5.1 Accruals or cash flows?

In presenting the accounting approach, Penman starts step by step with the valuation of a risk-free bond. His goal is to adhere to the finance principle that value is based on expected cash flows. This principle, is good in theory, however, can be only poorly implemented. Decision-makers with limited knowledge and bounded rationality cannot forecast all cash flows expected over the life of the enterprise, and therefore, the use of free cash flows (FCF) does take us far away. In fact, once capital expenditures are subtracted from cash flows of operations, negative FCF are determined which are useless for valuation. Therefore, “[u]sing FCF in valuation is not only odd, it’s perverse” (p. 47). Accrual accounting helps, since accounting earnings allocate capital expenditures over periods of time according to the matching principle, and the same does for all of other cash flows. Summing up, value is based upon expected cash flows, but accrual accounting is the heuristic that the intelligent man can use in real life. In other words, one can build up the best and most elegant models he can think of, but to be useful they should take into account real life. Furthermore, they should include the real firm that runs the business over time that accruals are designed to represent and control.

Here, it is worth noting that Penman’s view is implicitly consistent with cost-based accounting. In his views, accruals in fact mainly refer to a historical-cost model, where the cost of an asset is reallocated over time against future cash flows. But why is accrual accounting better than DCF? A clear rationale for accrual accounting is not offered by the author, but some reasons can be inferred First, there is the practical problem. If one works with DCF he must estimate a terminal value, and such a value can impact on the final result of the valuation (p. 44). Again, DCF operates by subtracting cash outflows (i.e. investments), and this could produce a negative cash flow which is hard to interpret. This, Penman says (p. 46) “… is odd because investments are made to add value, not reduce it.” On the other hand, FCF increase when liquidating a firm. Accrual accounting, the author says (p. 48), anticipates future cash flows, which also take into consideration costs without cash outlays, as for employee stock options. His rationale for using accrual accounting mainly refers to problems with implementation of DCF more than to an alternative theoretical background. Even the way DCF and accrual accounting record stock options could be a self-serving example, if one recalls that not only out-of-pocket but also opportunity costs must be considered in the DCF model (Brealey, Myers, & Allen, 2006). Another issue to be considered regarding the supremacy of accruals over DCF is related to the risk of speculation and the particular perspective the author takes, since DCF seems to the author too open to speculation. Anyway, according to the author’s position, the problem is not the model per se, but the way it is applied.

5.2 Residual operating earnings and the cost of capital

Accrual earnings is the basis for valuation, but what kind of earnings should be used? Net earnings, Penman says (p. 90), are not adequate. They suffer, since they consider the effect of capital structure choices. It is a common issue that leverage can increase the return on equity. In order to offset the financial leverage effect, the fundamentalists should focus on operating income (Chapter 4). This has two advantages. First, it is free of the effects of financial decisions. Second, it specifically refers to the business; and business is what a fundamentalist should be interested in.

Another fundamental pillar is that earnings are not earnings at all! What really counts, Penman continues, are residual earnings, where “residual” means that they are the earnings from operations minus the cost of capital (or the minimum rate of return required) on net invested capital. Here is a potential pitfall of the book. Since valuation depends so heavily on residual earnings, the minimum rate of return required on net assets, r, should be carefully determined. Let us consider that residual earnings could be null or negative even in the case of positive accounting earnings, depending on the rate r. The mechanics Penman uses for the calculation of that rate consists of adding a risk premium to the long-term bond rate. Chapter 6 clearly addresses that point, but without suggesting any “how to do” theories. It is full of interesting ideas about how it is important to take into account uncertainty, but it says nothing on that point.

In the section “Finessing the Cost-of-Capital”, the author says: “… the fundamentalist approaches the market, not with a precise cost-of-capital in mind to challenge the price, but with the question: What is my expected return to buying at the current market price?” (p. 131). Again the reverse engineering can be applied, starting from market price and the accounting-based formula. Such reverse engineering, however, also reverts the variable r, which was an input data in the previous chapter and is now the output.

