Comments to the
Securities and Exchange Commission

On the Proposed Rule:

Revision of the Commission's
Auditor Independence Requirements

File No. S7-13-00

Submitted by

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

July 31, 2000

Original comments dated
July 19, 2000


Thank you for this opportunity to speak my mind.

The QFR paradigm

My comment letter describes the Quality Financial Reporting paradigm that I have found to be useful. The letter explains that it is based on the relationships described in these four axioms:

  1. Incomplete information creates uncertainty

  2. Uncertainty creates risk

  3. Risk creates demand for a higher rate of return

  4. A higher rate of return translates into a higher cost of capital and lower security prices

The risk also creates friction in the capital markets that can be reduced by more complete information.

The implications

The QFR paradigm reveals that uncertainty is a problem and an opportunity. As I see it, we have tried to deal with it in these five ways:

  1. Public reporting is supposed to reduce uncertainty about the issuer.

  2. GAAP are supposed to reduce uncertainty about the issuer's reports.

  3. Audits are supposed to reduce uncertainty about the GAAP financial statements.

  4. Independence rules are supposed to reduce uncertainty about the audits.

  5. The SEC is supposed to reduce uncertainty about all of the above.

Therefore, I believe that the Commission must resolve today's independence issues in ways that reduce uncertainty about the quality of the reported information. All other considerations should be secondary.

What are the issues?

Of course, the framing of the issues is crucial to any debate. However, as I consider the subject of this hearing, I find only two significant issues that are officially on the agenda. I find two more that ought to be on the table, but aren't. Finally, I see a fifth issue that no one has explicitly acknowledged.

The two issues on the agenda are the so called modernizing of the ownership rules and the prohibition of certain nonaudit services.

I think the two issues that are not on the table are actually more significant:

First, we need to ask whether GAAP financial statements are worth being audited. This issue exists because many of the principles are very old and all of them are political compromises. If the financial statements cannot produce useful information, it doesn't matter whether they are audited or not. It certainly doesn't matter whether the auditors are independent.

Second, the extreme focus on the controversy over nonaudit fees has caused us to overlook the important question of whether audit fees compromise independence.

Despite their significance, I am confident that these issues won't be addressed in this venue.

The invisible issue

We are also faced with a troubling but invisible issue. Even though it is not addressed in your proposal, it is on everyone's mind. Specifically, I think the real debate is over the question of who is empowered to decide when independence is impaired.

  1. The Commission has had this power for decades and has used it to create and enforce rules that apply to all registrants and auditors.

  2. The AICPA has had this power and expressed it through ethics rules.

  3. The auditing firms have tried to create and enforce their own independence policies, with varying degrees of success.

  4. Finally, the ISB was created to help define independence, but its structure has limited its credibility.

So, where can we look for the answer?

Authority and power

On this point, I want to share an insight I gained one day from Clarence Sampson during a meeting with the chief accountant of Israel. At one point he asked whether Clarence had the authority to overturn the FASB. With characteristic directness, Clarence said: "I may have the authority but I don't have the power."

Those word's summarize the controversy. You have the authority to prohibit ownership and nonaudit services, but some people aren't sure that you have the power to do so.

So -- where might the power reside?

Over the years, I have come to have great respect for the power inherent in our capital markets.

My ultimate point is that the markets will make their own assessments of independence. In addition, they have the power to impose much larger and more effective penalties than you can.

So what?

This framing of the issue leads me to suggest that this particular rule making process can be a landmark, but for a different reason than some might think.

Here is what I mean -- independence rules have always been created by accountants after long discussions with other accountants. Should anyone be surprised that the outcome elevates accountants' interests over the public's interest? As a result, uncertainty has been increased and capital costs have been higher.

That's why I strongly encourage you to resolve these issues only after listening carefully to the investment community. You should not ignore accountants, but you should not be persuaded by them.

Now is the time to create rules that protect the public from accountants instead of the other way around.

