Financial accounting has emerged in response to the need of managers to communicate to owners [principal] regarding whether the principal’s capital entrusted to them is preserved intact, whether investment yielded income, and how much. This is the traditional role of an agent accounting to a principal, cloaked in modern attire. But with multiplicity of participants in the market place and with increased specialization, owners of capital had to delegate the conduct of business to professional managers.
It is this development that has made accounting a tool for guaranteeing that management fulfills its stewardship function toward stockholders: management accounts for how it safeguards and manages the owners’ resources. Seen from this perspective, accounting is a monitoring mechanism either demanded by stockholders or volunteered by management to minimize diversion of resources to activities that do not serve the purpose of owners. An unresolved issue is whether the resulting accounting intelligence is socially optimal, but accounting has evolved to provide information for resource allocation in the economy and consumption and investment decisions.
In this post, I am going to express my opinion concerning financial accounting theory and research. If you somehow just landed on this post and have been reading so far, please bear with me, share your thoughts and opinion through the comment form at the bottom of this page, it may useful or at least it inspires others. Again, bear with me and read on…
Nature of Accounting Theory and Its Environment
Unlike the finance discipline, we do not yet have anything approximating an accounting theory. An accounting theory must accomplish two objectives:
- First: describing, documenting, and explaining accounting practices as an equilibrium phenomenon within a market with diverse constituencies; and
- Second: paving the way for the normative improvement upon what currently exists
The fact that constituencies [investors, financial statement preparers, financial analysts, auditors, investment bankers] exert pressure on accounting rule-making bodies [e.g., the Financial Accounting Standards Board [FASB] in the USA] to establish accounting standards that would further their own self-intent implies that a formulation of an accounting theory is not feasible.
The role of accounting rule-making bodies [such as the FASB] in an environment of conflicting interests also should be to gain an understanding as to what corporations would voluntarily wish to disclose. For example: would self-interested, corporate managers have incentives either to withhold information that is important for resource allocation or to disclose misleading information?
Given the conflicts between owners and managers [and divergences of preferences], managers have an incentive to misrepresent. In fact, there is evidence of managers delaying the disclosure of negative information until such time when the auditor ”flushes out” the news. This possibility of misrepresentation gives explicit recognition to the asymmetry of information: managers possess private information gained from their intimate involvement in the operational activities, and can signal expectations of future cash flows contingent on their own actions when given incentives to do so. By “accounting” to owners for their actions and the assets under their custody, managers fulfill a “stewardship function” and an “informativeness function”. Stockholders require reporting not only as means for providing an agreed-upon measure of performance, but also to obtain information on the basis of which they can make future decisions. This can also be referred to as “the role of signaling”.
In the USA, class action suits brought under the Securities and Exchange Act of 1934 create disincentives not to report truthfully or in a timely fashion, and may deter managers from disclosing relevant information which they have no duty to disclose under present rules. Indeed, the optimal amount of misrepresentation induced by the intricate web of existing incentives and disincentives [compensation schemes, legal liability, and ethical rules of conduct] is above zero.
The role of the FASB [or other regulatory bodies], were they to identify the deficiencies in existing disclosure due to the lack of sufficient incentives, would then be to provide the proper incentives [including penalties] for the generation and provision of useful information. To minimize expected costs of misrepresentations, it would then be in the managers’ best interest to deviate from the guidelines only for the sake of truth-saying.
Accounting Standards: Uniformity versus Flexibility
Should management be afforded flexibility among alternative treatments that would fit the different circumstances in ways that satisfy accounting objectives?
Increasing the uniformity of accounting standards [reducing the flexibility of management in choosing among different accounting treatments] could reduce audit [monitoring] costs and the ambiguity of the resulting signal. The greater the ambiguity of the signal, the lesser the reliability of the resulting inferences.
But the more confining the generally accepted accounting principles and the methods of their application, the lesser the flexibility that management can use to convey its expectations of the firm’s prospects. This reduction in the ability of management to signal counteracts the reduction in monitoring cost produced by the enforcement of uniform standards.
I conclude, not surprisingly, that there is an optimal amount of uniformity of accounting standards that balances the costs of limiting management flexibility and signaling against the benefits inherent in the reduction of monitoring costs due to uniformity.
The Descriptive Role of Accounting
The “informativeness role of accounting” has both descriptive and normative implications. A descriptive finding that accounting information is used in decision-making by market transactors is a necessary, but not sufficient, condition for the claim that the manifested effects should guide the selection among accounting alternatives.
A large body of empirical evidence suggests that accounting information is used but that it explains only a small proportion of the variations in stock returns. But the suggestion that earning should be the sole criterion by which to judge the usage of accounting reports is fraught with hazard. The problem of “jointness” of factors influencing returns is a severe one and cannot be addressed adequately by empirical studies. Earnings is but one, albeit important, output of the financial accounting process.
