is expected to be stable, the value of the share can be written as the basic Gordon Growth Model.

The only unknown in the above equation is the required rate of return. The required rate of return is the implied cost of equity. One of the advantages of the implied equity premium is that it can be estimated for any market in the world. Historical data is not needed (Damodaran, 2000:8).

Estimating beta

Beta measures the risk of a share relative to a diversified portfolio. Most services estimate betas for emerging market companies using the local index. These betas mean very little if the shares are to be included in a portfolio that includes non-local shares. To be meaningful, the beta of these companies ought to be estimated not against a local index but against a global index.

An even more troubling problem with beta is that the betas for emerging markets can be completely counterintuitive.

The largest and safest companies could look riskiest; the smallest and riskiest companies could look safest. This phenomenon is caused by index domination. The weighted average beta for all shares is one. In Brazil for example, Telebras (Telecomunicacoes Brasileiras) which comprises 50% of the Bovespa share index, recently had a beta of 1.11.In the index, eight other shares had betas greater than one and 169 had betas below one.

Some of the solutions to the regression beta problem are:

  • Choosing a more relevant index e.g. The S&P 500;
  • Using a measure called the ‘relative standard deviation’, whereby each company’s standard deviation of share prices is divided by the market average. A number for relative volatility that looks very much like a beta is produced; and
  • Estimating beta from the bottom up i.e. taking a weighted average of unlevered betas from other similar companies and levering up using the company’s debt to equity ratio. This approach offers greater precision by substituting the average beta for the sector in which the company operates for the regression beta (Damodaran, 2000:9/10).

Because accurately estimating the company’s cash flows and choosing the appropriate discount rate is difficult, DCF analysis is often abandoned in favour of valuation by multiples. Valuation by multiples entails calculating particular multiples for a set of benchmark companies and then finding the implied value of the company of interest based on the benchmark multiples. Although many studies incorporate this approach, no multiple is uniformly accepted as the one on which to base valuation (Lie & Lie, 2002:44/45).

Earnings valuation approaches capitalise a maintainable earnings stream at a multiple so as to arrive at an equity value. Earnings valuations techniques include market multiples such as Price Earnings, Forecasted Price Earnings, Enterprise Value over Sales, Enterprise Value over EBITDA and Enterprise Value over EBIT (Lie & Lie, 2002:46).

4.1 Price Earnings Multiples
The most popular earnings based valuation technique is the Price Earnings Multiple or P/E Ratio. The P/E ratio approach assumes that the company will be worth some multiple of it future earnings in the continuing period (Copeland, et al. 2000:284).

The P/E multiple is driven largely by the cost of capital and the growth rate of earnings. Holding the cost of capital constant, the P/E ratio is primarily a function of the expected growth rate in earnings. High P/E companies are expected to grow their earnings quickly relative to today’s earnings (Lee, 2003:9).

In its relative form, where firms are ranked on the basis of the ratio of P/E ratio to expected growth, the rankings will provide a measure of relative value, if:

  • The length of the high-growth period is the same for all firms; and
  • All firms are of equivalent risk (Sharpe, 1981:397).

If these assumptions are not fulfilled, a direct comparison of the ratio of P/E to expected growth may not be justified because a higher risk firm should be expected to have a lower P/E ratio than another firm with the same expected growth rate but much less risk. Similarly, the P/E ratio for a firm where high growth is expected to last 5 years will be lower than the P/E ratio for a firm where the same high growth is expected to last 10 years.

4.2 Advantages
There are a number of reasons why the P/E ratio is used so widely in valuation:

  • It is simple to compute for most shares and is widely available, making comparisons across shares simple;
  • It is an intuitively appealing statistic that relates the price paid to current earnings.
  • The multiple is simple to apply as a valuation tool; and
  • It can be a proxy for a number of other characteristics of the firm including risk and growth (Damodaran, 1996:291).

