1 The theory of the firm

In developing my arguments I shall rely on Williamson’s work, which also forms a departure point for Brian’s exposition of his main thesis, namely, that business models provide a suitable basis for regulating financial reporting. The Williamsonian transaction costs approach states that firms emerge only when internal organization of activities is superior to transacting in the market. Organizing internally is advantageous whenever and only if doing so adds value to the firm’s owners. From this it follows that any firm in existence, assuming rationality of its owners, provides value added. In other words, there is a specific advantage to being a firm rather than transacting in the marketplace. This in turn implies, contrary to Brian’s proposition regarding the existence of two types of firms, type 1 and type 2, that there must exist only one type of firms: a firm that adds value by transforming assets—by transforming inputs into outputs.

In Brian’s words, borrowing from Penman’s distinction between two types of economic activity by firms, type one transforms inputs and add value to them whereas type 2 holds items that it does not transform and to which it does not add value; this type 2 form simply buys and sells the same assets at market prices (what goes into the firm is the same as what goes out). Nonetheless Brian admits the so-called type 2 firm in fact adds value, for example by providing asset management services. Their internal processes include asset selection, monitoring performance, etc.

In spite of this admission, however, Brian goes on to claim that “there is something significantly different about type 2 firms to the extent that what they are doing is holding assets rather than transforming them.” Then he states, in agreement with Penman, that this distinction between type 1 firms and type 2 firms requires financial reporting information to be used in different ways. Namely, for type 2 firms, where the assets are at market prices, the balance sheet shows the value of the firm. For type 1 firms,however, historical costs are the appropriate measurement principle, allowing investors to use the income statement as a basis for forecasting future income, and hence, by applying an appropriate multiple, to value the firm.

My disagreement with this conclusion begins with the premise that there exist two types of firms. Once we subscribe to the Williamsonian theory, we can see that no firm will emerge unless it adds value by transforming assets. Therefore, there is no room under this theory for any type 2 firm. No firm would be viable unless it transforms inputs into outputs; simply buying and selling the same assets at market prices can be accomplished more cheaply and efficiently through market transactions: there would be no need to conduct such activities within a firm that inevitably imposes costs of organization and hierarchy.

Let us examine this more closely by considering Brian’s (or Penman’s) example of an asset management firm, which, on the face of it, appears simply to buy and sell financial assets with no transformation taking place. But this first impression must be fallacious: as mentioned, with no transformation there would be no value-added and no justification therefore for any firm. There must be some transformation taking place. We should then be able to identify inputs that are transformed into outputs. A simple reflection suffices to make this identification. Two kinds of inputs are involved here. The first is the very large set of markets-priced financial assets that exist in the economy, and the second is an intangible: the specialized financial and valuation know-how that the firm’s management processes. The transformation process that takes place consists of using the specialized know-how to pick from among the vast array of financial assets a subset—a portfolio—that promises clients superior returns relative to a given risk profile. The value added in this case is reflected in the combining these two inputs into a desirable output, i.e., the creation of a portfolio of financial assets that increases the expected utility of the client beyond what it would have been without the service provided by the asset management firm.

It may be countered that the client could simply purchase the individual assets contained in the portfolio in the marketplace without the services of the asset management firm. This is not feasible however. Without the financial know-that the firm’s management brings to bear on picking the “right” individual assets that should comprise the portfolio, the client cannot replicate what the asset management firm does. He simply does not know which individual financial assets to purchase. Put differently, the specific advantage of the asset management firm in this case is the ability to select assets that in combination satisfy the return-risk preferences of the client and hence add value (in the sense of expected utility). This value-added justifies the fees that constitute the asset management firm’s revenue.

Indeed, intangibles (such as the specialized financial know-how of the asset management firm) are explicitly recognized by Williamson as specific assets that explain formation of firms:

Whatever the unit of analysis, operationalization turns on naming and explicating the critical dimensions with respect to which the unit varies. Three of the key dimensions of transactions that have important ramifications for governance are asset specificity (which takes a variety of forms- physical, human, site, dedicated, brand name-and is a measure of bilateral dependency), the disturbances to which transactions are subject (and to which potential mal-adaptations accrue) and the frequency with which transactions recur (which bears both on the efficacy of reputation effects in the market and the incentive to incur the cost of specialized internal governance). Given that transactions differ in their attributes and that governance structures differ in their costs and competencies, the aforementioned-that transactions should be aligned with appropriate governance structures-applies.