I also have suggestions concerning the two specific issues in your proposal.

My advice on stock ownership

On relaxing the ownership rules, my language is plain: don't do it. I see this move as a step out on a very steep slippery slope that produces no redeeming value for the public.1

As I consider this situation, it looks to me like the problem is the inconvenience the independence rules create for auditors in building and maintaining their personal investment portfolios. It seems to be hard for them to know what they or their relatives cannot own and when the status of specific companies changes and moves them on or off the list of forbidden stocks.2

If you were to change the rules as proposed, surely some auditors would be able to manage their portfolios more conveniently. However, what would be the cost of this convenience?

QFR tells me that the capital markets would respond to this change and the resulting additional uncertainty about the financial statements by demanding higher rates of return that will, in turn, create higher capital costs and reduce security prices. In effect, the stockholders would pay a large price for the added convenience of the auditors.

In addition, there is the difficult matter of overlapping jurisdictions. Unless state regulators change their rules to allow ownership, a change in the commission's rules will accomplish nothing. It is possible for your standards to be more restrictive than the states' rules, but the opposite cannot occur. (The same problem exists for the amnesty program -- anyone who confesses independence violations to the Commission is exposed to sanctions from the state licensing authority.)

I also wonder if ownership by other members and employees of the audit firm will expose them to the possibility of being considered to be insiders through the confidential knowledge that would exist within their firm, even if they did not participate directly in the audit. Auditors would certainly face temptations to give their coworkers a heads-up on extraordinarily good or bad news.

Further, one has to believe that ownership within the firm is simply not congruent with encouraging the perception of independence. The proposal is so far removed from the time-honored existing rules that it is imprudent to change them as you have proposed.

Finally, I have to point out that this part of the proposal appears to be an attempt to mold the rules to match existing behavior instead of taking steps to mold behavior to match existing rules. If so, the Commission's own integrity is at stake.

In light of all these conclusions, surely there are more productive strategies for increasing the auditors' convenience without creating so many problems.

In particular, I am convinced that there are sufficient economic incentives to encourage auditors to come up with their own strategies within the existing rules that would not expose their clients to the near certainty of higher capital costs.

Simply put, I encourage my practitioner colleagues to do what I have done and hire professional money managers to design and implement investment programs for partners, employees, and their families. The instructions to the money managers would include staying away from any client's securities.

My original comment letter also suggests that an auditing firm could hire a large fund manager to custom design a family of mutual funds that do not own any client securities. This arrangement will provide the auditors and their families with professional management, diversification, liquidity, and asset allocations according to their needs. It would also facilitate responding to changes in the client list -- for example, a new client's stock could be quickly stripped from the funds' portfolios with a single call from the audit firm. In contrast, without the funds, the act of obtaining a new client would trigger massive communication difficulties and a myriad of individual transactions as the securities would have to be purged from literally thousands of individual portfolios.

I am convinced that this voluntary approach will create adequate convenience for the auditors while reducing the markets' uncertainty about independence.

This approach has to be superior to changing the commission's longstanding rules.

Nonaudit services

Because I anticipated that so many witnesses would address the issues surrounding nonaudit services, I did not prepare many comments to present at the hearing. The comments that I do offer are limited to these two points:

Subsequent to the hearing, I have drafted a column for Accounting Today that describes and attempts to dismantle the three main arguments tendered in support of the status quo That draft is attached to this submission as an Appendix.

In summary, I find serious fault flaws in the three primary assertions by the representatives of the large firms at the hearing: (1) that nonaudit services improve the quality of the audit, (2) that banning certain nonaudit services will create serious obstacles to hiring new employees, and (3) that there is no proof any audit failure has been caused by independence that was impaired by nonaudit services.

The column challenges these arguments as follows. First, the witnesses never defined what a "higher quality audit" is. Second, there is nothing of substance in the claim about recruiting difficulties. The column identifies a number of other factors within and outside the scope of the auditors' control that impact the firms' human resource management efforts. Third, the lack of independence is certain to have negative effects other than audit failure on clients, investors, auditors, and capital markets.