It is noticeable (see the quoted sentence and the rest of book, of course) that the author insists on the concept of “precision” as if he was willing to stimulate a subliminal comparison to “relevance”, but recognising that a few basis points in the cost of capital can dramatically change the calculated value. Penman emphasizes that sophisticated models do not protect against bad outcomes. But the reader could be worried that too much reverse engineering could lead to parametric solutions that make things more complex and undefined. In the basic formula Penman proposes, at least two variables, r, and the expected growth rate, g, must be determined. If one uses reverse engineering to test the growth rate implicitly embedded in prices, as the author suggests in Chapter 3, r must be determined as an input variable. On the contrary, in order to know the expected return to buy at the current market price, g must be set. Knowledge of the business could support the determination of g, beyond the possibility of using the GDP growth rate as a benchmark. In any case, one should have some model for determining one of the two, or both, and the lacking of such a model is clearly a problem (Biondi & Marzo, 2011).

5.3 Knowing the business

Knowledge is important in the author’s thought: “Know your business” appears to be his credo. But the way knowledge can be used for deriving numbers or models for valuation needs to be further investigated, since the author only rests on accounting figures to present this valuation approach. Knowledge is not the simply Bayesian updating of previous information. In fact, in the presence of interpretative ambiguity, people tend to use beliefs in order to clarify or complete information. This way knowledge grows autonomously from information, and knowledge (belief) is used to interpret incomplete information (Fransman, 1994). So knowledge is open-ended, while information is closed-ended.

Knowledge cannot be presented under a short formula. However, being silent on the kind of knowledge an analyst should use for valuation and on the way he would do it can potentially generate the risk to ground analysis on accounting numbers more than on business knowledge. The ad hoc cases Penman chose for supporting his theses are introduced and discussed on the basis of accounting figures more than on business knowledge. Demonstrating how business knowledge can help fundamentalist would have been more fruitful for the reader. Resting on accounting could only generate some confusion, as throughout the book Penman advises the reader that accounting truth is not the “real” truth one would expect to know. After all “… residual operating income is an accounting measure, so it depends on how the accounting is done” (p. 104).

Moreover, the author often refers to aggregate data more than to firm-specific knowledge. For example, he recalls that rates of return are expected to revert towards the mean in the long run, when the competitive advantage period of a firm erodes. Simulation through reverse engineering is a way to challenge market participants’ estimates. But when will such a long run be achieved? The “long run” is a rhetoric concept more than a date. “In the long run, we are all dead”, Keynes said. Nevertheless, for those that in the long run will be still alive, what average growth rate may they find?

Clearly, the main concern of the author is moving away from speculative value which is not grounded on what you know but on speculation (Fundamentalist Principle no. 7). His defensive approach is to be cautious and prudent. Nevertheless, such an approach could generate some counterintuitive effects. For example, the author’s defensive focus is on short-term earnings more than on long-term growth. However, relying on short-term earnings may not be the right thing, it they were abnormal by chance. Knowledge should, therefore, help in the difficult task of discriminating between the various conditions. The way forward is to be discovered case by case.

5.4 Growth, speculation and risk

A great concern of Penman is about speculation. He is against speculative value, or to say it better, he criticises all valuation methods (developed within the finance theory) that push valuators to consider growth as being always possible and, in any case, free of risk. At least three Chapters (Chapters 4–6) focus on growth as a potential risk driver. Penman carefully argues that many kinds of growth exist, not all of which add value. Only growth that increases residual earnings adds value, and in any case at some price: a larger risk. In particular, he is interested in demonstrating that the apparent growth determined by financial leverage is unreal. Growth should be one of the first concerns for fundamentalists: Beware of paying too much for growth, states the Fundamentalist Principle no. 8.

That growth adds risk is, however, controversial. While growth usually calls for investing new resources at some degree of risk, a larger size can help companies to make their results more stable. Some competitive advantages, in fact, are clearly determined by the large size of a firm. The point could be somewhat reframed considering that growth calls for funding, and if new debt is used, then the firm is potentially riskier than before.

How to make a distinction between bad and good growth is still a mystery, however, as Penman does not clarify this point. It could be argued that considering growth as always bad is not so different than considering it as being always good. Some suggestions would have been useful in order to make such a distinction. Again, here is the echo of the prudent approach a fundamentalist should implement. Growth modifies things, and such changes could lead us far from the actual results. The short termism a fundamentalist must adopt is clearly consistent with the suspicion the author feels against growth.

