Comments to the
Securities and Exchange Commission

On the Proposed Rule:

Revision of the Commission's
Auditor Independence Requirements

Submitted by

Paul B. W. Miller, Ph.D., CPA
Professor of Accounting
University of Colorado at Colorado Springs

July 19, 2000

Four Axioms

Over the last several years, I have come to look at financial reporting through a new paradigm based on the relationships described in these four axioms:

1. Incomplete information creates uncertainty in investors' minds

2. Uncertainty creates risk for investors

3. The risk causes the investors to demand a higher rate of return from an investment

4. The demand for a higher rate of return produces a higher cost of capital for the investee and lower security prices

In and of itself, none of these propositions is remarkable. However, bringing them together has produced new insights into the practices of financial reporting, accounting standards setting, and auditing.

Although some might be puzzled why, present practices are in fact quite inconsistent with this paradigm. It may very well be true that today's approach to regulating financial reporting is also inconsistent with it.

A New Lens -- Quality Financial Reporting

In various forums, I have called this paradigm "Quality Financial Reporting." In simplest terms, it asserts that significant economic incentives should encourage managers to voluntarily provide greater quantities of more useful financial information than they now provide under compulsion. Of particular interest to the Commission is the implication that these incentives should cause managers to embrace the audit as an essential component of their financial reporting strategy because of its unique capacity to reduce uncertainty, risk, and thus the cost of capital.

The QFR paradigm has led me to see economic issues where I used to see ethical issues. For example, when a manager faces the choice between accounting for a business combination as a purchase or a pooling of interests, QFR shows that the better method is the one that more completely informs the capital markets about the transaction. Where I used to consider a decision to pool as unethical because it invariably shortchanges the markets, I now see it more as a suboptimal and unwise economic choice that brings about the consequence of greater uncertainty, higher risk, and lower security prices. (Of course, the decision to pool would be unethical if the manager's intent is to deliberately deceive the markets by keeping relevant but bad news out of the financial statements.)

As a result, I have put aside much of my previous outrage and replaced it with a positive attitude toward the educator's greatest challenge of expanding knowledge while eradicating ignorance. Instead of doubting the morality of all managers who make poor choices, I have come to question the quality of their economics training and even their common sense. Certainly, many managers and accountants do and will display ethical shortcomings; for this reason, regulation and enforcement are necessary, and will remain so indefinitely.

The QFR Vision

Based on these axioms, QFR has given me a vision for a financial reporting world in which managers learn to compete in the markets on the basis of the quality and quantity of the useful information that they provide. Thus, these managers see auditors as their vital allies in that competition because they can add immense value to financial reports by providing unquestionably independent verification of their truthfulness and completeness. The SEC and FASB will also be managers' allies in this system by continually raising the bar of minimum standards while encouraging innovations that increase quality.

Obviously, a huge gap exists between this vision and current reality, especially as shown by the expressed positions and actions of many managers and accounting practitioners. Instead of competing by increasing information quality and thereby reducing uncertainty, they seem to race to the lowest common denominators of accounting methods, presumably because they desire to use their reports to manipulate the markets into paying too much for their securities. In doing so, they appear to act without regard for the impact of the resulting uncertainty (and the associated mistrust) on investors' risk and the cost of capital. Instead of working cooperatively with what are surely efficient capital markets, they seem to aim at being the one in ten thousand who successfully escapes the markets' careful scrutiny. In doing so, they defy the odds and neglect the accountability that comes when the whole truth is eventually revealed.

Focusing the QFR lens on auditing

QFR shows us that audits are worthwhile only when they reduce the capital markets' uncertainty about a reporting company's condition, past results, and future prospects. Specifically, audits are performed to add value to financial statements by reducing the natural doubts about the trustworthiness of management's self-representations. In order to add that value, the auditors themselves must bring certain qualities to the task, including independence.

A Momentary Diversion

As an aside, I cannot continue without questioning whether the current attention directed at independence issues might be a case of straining out a couple of gnats while swallowing two large camels.

First, many speak and act as if financial statements prepared in accordance with generally accepted accounting principles (GAAP) actually contain useful information. Because many of these principles are quite old (dating back more than 50 years) and because all of them have emerged from clearly political processes, that assumption must be challenged. If it is not valid, perhaps all that an audit opinion can actually accomplish is to confirm for investors that the financial statements are not useful for decisions because they comply with GAAP.

