Positive Accounting Theory

Thanks for visiting this research page. At the moment, I am working on several projects that attempt to understand the political processes that generate new accounting regulations, how firms debate about their preferred standard, and how does that shape the regulations that we see in the real world. I am also interested in business cycles and their effects on corporate decisions, and capital structure choices.

PART I: On accounting standards, the politics and the business cycle

Below, a simple count of the output of accounting standards over business cycles (1975-2005, SFAS only).

What's going on?

Standard-setters produce the greatest quantity of output during the recession, and the lowest prior to the recession. Why? We do not know but there are several possible explanations: (i) political demands, (ii) cyclical financial innovation, (iii) information about problems, (iv) cyclical demands for more information.

Political Pressures as a candidate explanation

During the first signs of an adverse macro shock, many firms have bad future news. They optimally demand lower-quality accounting information/regulations. This reduces future information about the cycle. Then, bad projects accumulate as investors do not learn in a timely manner about a severe recession. When the severe recession is revealed by the increase in bankruptcies, that information now becomes public and there is no more value in implementing low-quality accounting to hide bad macro news. At that point in time, the large number of bad projects calls for more monitoring and better accounting.

For more, see Bertomeu and Magee (2011). A simple diagram illustrates what role accounting regulation plays during a financial crisis.

PART II: Predicting the (overdue) upcoming deregulation of accounting

We know that, now and then, bad regulations are passed. However, and far too often, we look after that regulation has passed, that is perhaps too late, and we do not attempt to better understand why it did pass. The Nobel-prize winner George Stigler described such political processes as "regulatory capture" meaning that interest groups may take hold of the ability to regulate.

How, then, does regulatory capture affect standard-setting in accounting?

An Example

Consider a firm who has some value, say x=10, and who has the ability to choose the information system for all firms (include itself) in the economy. What is the simple information system that maximizes its market price? The solution is to: force those with x<10 to disclose publicly, but for those firms with x greater or equal to 10, do not disclose anything. The dictatorial preferred choice is one that features non-disclosure of good enough news...

In Bertomeu and Magee (2012), Bob and I have looked at a more realistic process in which firms may propose new regulations but, then, that new regulation must be voted against the status-quo by all firms. Below, we plot as a full-line the regulatory choice defined as the threshold below which news must be disclosed. The dotted line represents the value/quality of the firm who had control over the agenda-setting for the period.

Discussion and Analysis

It is apparent that the regulation features cycles. Why? when there is very little mandatory disclosure, most firms are non-disclosers. What regulation wins a majority choice? One that induces disclosure of the worst firms to increase the market price of the other non-disclosers (periods 1-6). But, then, what happens when most firms are disclosers? Low-value disclosing firms band together to sharply reduce the amount of disclosure (period 7). Modeling this, we have seen that there is a choice to be made between the long-term persistence of such cycles, and very rigid regulatory processes that do go to full-disclosure but very slowly.

If our theory is correct, the long period of regulatory encroachment that we have seen over the last century should be followed by a major deregulation of accounting matters (e.g., like that which occurred in other industries, electricity, trade, etc.).

What other forms might be the successor to centralized accounting as we know it? Perhaps we will return to social norms as advocated in the recent works of S. Basu, P. Fischer and S. Huddart, S. Sunder and G. Waymire. In Bertomeu and Cheynel (2013), we've looked at two other different forms, an independent self-regulation body or two competing standard-setting bodies. The latter (competition) seems to be dominant on pure theoretical grounds because it is less prone to regulatory encroachment.

PART III: Explaining the predominance of impairment-like rules (and why it is likely a problem)

Motivation

This has become so much of a second nature of accounting rules that we never do ask why many accounting rules are asymmetric, i.e., tend to emphasize mandatory reporting over bad news. Instead, why don't we have a focus on symmetric measurements, in which we always report information that is material, regardless of the direction of the information. Or, to be entirely open, why don't we mandate early reporting of good news to avoid, say, problems in which the manager temporarily depresses prices to get a better deal on an equity grant or some stock purchases?

Maybe, it is that having more information about bad news is more important for decision-making; however, many casual examples suggest this explanation "does not fly." After all, don't we need to have advance information about good news if we want to decide to scale up an investment policy? And isn't the potential opportunity cost of not making great investments generally much larger (in magnitude) than what could be lost for not salvaging a non-performing investment, if salvage values are not that big in the first place? Yet, accounting rules over-emphasize impairments (bad news disclosures) in many places:

- impair inventory when the market value is below book, but not the opposite.

- impair construction in progress for the full loss when a loss is magnitude, but delay recognition of revenue until the project gets closer to completion.

- impair goodwill (and other intangibles) but do not reevalue that goodwill up if its value increases.

- recognize contingent liabilities when more likely than not, but ignore potential gains.

All of these rules are in contradiction with the stated objective of neutrality and, yet, they seem to be used for a reason.

Discussion and Analysis

In Bertomeu and Magee (2013), Bob and I show that impairments always emerge as a solution to a regulatory process if that regulation process is political, i.e., (a) regardless of details of the production technology (and even if information has no social value), (b) regardless of characteristics o of the information environment.

The intuition is as follows. Assume that rules are made in the political arena and suppose that firms have some private information about their final cash flows. Suppose a rule were not to require impairments (say, report when you make a gain) and no disclosure for bad news:

To set ideas, let's assume that those that disclose report all their information, and those that do not get the expectation (the average price between y and z, in the plot above). Can that type of rule emerge from a political process?

It turns out to be no! To see this, consider expanding the non-disclosure to remove disclosure over some good news, like that:

Compare who prefers the old rule (do not disclose between y and z) vs. the new rule (do not disclose beween y and z+something small). Note that almost everyone prefers the new rule, implying that there cannot be any disclosure of good news.

Therefore, under a political process, the optimal rule is an impairments: disclose when things are bad but do not disclose when things are good.

What goes wrong?

Note that the political process does not necessarily select what is ex-ante desirable, in fact, Bob and I formally show two problems with letting politics decide what standards we elect:

- The standard-setting process is unstable, if the costs of disclosure are low. Regardless what we agree upon, there is a new standard that would be preferred by almost everyone; so we never agree for the long-term.

- There is just too much mandatory disclosure. Why? In a political equilibrium, the non-disclosers vote as a group with a common interest to raise the non-disclosure market price. But, in doing so, they completely ignore the costs borne by those who do impairments. From an ex-ante standpoint, there are too many impairments and the economy would benefit from a relaxation of the rules.