5.5 What do losses imply?

Penman’s conservative approach also reflects the accounting model he proposes. Such a model is a kind which deserves much more attention to the peril of paying too much than to the opportunity to have a great deal. This probably leads to unanswered questions with respect to the valuation of firms which do report losses. How these firms could be valued is not fully explored. If one should use accounting income in order to focus on the short term, he could end with the conundrum of finding out a value when income is negative. The two points seem completely at the odds in Penman’s approach. Caution for paying too much for growth clearly translates into the fear of paying too much for a firm reporting losses. The two situations are clearly similar. In general, one could oscillate between two poles. One could maintain the basic accounting model the author develops and apply it to forecasted figures. This would place him in some trouble with respect to the fundamentalist principles, although would give him what he was looking for. On the other hand, he could decide to fully adhere to the fundamentalist principles, so searching for other investment opportunities.

5.6 Looking for residual earnings

Chapter 5 deals with the kind of growth, in residual earnings also, deriving from the way accounting is done. This chapter still underlines the particular position of the author on the assumed objectivity of finance with respect to the subjectivity of accounting. Accounting numbers do not tell the truth. Truth does not exist by itself, of course; and accounting creates its own reality as for the results a firm has obtained. Accounting figures are “real” only within the particular set of accounting principles and standards used for their determination. This means that, firstly, one should not consider accounting figures as truth; secondly, one should think of what good accounting is like.

Penman compares conservative accounting to non-conservative accounting and argues that a simple modification of the valuation formula can offset the differences generated by the two systems. The readily available remedy, however, does not make the author uninterested in what a good accounting should be like. Chapters 8 and 9 deal with this issue.

6.1 Against fair-value accounting

Recent years witnessed the rise and fall of fair-value accounting. Welcomed as the best tool for an investor’s sound decision-making (Hitz, 2007), it has been under attack for its procyclical and perverse effects on crisis (Bignon, Biondi, & Ragot, 2009).

Clearly, comparing fair-value to historical-cost accounting seems a battle already lost. In economics and in finance, words are value-laden (Frankfurter & McGoun, 1999), and they exercise a relevant power on accepting ideas. So, how can you dare comparing “historical cost” to “fair value” and still continuing to think that what is “old”, “antiquated”, “outmoded” is better than what is “impartial”, “rational”, “unbiased”, “just”? Despite such appearance, fair-value accounting is far away from good accounting, Penman argues. His position is supported by a list of criticisms against it, which find their reasons in the market and the firm functioning.

Fair value finds its justification on Efficient Market Hypothesis (p. 170). “Fair value accounting pulls the information in prices into the accounts immediately”, Penman states. Nevertheless since markets seem to be not as efficient as theory assumes, fair value itself should be not so fair. The fundamentalist principle no. 9 preaches: “When calculating value to challenge price, beware of using price in the calculation.” Clearly the use of fair-value accounting is at the odds with the fundamentalist Decalogue. If the fundamentalist goal is challenging price, he cannot rest on price, he needs information which is different than that embedded in price.

From this perspective, fair value increases the systematic risk due to feedback loops that can add crashes to crashes, he says (p. 177). When fair-value accounting enters financial statements, the risk borne by investor increases. Two reasons support this sentence. First, “leverage … is no longer leverage against hard assets, but leverage against expectations.” (p. 178). Fair value puts ambiguity in a firm’s performance. Since fair value embeds expectations, leverage is no longer against hard assets, but against expectations. The reader would immediately agree in the case an increase of fair value would increase the value of firm’s assets. In this case, in fact, a large part of debt would be secured by soft expectations more than by real hard assets. But Penman argues that it happens also in the opposite case, when fair value falls, if firm’s debt is fair valued on declining estimates on credit worthiness. The reported firm’s position clearly deteriorates.

A comment could be here added, considering that the calculation of leverage based on market values is the typical approach adopted by finance and that the fair valuation of assets (according to the three levels of fair value) echoes that perspective. It has been noted (Barclay, Morellec, & Smith, 2006; Marzo, 2012) that adopting the market value of equity for the calculation of the firm’s financial leverage can produce ambiguity and perverse effects. A value-based perspective in the calculation of that ratio is closer to the way value can be distributed among financial stakeholders (in a liquidation) more than to the sustainability of debt (in a going concern) (Marzo, 2012). Penman offers a similar view.