Second, the commission's proposal directs a great deal of attention to the conflict raised by the fees paid to auditing firms for nonaudit services. In focusing on that problem, is it possible that we are overlooking the more obvious problem of the conflicts that are created by the fees paid to the auditors for audit services?

While I am not content to leave these questions unasked, I realize that their answers obviously need to be addressed in another venue. Nonetheless, I encourage the Commission and the staff to keep them in mind as they consider the issues.

The Proposed Rules -- their main strength

The commission's proposed rules are rightfully based on the premises that independence is a crucial contributor to the credibility of audits, which, in turn, are crucial to the credibility and usefulness of public financial reports. Although there is nothing new in these points, they are worth making as clearly and strongly as possible.

In my estimation, your greatest contribution is the conceptual definition of nonindependence (on p. 14) as comprising situations in which the accountant:

1. has a mutual or conflicting interest with the audit client

2. audits the accountant's own work

3. functions as management or an employee of the audit client, or

4. acts as an advocate for the audit client.

These concepts should prove to be useful to the Commission as it writes, applies, and subsequently modifies independence rules.

Perhaps more importantly, they can also guide managers and auditors in dealing with independence issues at the micro level. On the other hand, if they are looking only for specific bright-line rules as to what makes auditors independent or nonindependent, they will be confused, disappointed, or both.

The Proposed Rules -- their main weakness

I am not as impressed by what the Commission's proposal refers to as the effort to "modernize" the independence rules. Even a mildly skeptical observer could consider that word to be a euphemism for the process of molding new rules to fit existing behavior instead of molding new behavior to fit existing rules. I am troubled most by the proposed relaxing of the rules on stock ownership.

Another weakness in the proposal is its tendency to emulate the existing authoritative ethics literature by splitting hairs between situations that do and do not destroy independence. I suppose that such fine distinctions are essential for enforcement, but I have admonished accountants and students for years with this statement: "If you have to look in a book to know whether you're independent, rest assured that you're not." In other words, if someone relies on rules to appear independent, they will fail.

The safety net for the commission (and the public) is, of course, the incredible ability of the capital markets to look after themselves. Without condescension, I can assure you that the markets will reach and act on their own conclusions about independence without regard to what you think, decide, or implement through the rules. As a theme that will return again in a moment, the Commission must establish mechanisms that will help market participants arrive at the desired condition of "knowing all the relevant facts and circumstances" (Proposed rules, p. 6). Even if the rules fail to produce those mechanisms, we can all be confident that the markets have their own means for uncovering those facts or, as QFR teaches us, exacting a higher return to compensate for the greater uncertainty produced by the lack of facts.

As an example, the proposal solicits input on the question of how much ownership is sufficient to create the appearance of compromised independence:

Is the five percent threshold for financial interest in an audit client by persons who do not influence the audit appropriate? For example, would reasonable investors perceive a firm's independence to be impaired if a partner or employee in an office that did not work on the audit, held four percent of the audit client? If the five percent threshold is not appropriate, what threshold is appropriate, and which individuals should be subject to the restriction?

While those questions are interesting and surely challenging, the emphasis on a uniform threshold for all investors misses the point that some reasonable individuals will find even 0.1% ownership to be too large while other reasonable persons might have upper limits of 1%, 2.5%, 5%, or even more. By selecting 5% (or any percent other than zero) as the one size that fits all, you would create the situation in which a registrant's proclamation that its auditor meets the SEC's independence standards would leave investors guessing where in the range of 0% to 5% the actual ownership falls. As a result, comparability will be lost because situations that are radically different in investors' eyes will all appear to be the same. On the other hand, if the Commission were to require disclosure of the actual percentage in all cases, then the needs of all financial statement users would be satisfied. In QFR terms, the more complete information will reduce uncertainty and allow more informed assessments of risk and return. Further, requiring this information to be reported publicly will reduce the private costs of discovering it while avoiding the possibility that some investors will merely speculate about the actual ownership proportion and go on to discount the issuer's securities because of the greater uncertainty.