4.3 Practical Considerations
4.3.1 Multiple to be Used
Merrill Lynch & Company views the hierarchy of earnings as being shaped like a pyramid. From the top, sales tends to be more stable than earnings before interest, taxes, depreciation and amortisation (EBITDA); EBITDA tends to be more stable than earnings before interest and taxes (EBIT); EBIT tends to be more stable than cash flow; cash flow tends to be more stable than operating earnings per share (hereafter EPS); and operating EPS tends to more stable than reported EPS. As one comes down the income statement, earnings become more variable and unpredictable. Analysts should focus on the bottom of this pyramid i.e. reported EPS and look for signs of variability, even though the company would want analysts to focus on the top of the pyramid. Analysts should look at the variability of earnings, and variability is masked as one moves up the income statement because the trend as one moves up is toward greater stability. The cyclicality of earnings is not nearly as important as the variability of earnings (Bernstein, 2002:51,53).

4.3.2 Quality of Earnings
The “Quality of earnings” directly affects the valuation of a company and the accounting treatment associated with inter alia revenue and costs recognition directly affects the quality of earnings. Therefore to avoid accounting discrepancies under current GAAP, analysts must carefully examine the accounting choices a company makes when constructing its financial statements and assess the effect of those choices on its reported earnings quality (McConnell, 2003:33).

4.3.3 Comparable Firm
The definition of a “comparable” firm is essentially a subjective one. The use of other firms in the industry as the control group is often not a solution because firms within the same industry can have very different business mixes, risk and growth profiles. There is also plenty of potential for bias. To the extent that companies operate in several industries, the high subjectivity involved in choosing comparable companies could render this method impractical (Damodaran, 1996:304).

When markets make systematic errors in valuing entire sectors, using the average P/E ratio of these shares will build that error into the valuation (Damodaran, 1996: 291).

Alford (Lie & Lie, 2002:47) studied the effects of the choice of comparable companies on valuation accuracy when the P/E multiple is used and found that comparable companies chosen on the basis of industry yielded the smallest estimation of error when valuing companies by use of the P/E multiple.

4.3.4 Companies with High Intangible Value
Companies with a large part of their value in intangible assets, such as high-technology companies and companies with substantial research and development activities, may be particularly hard to value because a large portion of their value derives from uncertain future growth opportunities. A further complication with such companies is that their high research and development expenses reduce current earnings and in such a case, current earnings may be a bad predictor of value.

Dot-com companies are typical examples of the above as most of their value is tied to their potential to become major players in the future. Multiples are not suited for valuing these companies and analysts should resort to valuing the vast real options embedded in these companies’ assets (Lie & Lie, 2002:51/52).

4.3.5 Accuracy
Graham (1995:209) concluded that a large part of the discrepancies between carefully calculated formula values and their market prices can be traced to the growth factor, not because formulae underplay its importance, but rather because the market often has concepts of future earnings changes which cannot be derived from the companies past performance.

4.3.6 Comparisons of P/E Ratios
Comparisons are often made between P/E ratios in different countries with the intention of finding undervalued and overvalued markets. It is clearly misleading in these cases to compare P/E ratios across different markets without adjusting for differences in the underlying variables such as interest rate differentials and expected growth (Damodaran, 1996:299).

4.3.7 Losses
P/E ratios are not meaningful when the EPS are negative. While this can be partially overcome by using normalised or average EPS, the problem cannot be eliminated entirely (Damodaran, 1996:307).

Earnings and cash flows, though related, measure fundamentally different things. Cash flow measures inflows and outflows within a time period, regardless of the state of the operating cycle. Earnings measure inflows and outflows from operating cycles that the company has completed within a time period, regardless of when the cash flows occur.

Predicting the future from the past is always a risky thing. In the case where the company is neither growing nor shrinking the scale of its operations, cash flow and earnings reflect the same results. When the company is scaling up or down, earnings reflect the operating cycle that the company may replicate in future, while cash flows mix this operating information with information about investment and divestment (O’ Brian, n.d. :).

5.1 Caveats
Some of the pitfalls of using earnings to predict future cash flows are:

  • The company may not replicate its current business activity in the future. It may change strategic direction, pursuing new or different product markets;
  • Accounting numbers in most major economies are not adjusted for the effects of inflation. For short operating cycles in low-inflation economies, the time lags for most revenue and costs are not substantial, and therefore inflation creates only minor distortions; and
  • Earnings involve a great deal of judgment. Judgment enters for example in defining when operating cycles are complete and in identifying costs with sales (O’ Brian, n.d. :).