Moreover, fair-value accounting invites distribution of unrealised earnings. Fair-value accounting generates “paper” profit, i.e. profit that is still unrealised although recorded. Expectations can artificially inflate profits, so stimulating the distribution of unrealised value. A possible solution could be to create side pockets until uncertainty on those recorded profit will be resolved (p. 179), but in Penman’s view the right solution is to abandon fair-value accounting in favour of historical-cost accounting.

Criticisms against market efficiency apart, Penman also underlines that fair-value accounting serves self-interested managers (p. 173). Fair value, with its three levels of implementation, allows managers to put their estimates in financial reporting. Of course, the expectations of managers are important for all of those interested in a firm’s valuation. However, those expectations should be left apart from accounting, not embedded in it.

One could agree or not agree with the above remarks. After all it seems that much of the debate on market efficiency rests on a question of faith more than on question of facts. A sounder reasoning is, however, proposed by Penman, on why historical cost, and not fair value, is more suited for firms. “As exit prices, fair values are liquidation values, typically reported when a firm is failing. Going-concern value is quite different; the exit price for an on-going business is the price it can sell its product for after adding value through the business process – revenues in the income statement” (p. 176). Fair-value accounting is not able to correctly identify margins gained by an enterprise, just because those margins uncover themselves only when selling prices are compared to historical cost. The fair-value adoption generates anticipation and deferring of revenues and expenses which make it difficult to understand which is the current profit realised by an enterprise.

The point is a good one, since it refers to the economics of the firm. While all of the other criticisms against finance and accounting standards draw on the role of either the inefficient market or the rationally bounded man, this insight is strongly linked to the way one expects a firm to generate value. Traces of a similar position, but with some ambiguities, can be now found in the IASB proposal of adopting fair value according to the business model of the firm.

Penman’s position is clearly addressed. In a few words, fair-value accounting is not adequate for the going concern, since its use denies the going concern. If you are interested in valuing a going concern, fair value is exactly what you do not need. Historical accounting, says Penman, is what an investor needs. Historical accounting does not mix facts with expectations. It is against speculative value, since it records earnings only when realised. Finally, it adheres to the way value is generated by the going concern. The example Penman presents (p. 175) is clear enough to understand the paradox the fair-value accounting can generate. If a steel maker firm has a pile of coal, would it get better if the coal price increased? Fair-value logic would suggest yes, since the fair value is now higher than before. But, Penman argues, fair value is an exit price, i.e. a good reference point for whom selling coal. The steel maker, instead, uses coal in its production process, and a higher price means higher costs in the future, and therefore lower profits.

Here, a different logic is presented by Penman as for the conception of firm. The conception of firm at the core of FASB/IASB is similar to a nexus of contracts signed with market more than to a system of resources and people. In the first case, the value of the firm is the sum of the value of its assets, this being simply their market value.

According to the different conceptions of the firm Penman evokes, the value of any asset has to be determined in relation to the system of assets and resources it belongs. Paradoxically, a specific resource supporting a firm’s value generation could have a null value for the market (Denrell, Fang, & Winter, 2003). Economics of firm is different than the economics of market (Biondi, Canziani, & Kirat, 2007). Fair value neglects this profound difference, and it can generate bad effects, without supporting investor decision-making. Only firms for which the one-to-one principle holds should adopt fair value. That principle indicates the situation when the shareholder value moves one-to-one with the market price. This is the case of speculator, for example, or share trader.

6.2 In favour of historical-cost accounting

Once fair value has been criticised and identified as the potential source of problems and ambiguities, historical-cost accounting rests the more suitable alternative. Penman is aware of the criticisms many authors have addressed against historical accounting for it being backward looking and for leaving some important assets (the intangible ones) outside the balance sheet; therefore, he tries to offer some considerations to support historical-cost accounting.

Providing that there is no growth in unrecorded intangible assets, the omission of value in the balance sheet is completely offset by the omission of value at the end of the period, and therefore the error cancels, and the stock return can be calculated as the earning plus the change in the price over the book value for the earning period (p. 180):

Therefore, “… it does not matter if intangible assets are missing from the balance sheet if earnings from those intangible assets are flowing through the income statement” (p. 181).