The key factor here is whether the transaction in question is supported by investments in transaction-specific assets. Such specialized investments may take the form of specialized physical assets (such as a die for stamping out distinctive metal shapes), specialized human assets (that arise from firm - specific training or learning by doing), site specificity (specialization by proximity), dedicated assets (large discrete investments made in expectation of continuing business, the premature termination of which business would result in product being sold at distress prices) or brand-name capital. (Emphasis mine).Footnote 1

According to Williamson, the governance approach to describing a firm (as contrasted with mechanism design, agency theory, and property rights theory) appeals to law and organization theory in naming incentive intensity, administrative control and contract law regime as three critical attributes. Coase’s theory of the firm suggests that organizing a firm economizes on the cost of price discovery because accounting measures can be used for transfer pricing or for transferring a good or service. Under a firm’s hierarchy, incentive intensity is less, administrative controls are more numerous and discretionary, and the internal dispute resolution supplants court ordering. Adaptation is taken to be the main purpose and the need for coordination builds up as asset specificity deepens. The use of credibility is required to support those transactions when asset specificity and contractual hazards are an issue. Such use is not required where the general purpose production technology is employed.

Examples of asset specificity extend to brand-name capital (Klein 1980), the integrity of which can be compromised. Essentially, the firm is an idea and the implementation of an idea becomes a viable firm if and only if the idea creates specific assets. The idea is typically described by a “business model”. The business model lays out the nature of the specific assets and the specific transactions that are envisaged to carry out the objectives embedded in the idea. Along with transaction and asset specificity contractual complexity inevitably ensues (Williamson 1971, 1979, p 239).

Going back to the asset management firm example, what is the idea? It is to construct asset portfolios that enhance the expected utility of customers. What is the specific asset? It is the human capital—the intangible—enabling the construction of optimal portfolios. Clearly, the contracts that regulate this activity delineate the rights and obligations of management and clients and specify remedies for recklessness or bad faith perforce would have to be complex. At the conceptual level, this example does not differ significantly from a manufacturing operation. In the case of manufacturing, specialized machinery may substitute for specific human capital, and internal administrative controls and incentive mechanisms would provide much of what otherwise would have required complex contracts. Thus, at this level of abstraction, there are no differences between manufacturing and asset management. Both transform inputs to outputs; it is only the nature of these inputs and outputs that differs across the two organizations. In the case of manufacturing inputs are raw materials, labor, and other services and the output is a saleable product. In the case of asset management the inputs are traded securities and the specific human capital of management manifested in the ability to select securities, and the output is portfolios constructed for different clients.

2 Implications for financial reporting principles

A profound implication of this essential similarity between the two kinds of firms (such as manufacturing and asset management) is that accounting for these two businesses or any other businesses for that matter should not be based on different principles. Specifically, recognition and measurement principles should be invariant to the business model employed. Accounting should reflect the transformation of inputs to outputs in such a way as to best meet the objective of helping stakeholders predict future cash flows, their timing, and associated uncertainty. Hence, in disagreement with Penman, I would argue that the need for fair values (specifically, exit values) and historical costs arises with equal force with respect to both firms. In other words, if historical costs are seen as the preferred measurement principle, they should be used in both kinds of firms, and conversely, if exit values are deemed to be the solution, they should be the basis for both kinds of firms as well. I argue below, however, that a compelling implication of the theory of the firm is that both measures should be provided for each firm.

If we adopt a governance perspective of the firm, one that is based on asset and transaction specificity, mutual dependency among players, and potential for conflicts and contractual disturbances, we can make reasonable inferences about the financial reporting elements that meet the accounting objectives. Since mutual dependency and contractual disturbances as well as the potential for conflicts can be viewed as byproducts of assets and transactions specificity, it is best to focus on the latter to guide financial reporting principles.

Asset specificity points to two distinct useful-to-stakeholders measures of value. The first is use value—discounted expected cash flows when the asset is employed as intended by management, i.e., in accord with the business model. The second is the exit value—the proceeds if the asset is sold in the outside relevant market. The former is the measure of value to shareholders—their expected benefits from the firm employing the assets in its intended use. The second is a measure of what the firm can recover if things go wrong; it also reflects the opportunity cost of operating the assets within the firm. And while discounted cash flows reflect expected benefits—and ultimately expected returns to stakeholders, exit values, when juxtaposed against discounted cash flows reflect risk. In particular, the difference between discounted cash flows and exit values quantifies the risk of business models failure: the amount by which expected benefits decline if things go wrong. Changes in exit values over time would thus reflect changes in risk: increases in exit values relative to discounted cash flows result in decreased risk, and conversely, decreases in exit values result in increased risk. Consequently, the two measures in combination describe the two sides of transactions: expected return and risk—the two primary parameters investors need for selecting portfolios of securities. This is not to say, of course, that the two measures capture everything one needs to know about risk and return. (One can imagine for example the provision of probabilistic financial statements, i.e., one set of financial statements for every sequence of cash flow realizations contingent on employment of the firm’s resources when a well-defined set of states of the world (event) occurs. This would require as many financial statements as identified events with a probability measure attached to each one of these. But such a radical departure from the existing accounting model today seems like a futuristic fantasy.