The Proposed Rules -- the main controversy

The publicized preliminary responses to the proposal (primarily from some leaders in the public accounting profession, as well as several members of Congress) show that the proposed limitations on nonaudit services have hit a sore spot (see "Levitt strikes back," Accounting Today, July 12-23, 2000, p. 1). As I see it, the outcry has its roots in the perception that the Commission is making a major commitment to the longstanding hypothesis that nonaudit services diminish independence, both in fact and appearance. There is no surprise in the fact that the opponents have challenged you to prove this hypothesis with empirical evidence that audit failures have occurred where these services have been present.

The proposal's preemptive response to this challenge (p. 10) is adequate, but it could be strengthened. For example, those who raise this call for proof should be asked to demonstrate why the only manifestation of compromised independence is an audit failure. That is, QFR suggests that capital costs are higher in any circumstances in which uncertainty about the auditor's independence creates risk for investors. Further, they should explain whether they accept the substantial empirical evidence (and common sense) in favor of the hypothesis that the capital markets are efficient and incapable of being misled by accounting policy choices. Perhaps it would be useful to point out that they cannot have it both ways. On the other hand, this inconsistency in their arguments may simply reveal their real motives.

At the same time, I assert that the QFR paradigm actually lays out a higher standard of proof. Specifically, it challenges the proposal's opponents to demonstrate (logically and empirically) how the presence of nonaudit services (and fees) would decrease uncertainty about the financial statements' reliability and thereby reduce the cost of capital. Through intuition and the sources described in the release, there can be little doubt that these services do create uncertainty and thus negatively impact shareholder value.

Of course, the absence of proof for either position on the issue should not deter your intention to rein in the virtually unbridled ambitions to expand services without regard to the resulting impact on the credibility of all auditors' opinions and all financial statements. It is not difficult to speculate that the aggressive erosion of confidence in the integrity of audits for one or a few clients would spill over into uncertainty about other clients or indeed all registrants.

With regard to the specific categories of services that would be restricted or prohibited, I concur that you have identified those that are most likely to diminish real and perceived independence. I also concur that outright prohibition is worth considering. As an alternative, requiring full disclosure of the existence of these questionable but permitted services and the resulting fees would give the markets relevant information that they do not presently have. (Momentarily, I comment further on a recommended disclosure strategy.)

The Proposed Rules -- a troubling point

Although the proposal does address one other problem briefly, I have to confess that I am thoroughly handcuffed by it. Specifically, we are all aware of the bewildering array of multiple authorities that govern CPAs. Even if the Commission does choose to amend its independence rules by allowing some ownership to occur, the fact remains that such a condition is currently unacceptable in all 54 licensing jurisdictions, not to mention an equal number of state-level professional societies. Simply waving a wand of permission over accountants with SEC clients will not make them immune against enforcement actions at home that could possibly take away their licenses. I trust that someone will address this issue soon.

My Proposal -- Client Ownership Limits

Except for the commitment to find reasonable ways to deal with inadvertent temporary holdings, I do not see any redeeming value in the Commission's proposal to permit auditing firm personnel (and their families) to hold securities issued by a client. I simply do not find the proffered rationale of dual-income families and other demographic changes to be persuasive. As you might expect, my point of view is shaped by the QFR paradigm.

Individuals who aspire to be independent auditors essentially set themselves apart from the rest of the economic community so that they will be qualified for their peculiar task of attesting to veracity. As a result, they cannot enjoy this unique status and integrity while simultaneously engaging in all the same sorts of economic activities that nonauditors engage in. While QFR tells us that they certainly can be free to act like everyone else, it also shows that they cannot expect to add as much value to their clients' financial statements if they do so. Instead, the four axioms show that they are likely to obtain higher billing rates when they remain objective and independent. Thus, the argument against allowing ownership is more economic than ethical. Nothing has changed in our demographics to alter the pure economic power of the law of supply and demand as it applies to audit assurance.

This analysis leads me to first encourage the Commission to not accept anything other than temporary inadvertent ownership. If you cannot make the acceptable level zero, then I suggest that you create a regulatory environment in which auditors are confronted with the negative effects of the conflict of interest associated with ownership. Eventually, they would come to understand that any level in excess of zero creates uncertainty, risk, and higher capital costs for their clients, and lower fees for themselves. One key to building that environment is implementing the disclosure recommendation described in the following section.