However, such an approach risks generating some problems, for three reasons at least. First, the way Penman supports historical-cost accounting may generate some anxieties against it, since the cancelling error principle only works for firms with no-growth assets. This is certainly the case of some firms, but it cannot be generalised. One could eventually accept little un-cancelling errors for firm with growing assets, but perplexities still remain.

Second, the un-recording of some assets (e.g. intangible assets) comes more from the conservatism principle more than from historical-cost accounting. One could still base his accounting on historical cost but record intangible assets at their cost, providing the usual amortisation process. This possibility would open the door to the introduction of assumptions about future in the financial statement, but still resting on historical cost. In order to prevent the distribution of unrealised earnings, the distribution of the part of earnings generated by the capitalisation of intangible assets could be forbidden.

The third reason refers to the examples Penman uses for illustrating the cancelling error principle. Microsoft and Dell appear to be two valid cases, since one could argue that if the principle holds for such intangibles-intensive companies it can be universally employed. The two examples suffer, however, from putting a required return without any specification of why the number employed is the right one. In the case of Microsoft, Penman adopts a return of 9% (while it is 10% for Dell). The Microsoft equity value (in the fiscal year ending June 2008) is the sum of enterprise value and the value of available cash ($23,662 million). Enterprise value can be determined summing the book value of net operating assets ($12,624 million) with the discounted value of the residual income expected in 2009. This latter addend is calculated discounting as a perpetuity the excess of 2008 operating income (used as a proxy for the operating income expected in 2009) over the cost of capital, calculated employing the 9% rate against a net operating capital of $12,624 million. The presented value of such a residual income is performed at the 9%. The final value is $210,718 million, or $23.03 per share, which is very close to the $25 per share at which Microsoft was traded during that period.

Nevertheless, a perplexity arises. Assuming that the robustness of the calculation is measured by its closeness to the market price (I will come back on this point later on), the obtained result is determined in large part by the rate of return Penman adopted. With a rate of return equals to 5%, 7%, 11% or 13%, the value per share would be $39.38, $28.87, $19.31 and $16.74, respectively. How can we be sure that the rate of return Penman adopts is the right one?

The role of the rate of return is really the focal point of the model (and this commentary focused on the problem of its determination also in Section 5). Different rates of return give different results. How choosing the one to which investor decision-making can refer?

Paradoxically, the path Penman opens is not acceptable. Referring to the result of his calculation, Penman says: “… although considerable value is missing in the balance sheet, the accounting that includes earnings explains almost all the value that the market sees in its $25 price.” In order to choose the right rate of return, should the reader rest on the convergence of the calculated value with the market price? A positive answer cannot be given to this question except in the case where one desires to abandon fundamentalist principles Penman has though throughout the book. Fundamentalist must challenge market price, and doing so he must beware of using price in calculation. Now price is outside the formula, but it enters the result as a reference point. Using price as a control for his own valuation seems at the odds with the fundamentalist approach.

Moreover, fundamentalists should feel worried if the value he calculates falls very close to the market price. In this case, in fact, there would be no stance for arbitrage nor for profit.

Really, the reader is left without any model that could help him finding out the rate of return he can use for challenging market price.

6.3 A renewed interest in income statement

The second criticism addressed towards accounting standards is related to their focus on the balance sheet. Accounting standards setters have pushed the idea that only the balance sheet matters. This is because fair value seems so useful. Fair-value application reduces the gap between market and book value, so generating the illusion that financial statements are now reliable, since they report the true value of the firm, the one formed in capital markets. The Market-to-Book ratio should not be thought of as a gap that should be closed. Nevertheless, many researchers clearly focus on that gap as a symptom of inadequate financial statements. For instance, many researchers interested in Intellectual Capital and intangibles share the idea that the gap between market and book values should be filled in. “A sign of the loss of relevance of accounting information is the increasing gap between the market value and the book value of equity of companies in financial markets” (Cañibano, García-Ayuso, & Sánchez 2000). But market value refers to the firm as a whole, whereas the balance sheet shows the value of single assets (Walker, 2009). Moreover, many other reasons could be listed for that gap to remain, according to the theory of the firm one chooses to refer to (Marzo, 2013).