5.2 Earnings and Credibility
Cash flows are more credible than earnings because they involve less judgement. They are, on the other hand, less relevant than earnings in measuring replicable operating performance. This leaves us with the classic accounting conundrum: we want both relevance and reliability, but no single number provides both.

The best number for forecasting future cash flows is the one that credibly conveys the results of replicable operating cycles. Earnings measure the results of operating cycles, but involve judgement, which reduces their credibility. Cash flows involve less judgment, but do not measure the results of operating cycles (O’ Brian, n.d. :).

The P/E ratio and other earnings multiples, which are widely used in valuation, have the potential to be materially misused. This method, whilst not fundamentally flawed that it cannot be relied upon, has major shortcomings. A failure to adjust for differences in the fundamentals can lead to erroneous conclusions based purely upon a direct comparison of multiples (Damodaran, 1996:315).

Rarely will earnings and cash flow depict the same result. It is in these cases only that earnings can be deemed a surrogate for cash, and earnings based valuation techniques can be used in conjunction with cash based techniques so as to cross correlate results. In all other cases, the DCF method is superior to earnings based methodologies.

The value of a company is a function of the company’s cash flows, its expected growth, the length of the growth period and the cost of capital. Naïve attention to accounting earnings will often lead to value destroying decisions. Investors, who use the DCF approach to valuation, focusing on increasing long-term cash flow, ultimately will be rewarded by higher share prices (Copeland, et al. 2000:87).

BERNSTEIN, R 2002: Reported Earnings: The Ultimate Driver of Stock Performance. [On-Line]. Accessed on 18/ 7/ 2003. Available: http://www.aimrpubs.org/cp/issues/v2002n3/pdf/p0020049a.pdf

BLACK, A; WRIGHT, P & BACHMAN, J 1998: In Search of Shareholder Value. London: Pitman Publishing.

BREALEY, RA & MYERS, SC 2003: Principles of Corporate Finance; seventh edition. New York: McGraw-Hill.

COPELAND, T; KOLLER, T & MURRIN, J 2000: Valuation: Measuring and Managing the Value of Companies; third edition. NewYork: John Wiley & Sons, Inc.

DAMODARAN, A 1996: Investment Valuation. New York: John Wiley & Sons, Inc.

DAMODARAN, A 2000: Value Enhancement and Cash Driven Valuation Models. [On-Line]. Accessed on 22/ 7/ 2003. Available: http://www.aimrpubs.org/cp/issues/v2000n1/pdf/p0000004a.pdf

FRANCIS, J; OLSSON, P & OSWALD, D 2000: Comparing the Accuracy and Explainability of Dividend, Free Cash Flow and Earnings Value Estimates. Journal of Accounting Research, Spring. 2000.

GRAHAM, B 1995: Formula Valuations of Common Stocks. Financial Analysts Journal. January-February 1995.

LEE, CM 2003: Choosing the Right Valuation Approach. [On-Line]. Accessed on 20/ 7/ 2003. Available: http://www.aimrpubs.org/cp/issues/v2003n2/pdf/p0030004a.pdf

LIE, HJ & LIE E, 2002: Multiples Used to Estimate Corporate Value. [On-Line]. Accessed on 22/ 7/2003. Available: http://www.aimrpubs.org/faj/issues/v58n2/pdf/f0580044a.pdf

MCCONNELL, P 2003: Quality of Earnings: Avoiding the Accounting Landmines. [On-Line]. Accessed on 22/ 7/ 2003. Available:

O’BRIAN, P n.d.: Cash Flow versus Earnings.

OTTOSSON, E & WEISSENRIEDER, F 1996: Cash Value Added – A New Method for Measuring Financial Performance. Study No 1996:1, Department of Economics, Gothenburg University. [On-Line]. Accessed on 25/7/2003. Available: http://www.anelda.com/articles/e2.html

SHARPE, F 1981: Investments; second edition .New Jersey: Prentice-Hall, Inc.

STEWART, GB 1991: The Quest For Value. USA: HarperCollins Publishers Inc.