A disclosure requirement

Toward the end of the hearing, it became apparent that the commissioners were starting to embrace the possibility that new disclosures would contribute positively to the situation. I had prepared the following comments prior to the hearing but the exigency of the 10 minute limit caused me to rush through them.

Regardless of what you decide on both issues, I think you should expand your proposal for disclosures about independence.

Furthermore, this disclosure requirement would cause the auditors and the managers to explain why such things as ownership or using the audit firm for nonaudit services do not create uncertainty about independence. Perhaps if they have to explain their decisions, they will think again before engaging in these controversial behaviors.

I emphasize one more time that the key to improving the current situation is reducing the markets' uncertainty by eliminating their need to guess.


In closing, I want to say that I believe that you currently possess both the authority and the power to forbid ownership and incompatible services.

You also have both the authority and the power to require managers and auditors to more fully inform the markets through mandatory disclosures about independence. It is clear to me that you need to act on these issues in a way that protects and promotes the public's interest in reducing uncertainty about independence.

I also think you have the obligation to produce these requirements despite being threatened by some constituents and harassed by some members of the Congress. These issues are too important to the economy and the capital markets for that kind of extraneous and irrelevant political pressure. The premise of those who apply the pressure is that they can boost the securities' market values by controlling the contents of the financial reports that management publishes. In fact, the absence of information increases uncertainty and drives security prices down, not up. In addition, the markets have the ability to discover useful information through other means and adequate incentives to apply that ability.

Thus, the pressure is clearly not intended to force you to take steps to produce more fully informed and thus more efficient markets. Rather, it is intended (vainly) to shelter certain managers and auditors from fully informed public scrutiny and economic sanctions in the form of higher capital costs. There is no justification in responding to that stimulus.

The ultimate irony is that you are being opposed by auditors for offering to do something that will ultimately increase the value of their audit services by enhancing their ability to reduce uncertainty about their clients' financial statements.

Thank you for the opportunity of appearing before you, and I hope that you will produce a solution that reflects the public interest in these issues.

Appendix -- Draft of column to appear in Accounting Today


Paul B. W. Miller and Paul R. Bahnson


In late July, Paul Miller had the experience of testifying before the SEC on its proposed independence rule changes. After his fifteen minutes of fame, he stayed around to listen to others who testified.

The big issue of the day (but not the biggest independence issue) was the question of whether the commission should prohibit certain nonaudit services by a registrant's audit firm. The guiding principles would be that the auditors should not audit their own work, should not play a management role, and should not have a mutuality of interest with their client. The sorts of services that are at issue include system installations, valuations, and internal audit outsourcing.

Anyone who's been around for more than a few years knows of the long running debate over this apparent melding of two distinct and incompatible roles for the auditor as watchdog and concierge. The commission deserves praise for finally tossing this bone into the ring where it can be dealt with officially.

Of course, the really interesting discussions continue to occur elsewhere, like at the Public Oversight Board, which had to get help from SEC chairman Levitt in maintaining its own independence after the SECPS executive committee tried to stop the funding for special investigations into stock ownership by auditors. They are also occurring in Congress where certain members are discovering a latent interest in auditing issues. Can anybody wonder what caused them to suddenly get motivated to express themselves?

In any case, one of the high points in the hearing was a panel of three partners from Deloitte & Touche, KPMG, and Arthur Andersen. (The other two firms, Ernst & Young and PwC, have already weighed in as generally not opposing the proposal.) The witnesses presented various opinions, but really did not display significant disagreements among themselves. Three main arguments kept coming up time and again.