As Penman says “… there is also an income statement and accounting for value employs both the income statement and the balance sheet” (p. 180). Once the importance of the income statement has been recognised, searching for good accounting does not end either in fair value or in the balance sheet. What an investor needs, Penman argues, is a renewed interest in the income statement.

This position is very close to the Italian accounting tradition developed at the beginning of 1900s by Gino Zappa, the founder of Economia Aziendale (Zappa, 1937; Biondi, 2002; Canziani, 2007). According to his vision of the firm as an entity, the income statement is more important than the balance sheet. Income is the results of the system of resources that identifies the firm. Balance sheet is an operational way for suspending costs from the determination of the income of the period, since they do not find in the same period their matching revenues. Therefore, the balance sheet plays an ancillary role with respect to the income statement. Income expresses the dynamics of the entity-firm as a coordination of revenues and costs. It is not income deriving from capital, as in the Hicksian approach, but it is the value of capital coming from income.

Along a similar vein, Penman argues that “Individual assets and liabilities cannot have stand-alone fair values, nor can the sum of fair values express the value of using assets together” (p. 186). Working on the balance sheet only will add little value to accounting. The income statement calls for being considered as an important tool for valuation.

6.4 Adding value to accounting

Now that the position of the author with respect to the two important topics of fair-value versus historical-cost accounting and to the role of income statement has been presented, a final question arises: How to add value to accounting? Chapter 9 is devoted to answer it.

Penman articulates his thesis around two steps. First, he demonstrates that investing in good accounting is still useful. Finally, he proposes some ways for improving accounting.

Penman presents two studies showing that markets naively use accounting information. Besides some features of the human beings that trap them in habits and routines, this is probably due to the fact that actual accounting is not so very useful for investment decisions. This is why adding value to accounting is possible and also due.

According to Penman’s view, the nowadays practice of accounting has become a matter of mere definitions of accounting phenomena and application of those definitions. “Foundation principles are important, of course, but determining accounting on the basis of how it conforms to accounting definitions can lead to accounting standards that read much as a thesaurus: accounting references themselves.” Penman argues (p. 192).

What would really help, Penman says, are at least five principles referring to both the balance sheet and the income statement. “First, the balance sheet that anchors the valuation must be ‘hard’. … Second, the income statement must anchor the forecasts of near-term earnings … Third, both the income statement and balance sheet must distinguish between operating activities and financing activities. … Fourth, conservatism applies … Fifth, ‘below the line’ disclosures focus on information, excluded from accounts, that assists investors in their speculation” (pp. 196–198).

Application of the five principles to both the balance sheet and the income statement are very interesting. Here, some examples are given.

According to Penman’s view, the balance sheet is not as close to the shareholder’s view as it is supposed to be. First of all, it does not clearly distinguish between debt and equity nor does it report the total cost of borrowing, as in the case of convertible debt. Second, the balance sheet does not provide a clear distinction between financing and operating activities. The short versus long-term perspective the balance sheet actually supports is of some interest for creditors, but not for shareholders.

Historical cost focusing on transaction with impairment of the carrying amount in case of expected losses makes the balance sheet harder, Penman says. Finally, referring to transactions (the basic unit of analysis of historical-cost accounting) hinders managers from putting their self-interested expectations in the balance sheet.

The regained relevance of the income statement also calls for thinking about how to improve it. Despite the common wisdom that the balance sheet is deficient, Penman argues that the income statement is, because the five or six line items it is made of are not enough. A lack of transparency is, therefore, at the stake.

Moreover, revenues recognition under GAAP is not particularly adequate for complex arrangements in which many companies are involved, thus giving managers large room for manoeuvring. A transaction-based approach can help as a remedy. This way both the balance sheet and the income statement are made harder.

Accounting rules should clarify the sustainability of profit margins. Now, on the contrary, managers have the possibility to write down and restructure on their arrival just to report higher profits later.

Conservatism is important in Penman’s view. However, he says, “Excessive conservatism – sometimes called discretionary conservatism – is as much a problem as no conservatism at all” (p. 205). Excessive conservatism is, however, difficult to contrast. Therefore, a couple of solutions could be adopted. First, a “quality of earning” statement of management could help to assess the effects of accounting choices. Second, write-down could be capitalised and then amortised over the next periods in order to smooth income.