How do business models enter into this picture? What is their role in financial reporting? It should be apparent by now that the formation of expectations about future cash flows depends on the business model. In essence, the business model is a set of strategies and tactical moves that management devices in order to meet its objective. Thus, the discounted cash flows measure would be derived from projections based on the expected outcomes of the plans envisaged in the business model. In other words, recognition and measurement principles (such as the principle of providing both discounted cash flows and exit values) should not depend on the business model. Rather, it is the quantum of expected cash flows that naturally depends on managerial plans embedded in the business model.

Invoking again the theory of the firm, we may now ask in what way are the two proposed measures, discounted cash flows and exit values responsive to the assets specificity view of the firm? Being mindful of the fact that greater specificity spells higher uncertainty, to predict the potential risk and return associated with specificity, stakeholders would benefit from information about the degree of specificity and the associated risk. This information would be reflected in the juxtaposition of discounted cash flows and exit values. The larger the gap between discounted cash flows and exit values, the greater the asset specificity and the associated risk that could materialize if management’s plans embedded in its business model go awry. Compare for example a retail firm having no significant intangibles with an intensive research and development firm whose main (specific) asset is the expertise of its researchers. The first would have a small-gap between the exit values of assets and discounted cash flow: most of its assets (such as inventory and real estate) have universal uses that command relatively high exit values that do not differ significantly from the assets’ discounted cash flows (the use value). The small difference between exit values and discounted cash flows will signal low specificity and correspondingly lower risk. By contrast, the research and development intensive firm will feature large discounted cash flows relative to the exit values of its assets, signaling a high degree of specificity (and hence expected return) and higher risk.

Do historical costs play any role under the theory of the firm? The answer must be yes. The demand for historical costs makes its appearance at the governance phase, that is, the governance of ongoing contractual relations (contract implementation). Once the business model has been laid out, implementation of contractual activities begins. Governance—the set of incentives, administrative control, mechanisms for resolution of conflicts, and monitoring is then required to minimize contractual disturbances and to assure adherence to goals on which the business model is premised. Historical costs, perhaps better described as the reporting of transactions’ realized prices (purchases or sales) then come into play. Realized transactions prices are necessary to provide measures of actual performance and of whether management plans resulted in the anticipated outcomes. Thus, the juxtaposition of historical performance with discounted cash flows tells us to degree to which management was successful in carrying out its plans according to its business model. It provides information about the actual return, whereas discounted cash flows reflect expected returns. But realization prices provide more than that. They also allow stakeholders to assess the forecasting ability of management—to what degree they were correct in forecasting future cash flows contingent on their plans. Our current historical cost accounting system provides measures of factual events and transactions, but fails to calibrate the degree to which management was able to implement its plans. Nor does it enable stakeholders to form a judgment on the reliability of management’s forecasts given the absence of a record of deviations from forecasts. The latter is possible only when both forecasts (discounted cash flows) and actual realizations are provided.

Consider the asset management firm again. Its specific asset is its management’s specific advantage reflected in the superior ability to pick securities. While the exit value of its assets would be the market value of the chosen portfolio, the discounted cash flows would reflect management’s expected return over a horizon during which management will use its superior ability continually to pick winning stocks in the future. The gap between discounted cash flows and exit value with reflect the risk of management not being able to pick winning stocks. At the same time, the cumulative actual returns since the purchase of stocks comprising the portfolio that are reported in accordance with the historical cost principle provide the measure of factual performance which can be compared with the forecasted return inherent in the discounted cash flows. These actual returns (the difference between input prices and output prices) would also provide stakeholders with the independent ability to evaluate management’s forecasts embedded in the discounted cash flows.

In the case of the manufacturing firm, as well, realization prices (historical cost principle) are needed to provide actual differences between input and output prices as a check on the forecasted returns embedded in discounted cash flows. Thus, for both manufacturing firms and those that (superficially) only buy and sell the same assets, an ideal accounting system comprises the three measurements: discounted cash flows, exit values, and realization prices.Footnote 2

To sum, business models should not dictate financial reporting measurements, but the quantum of one of these measures: discounted cash flows does and should depend on the business model. The willamsonian theory of the firm irrefutably leads us in the direction of providing these multiple measures Independently of the business models.Footnote 3 Just as the life of an individual cannot be dramatized by displaying his past alone, or present alone, or future aspirations alone, the evolution of a firm and its prospects cannot be characterized by its past (historical costs) alone, Present (exit values alone) or future plans (discounted cash flows) alone; all three dimensions contribute in combination. Arguments that investors would be confused by a plurality of measures cannot survive scrutiny. Once users are made aware of the different meanings and implications of each of the measures, they would recognize the benefits afforded by the richer menu.