However, producing earnings requires generating contemporaneously available accounting information other than earnings that also contribute to explanation of returns. From this perspective, the cumulative effects of balance sheet items as well as other accounting variables can be viewed as the informational outputs of the “earnings” generating process. An emphasis on the earnings generating process as the informative source also implies that the earnings measurement process is designed more to reveal a firm’s economic structure than simply to release each period’s data. That is, the earnings process may be viewed as building an archival history to facilitate predictions of how a company will respond to new external developments.
Inference of usage, however, does not imply usefulness. Usage does not address the desirability of the resulting resource allocation and the social utility derived from risk sharing induced by the “used” accounting information. Securities’ prices would reflect the net benefits of accounting information if such information were sold at a competitive market price. But accounting information does not constitute a “private” good in the sense that it is exchanged in the market place. It cannot be sold to consumers [users of accounting information] because of the difficulty of guaranteeing exclusive access to the information if it were sold. As a result, they cannot be used to assess the net benefits of different accounting alternatives.
Interpretation of the Evidence’s Pitfalls
Aside from the impossibility of inferring desirability, assessing the “effects“ of accounting and auditing methods is fraught with hazards. Observing an effect [e.g., an association of an accounting number, change in method, or regulation, with a price movement] need not be attributable to the accounting “event”. Consider the two cases:  when an association is observed, and  when no association is observed.
When An Association Is Observed
In this instance is consistent with one or more of the following:
- The observed price change is induced incrementally by the accounting event reflecting a change in the fundamental value of the firm that would not have occurred in the absence of the accounting event.
- Publicly known non-accounting events would have induced the same price change in the absence of the accounting manifestation.
- The accounting manifestation was not a cash flow event, but it induced a price change because it was perceived as a credible signal of management’s expectations about the future prospects of the firm.
- Without the accounting information traders will have engaged in the acquisition of substitute private information which will have resulted in similar [or different] equilibrium price changes. That is, the observed associated [with the accounting event] price changes need not reflect the incremental impact of the accounting event.
- Managers misrepresent and traders rely on the misrepresentation as if it were truth in light of high costs of private information acquisition and/or the promise of recovery under legal statutes and the case law, so that the price change coincident with the accounting misrepresentation is a distorted reflection of changes in the fundamental value of the firm.
- The price change associated with the accounting event, by itself, does not fully reflect the information content implied in the event. Other measures must be incorporated, some observable [such as trading volume reaction] and some possibly not easily observable [dispersion of traders’ beliefs].
- The price change associated with the accounting event does not fully capture the information content of the publicly announced accounting event because in a noisy rational expectations capital market equilibrium prices do not fully reflect publicly available information, i.e., the capital market is not semi-strong efficient. [see Dontoh et al., 1994].
Thus, independently of the magnitude of the explained variation [R2 values, which were observed by Lev  to be very small], the observation of an association between the two does not tell us much about the “usage” of accounting information.
When No Association Is Observed
Symmetrically, in this instance, the observation of “no effect” need not imply that accounting information is not used:
- First, even if there is no observed effect on market prices, the accounting event may have an effect on utility transfers among individuals as manifested in trading activityclearly a relevant factor if wealth distribution is considered important.
- Secondly, the effect of an accounting manifestation may become visible subsequent to the issue of the accounting report. Thus, a disclosure that management took a particular action [or a financial statement manifestation of such action] may have little or no effect at the time of disclosure. But later, the occurrence of an unexpected event significantly—altering—cash flow–expectations given that management had taken the said prior action, would affect stock prices.
One cannot argue that the effect observed on the security prices contingent on the disclosure of the information about the event’s occurrence is solely attributable to the information on the event. The effect is a joint result of:
- the occurrence of that particular event; and
- the action taken by management that had previously been disclosed in the accounting reports.
However, the disclosure of the action may have an “effect” that is conditional on additional information which becomes available only later. In this circumstance it is wrong to claim that the disclosure of the action in the annual report had no effect. In other words, the observation of security market prices cannot be used definitively to assess whether the accounting information has an effect on market equilibrium.
Thus, event studies that focus on the association between market returns and accounting numbers do not shed much light on the question as to whether accounting numbers are used. Even if they did, we would still be in the dark as to whether any accounting numbers, used or not used, are useful. Existing empirical evidence, as well as evidence expected to be yielded by future empirical studies of the type conducted in the past, provides us [or will provide us] with little, if any, guidance as to how to structure the accounting model, how to report, and how to formulate the objective, scope, or procedures of accounting and auditing.
We are left with what may be the only promising path of research toward construction of a financial accounting theory: normative research [including empirical or experimental tests of the assumptions underlying the normative analysis]. And this should not exclude consideration of social welfare implications of generally accepted accounting principles [GAAP] alternatives. To the extent that such implications are best flushed out in the political process, the aim of research should be to provide valuable inputs to the political debate. This research should certainly allow for well-reasoned, well-thought-out, hypotheses as to which accounting models would best serve society’s needs.