The first was that nonaudit services (NAS) produce higher-quality audits. Sorely missing from the discussion was any definition of what sort of quality they were talking about. Did they mean "faster," "cheaper," "more thorough," "more profitable," or "more assurance to users that the financial statements comply with GAAP and thus cannot be trusted"? No one said exactly what they had in mind. The three had a hard time refuting an earlier witness, Mike Cook, formerly of D&T, who said that this alleged direct connection is not confirmed by the fact that fewer than 25% of SEC registrants use their auditors for NAS. (It has be true that much fewer than 25% use the disputed services.) They also could not explain away the statement of John Biggs, CEO of TIAA-CREF, that his company has an inviolable policy of never using its auditors for anything but auditing.

About all the witnesses could claim was that the NAS enabled them to learn more about their clients. Excuse us, but isn't that what an audit is supposed to accomplish? One even asserted that outsourced internal audits are a "natural extension" of the external audit. A quick glance around the audience saw heads moving back and forth horizontally on that one.

There may be some good synergy argument, but it hasn't been made yet. To those who want to make it, we encourage you to start by at least defining what you mean by higher quality.

A second recurring argument was that the ability to provide NAS to audit clients was essential for hiring the so-called "best and brightest" students at the entry level. All three witnesses claimed their firms would be starved for good people if the SEC implemented the ban. The commissioners who asked questions on this point were clearly unpersuaded, despite the auditors' vehemence. It seems to us that disconnects abound in this argument. The SEC is proposing a ban on providing NAS only to registrants who are audit clients; thus, the services can be provided to any non-SEC client and to SEC registrants who are not audit clients. We suggest that it's harder to recruit good students because they are capable of identifying what public accounting is like. They know they can expect long hours, average compensation, high turnover (90%), a tough CPA exam, and now the nearly universal need to remain in school for another 30 hours and forego a year's salary. Not to mention the fact that the auditors' main product is a boiler plate opinion that attests to compliance with GAAP and thereby affirms that the information in the financial statements is not useful to those who receive them. The problem is not as simple as being able to provide NAS to SEC audit clients.

The third argument was that there is no "empirical evidence" that any audit failure has been caused by NAS. To consider analogous arguments: "there is no need for a guard rail on this highway because no one has driven over the cliff yet," or "we're cutting back on security in the bank because we've never been robbed." The SEC's obligation is to ensure that no audit failures occur in the future; the auditors' response is, to no one's surprise, to look at the past.

More importantly, we also point out that audit failure is the wrong factor to consider. We can't imagine that the only purpose of an audit is to avoid failing. Instead, the real objective is to bring credibility to management's otherwise untrustworthy self-representations in the financial statements. The issue is not whether the auditor can avoid catastrophic failure but whether the audit can increase the credibility of the statements enough to make investors perceive a lower risk of being misled. Thus, the question that ought to be debated is whether the occurrence of certain NAS will cause that perceived risk to go up, go down, or remain the same. The answer, of course, is that the ambiguity created by these services can only increase the uncertainty and the risk.

So, we think the big three's big three arguments are void of substance. We don't doubt that the witnesses and other leaders in their firms sincerely considered the arguments and found them persuasive. The fact that they believe them doesn't prove anything to us except that it can be very hard to accept what one considers to be unacceptable. Psychologists call that reaction "denial" because the individual denies clear evidence of what is true.

We don't think we can convince the adherents of NAS for audit clients to see the flaws. However, we can encourage the SEC and others to not fall victim to those empty claims.

Paul B. W. Miller is Professor at the University of Colorado at Colorado Springs and Paul R. Bahnson is Associate Professor at Boise State University. The authors' views are not necessarily those of their institutions.

1 As an exception to a total ban, it makes great sense to create a reasonable time period to allow auditors and their families to deal with inadvertent and other temporary situations.
2 Clearly, the existing rules do not deny auditors the ability to participate in lucrative investments. Instead, they merely constrain the auditors against certain investees. Further, a substantive public policy reason supports creating these constraints, namely the desirability of reducing friction in the capital markets by assuring investors that the auditors can be trusted.