The inherent subjectivity of business models reinforces the inappropriateness of having them dictate financial reporting principles. I agree with Mike’s characterization of business models presented as tools for propaganda. The problem this presents is the difficulty of checking the veracity of a non-quantified “business model” narrative. An attempt to do so would normally hinge on the comparison of depictions of the business model with what eventually transpires. Unfortunately, however, unexpected states of nature interfere: how to distinguish between prevarication (and overly optimistic intentions) and post-statement events that interrupted the implementation of plans? Of course, if such states of nature are random one would expect average (over multiple periods) deviations from the narrative to become more clearly attributed to prevarication as the bias introduced by unexpected events washes out. Even then, however, a major obstacle remains: how can one compare verbal narratives (such as of the kind that Mike illustrated) with later narratives of what transpires? Despite the advances of linguistics and computer technology, I do not believe comparisons among narratives can lead to reliable judgments. I conclude, therefore, that business model narratives can be verified only if quantified. Folding expected outcomes of the business model into discounted cash flow measures is one such quantification.

A thorny obstacle to implementing my proposed financial reporting model is the potential bias in management’s forecasts (needed to derive the discounted cash flow measure). This can be tackled by creating incentives for management to tell the truth, i.e., to provide their genuine and honest forecasts given their prior information. But this is a subject for a different conference, maybe the next one in Siena?

3 Implications for financial reporting regulations

Mike has touched upon the different views of “value”: total economic surplus generated by the firm’s activities versus shareholder value. This distinction is parallel to one concerning the objectives of financial statements. Are financial statements designed to provide information that promotes efficient allocation of resources in the economy, or to inform investors’ decisions so as to maximize the value of their investments? The two formulations of the objective, while not mutually exclusive, have different implications for the content of financial statements. And, indeed, the ambiguity as to which of the formulations guides the FASB’s and the IASB’s policy deliberations lies at the core of the recent conflicts that flared up between the FASB and the Congress surrounding fair value accounting.

If efficient resource allocation is the objective, financial reports then should include information about externalities. For example, firms should provide information about benefits they receive from activities of other entities, and conversely, about the costs they inflict on the other entities that result from their own activities.Footnote 4 This information would provide valuable input to regulators and policymakers who are interested in devising mechanisms for regulating economic activities through subsidies, taxes, and other incentives. But even though the conceptual framework purports to base its articulated objectives on the overall goal of efficient resource allocation, the GAAP promulgated by the FASB fail to require disclosures about externalities.Footnote 5 Worse than that, efficient resource allocation is foiled by some of the standards, notably the requirement that wherever fair values are required they should be measured as the exit values. In an illiquid market exit valuation inflicts harmful externalities—systemic risk. Unjustified write-downs to near-zero exit values when markets are illiquid by one entity, gives rise to losses, rating downgrades, cash collateral demands, and, in turn, additional write-downs by other entities that may have written credit default swaps or other derivatives on the first entity’s debt, restarting the pernicious cycle all over again.

With this background in mind, it is not difficult to understand why the IASB, and to some attenuated degree, the FASB, are now proposing to allow for managerial intent and ability to hold financial assets to maturity to affect asset quantification. The IASB favors amortized cost in these circumstances, whereas the FASB retains exit valuation but prevents exit value changes from affecting income. What my earlier discussion points to, however, is that neither of these approaches is satisfactory; the combination of amortized cost, discounted cash flows, and exit values, is incrementally informative over any one of these three measures in isolation.

The pressures brought to bear by Congress on the FASB in 2009 to amend FAS 157 were designed to mitigate the systemic risks generated by exit values in reaction to fierce lobbying by the financial institutions. The institutions credibly claimed they were not about to sell their so-called “toxic” assets into an illiquid market, fetching ridiculously low prices. (It is indeed telling that the original TARP program’s strategy of buying the assets from the banks failed because the exit values attached to the assets seemed too low in comparison with what the banks could get for these same assets if held to maturity).

Political pressures will always be exerted on the standards setters as long as the latter’s objectives diverge from those of the policymakers. The fair value controversy reflects the clash between the policymakers’ objective of promoting growth and recovery and the FASB’s objective of informing investors to the exclusion of considerations of allocative efficiency. (Of course, in this case exit values constituted misinformation to the extent assets were to be held to maturity, in which case exit values do not reflect value to shareholders). In the case of exit valuation, for example, additional regulatory capital requirements that are based on the GAAP measures of equity and assets, especially in illiquid markets, would most likely result in decreased lending, hence impeding recovery. There are only two solutions to this conundrum: either harmonize the objectives of the standards setters and the policymakers, or completely divorce regulatory accounting principles (“RAAP”) from GAAP while requiring that GAAP mandate the reporting of externalities.