New investment media

While allowing ownership, the Commission is attempting to respond to auditors' needs for greater access to the capital markets for managing their own portfolios. Instead of changing the ownership rules, I suggest that it makes more sense for the Commission to encourage auditors to create new two investment media: blind trusts and special mutual funds.

A blind trust would be created by transferring assets to a trustee who would be expected to earn a satisfactory return for the beneficiary while keeping the identity of the trust's investments a complete secret. By using one of these trusts, the auditors would be able to claim with more credence that they simply do not know whether they have any conflict of interest through their individual holdings. Such trusts could be most easily created for individuals, although one might be used for small groups. The greatest risk, of course, would be maintaining actual and apparent secrecy.

A second alternative offers advantages over blind trusts. Specifically, I think that it would be feasible and advisable for each larger firm to engage money managers to create customized mutual funds that would hold only securities issued by companies that are not clients or business affiliates. This arrangement would allow the auditors to enjoy the same benefits as other investors (professional management, diversification, and asset allocation for different investment goals) without the risk of losing their independence. This arrangement would also simplify the firm's procedures for preserving independence. For example, the event of obtaining a new client would trigger an order to the fund manager to immediately remove any of that company's securities from the portfolio. This single instruction would obviate the need for the word to trickle down through the organization and trigger hundreds of sales. Similarly, the loss of a client would allow the money manager to consider adding its securities to the portfolio. Thus, using one or more of these funds would greatly reduce the demands placed on the firm's quality control system for monitoring investments. It would also eliminate the need for secrecy about the portfolio contents. In fact, QFR tells us that full disclosure of the contents would reduce uncertainty, risk, and the capital costs of incurred by the firm's clients.

I cannot be sure but it seems highly likely that this service could be provided at relatively low cost by virtually any of the large mutual fund families.

My Proposal -- Mandatory Disclosures

The Commission's proposal calls for new disclosures in proxy statements about nonaudit services and fees as well as the existence of any arrangements that lead to "leased" auditors. It is possible that the Commission and staff have underestimated the potential advantages of public disclosures about independence.

To help clarify, I point to the difference between negative and positive confirmations of receivables and payables. The former are sent out with the instruction to reply only if an error exists. The latter carry instructions to respond affirming that the balance is correct or explaining why it is incorrect. The limitation of the negative confirmation is the ambiguity of the nonresponse -- was nothing received because there was no error, because the recipient was not motivated to reply, or because the account is not legitimate? The silence creates uncertainty that makes the negative confirmation process ineffective for reducing risk.

In the same way, investors face great ambiguity when financial reports are silent on the issue of whether the auditors were independent. As indicated earlier, this effect is exacerbated if new independence rules were to create an acceptable range for ownership (such as 0 - 5%) when investors want to know specific actual percentages within that range in order to assess the risk of nonindependence.

This analysis leads me to propose that the Commission require two separate disclosures in every financial report containing auditors' assurances. One would be prepared by the auditors to proclaim why they consider themselves to be independent of the specific client and the other would be prepared by the registrant's management to explain why they believe that the auditors are sufficiently independent to justify confidence in the opinion and the financial statements. QFR shows that these disclosures will provide more complete information to investors, reduce their uncertainty and risk, lower the demanded rate of return, reduce the cost of capital, and increase security prices. The ultimate result would be greater efficiency in the capital markets and the economy.

With regard to the specific content of these reports, I suggest that the Commission could build on the foundation established in the Independence Standards Board's first standard. The rule would go well beyond that standard by making public the mandated dialogue between the audit firm and the audit committee. In effect, this requirement would shine a bright light into the details of this important relationship.

Without getting too specific, I suggest that the disclosures could address the following points:

Ownership -- describe which client securities (stock, bonds, hybrids, options, others, whether voting or nonvoting) are held by whom, for how long, and why.

Relationships -- describe all potentially compromising personal relationships between client and auditor personnel, whether family, social, or otherwise; both the auditor and the client would aver why these relationships are not detrimental to independence

Business relationships -- describe all potentially compromising business relationships between the client and the audit firm or audit personnel, including nonaudit-related transactions as well as such things as loans and credit card balances owed the client by auditors

Former relationships -- describe all previous auditors who are now client employees and all auditors who were formerly client employees; again, the parties would be compelled to provide sufficient detail to explain why independence is not compromised

Quality control systems -- describe the systems used to prevent compromising the auditors' independence, including an assessment of why they can be considered reliable

Fees -- describe the amounts of both audit and nonaudit service fees

Nonaudit services -- describe all nonaudit services with sufficient detail to eliminate the perception that they may have compromised the auditors' independence

Audit services -- describe all audit-related services provided by the auditing firm to the client; this section would describe any accounting policies put into place at the auditors' suggestion, especially if such policies embrace practices that have been identified as or might be considered to be less preferable than other generally accepted practices; these practices might include such things as poolings of interest, excessive allocations to in-process research and development, off-balance-sheet financing, the indirect method of describing operating cash flows, and off-income-statement reporting of stock options expense; both parties would be tasked with explaining why implementing these policies was consistent with the goal of providing full and fair disclosure

Other topics might be considered, as long as the information would help the financial statement users assess the auditors' independence from the client.

Advantages

The explicit information provided in these disclosures would help the capital markets avoid guessing about the conditions. As mentioned earlier, different investors have different criteria and this approach would not rely on the "one size fits all" approach inherent in drawing bright-line criteria for independence.

This requirement would also be advantageous for its ability to make auditors and managers conscious of the advantages of dealing openly with independence issues. The need to disclose could cause the two parties to systematically identify, address, and resolve problems involving potentially compromising activities and situations that they presently may not even consider. In effect, the disclosure process will make them more aware of their vulnerabilities and the need to protect this most valuable attribute. The ultimate consequence will be a greater degree of both actual and perceived independence.

A third advantage of this approach is a diminished burden on the Commission and the ISB (and other authorities) for developing detailed descriptions of relationships that are tolerable or intolerable. While enforcement demands minimum requirements, the disclosures will allow financial statement users to decide for themselves whether the auditors have certain relationships that they consider to be compromising.

Cost-based arguments

I do not believe that the costs of the disclosure requirement would be material because the disclosed information should already be available. To the extent that it is not, the requirement would produce additional benefits.

I encourage you to avoid lending credence to two cost-based arguments that will be heard if you propose this disclosure. First, some will claim that it will produce information overload for investors. If professional money managers are investing millions of dollars, it seems clear to me that they will want access to as much information as possible. While overload is theoretically possible, I am far from convinced that we are anywhere near that limit.

Second, others may suggest that this much disclosure will invite litigation. There is little reason to be sympathetic with this complaint. The best defense against claims of deception is to tell the truth as you understand it, even if the truth is that you don't know what the truth is. If the opponents fear that events will prove that the original assessment was wrong, they should realize that their silence about independence is tantamount to claiming that it exists. Which would offer a better shelter against ex post recrimination: saying what you know or withholding relevant information and hoping that nothing happens?

An Additional Suggestion

In conducting the due process on this proposal, I believe it is essential that the Commission solicit, receive, and act upon input from nonaccountants, specifically financial statement users. Perhaps the main reason for the present predicament is the fact that the definitions of independence and its limits have been established through political processes in which accountants talked with accountants and then struck their own compromises. The users' and the public's interest were not directly considered because these constituents were not invited to enter the debate, much less empowered to shape the resolutions of the issues. This advice is in line with recent suggestions that the Independence Standards Board be reconfigured to include a majority of nonaccountant members.

Of course, I am not suggesting that accountants be excluded from the due process. They are entitled to participate and surely will do so. My point is that what they say must not be the only driving force in reaching your final conclusions.

Summary

I commend the Commission for attacking a major issue with a bold proposal. The boldness is evident to those who have been concerned over these issues; it is also made evident by the opposition that has been mustered to hinder the reform.

The proposal deals well with some pieces of the issue, particularly the definition of the compromising forces that produce nonindependence. I think you can do other things to make it stronger, including the continuation of the existing ban on ownership, the limitation on nonaudit services, and the proposed expansion of mandatory disclosures.

Once you have settled these matters, perhaps you will be in a position to address the much bigger issues of (a) whether GAAP financial statements actually contain information that is sufficiently useful to justify the cost of the audit and (b) whether the payment of audit fees by the client compromises the auditors' independence.

When you decide to attack these issues, I can assure you that I will be ready to testify again.