Asymmetrical Information

Another professor of accounting at Stern/NYU, Baruch Lev, had a simpler idea for making markets work. He testified to Congress that auditors should be hired by shareholders, not by corporate management.

Under the status quo the CEO not only naturally tends to get what he wants in the audit report from the auditor whom he picks and pays, but also controls the flow of information that gets out to shareholders and investors.

But if shareholders and investors hired, fired and paid the auditor, the auditor would have a very different priority about whom to please. And the flow of information to shareholders and investors would be much more complete. With no new regulations necessary.

As a second point, for some reason I don't understand (but suspect has to do with lobbying by incumbents) the federal government requires audits to be conducted by partnerships. Yet partnerships have notorious command-and-control problems, as each local office is in effect a semi-independent business that can defect with its valuable clients if bothered too much by headquarters. So HQ often can only ask that local offices do something in a given way, not give orders. And the locals can blow off HQ. This was a notable problem for Andersen not only with Enron but in several other cases as well.

But nobody in a local office of IBM, GE or Microsoft ever blew off Gerstner, Welch or Gates, so letting corporations into the competitive market to conduct audit work, to get a little command-and-control discipline involved, might be a good idea too. Other professions (law, medicine, etc.) now function through professional corporations, so I don't see why audit accountants couldn't too.

Who gets the insurance payouts and under what conditions?

For example, if the stock price drops by 50%, does anyone get money? If so, who? If not, why not?

If the company goes out of biz, does anyone get money? Again, why, who, why not?

I suspect that the intended answer is "it depends".

One of the causes for the current audit failures is that CPA firms are viewed by the public (and the firms themselves) as insurance companies. Here's how it works now: Has something "bad" happened at the audited company? Would a properly conducted financial audit have prevented the problem? No? Too bad, the auditor is STILL liable. Since audit firms CANNOT protect themselves from liability by doing a good job, they spend less time and effort in doing the actual audit.

In full disclosure, I used to work for Andersen (but not on the audit side of the business) and one other major accounting firm. I noticed a MARKED difference in the attitude of auditors between 1987 when I started my career and 2000 when I left big accounting firms for good (I hope). In 1987, virtually all the auditors were proud of what they did and would have walked through fire rather than violate what they viewed as a sacred duty to the public. By 2000, many auditors were more cynical. While I think most would not have done anything they thought was clearly wrong, their loyalties had changed. By then, virtually all the major firms had been held liable despite doing what they thought was a good audit. The loyalty was now to themselves and not the public. Since the public was going to hold them liable no matter what, it made more sense to make as much money as possible and hope they got out before the house of cards fell down.

I don't get it, David. Are you saying that even though the auditor gets paid to do a thorough audit, they might not in fact do the job right because they think it's futile to do so? This sounds a bit like fraud; perhaps the companies who operate in this way ought to find another way of making a living.

Bottom line is the companies got paid to conduct an audit. Charging for providing a service and then not providing that service sounds like a good scam to me, but not one that a reputable company should indulge in.

I haven't yet read the linked article, but here are some thoughts.

A fair amount of "auditor insurance" already exists in the form of E&O coverage and some experimental coverage more along the lines of what you describe. The problem is scope of coverage and availability. For fun, read this story of insurance on accounting for a disastrous Lloyd's syndicate.

In theory, insurance companies are "the house" and win all the time. In reality, they write coverage at a loss most of the time (the industry's "combined ratio", or losses to premium is always over 100%). Competition and over-optimism about the return opportunities in the investment of reserves tend to keep these prices down. A "soft" market has been the norm for the last 30 years (apart from 1980, 1991-1993 and now) with insurer investments keeping the doors open.

We are now in a rare "hard" market for insurance. You can't get medmal, you can't get financial institutions E&O (including accuntants) etc. at a reasonable price. Customers are self-insuring these risks through captives or straight retention. This is a lousy environment for "audit insurance", and the industry would need substantial new capacity to write it at a reasonable price.

As an underwriter, it would be difficult to price and police the moral hazard in such a contract. Would insurance companies be better policemen than shareholders?

It's also too close to insuring a against your own criminality, which is illegal.

David Perron:

My original post may have been misleading. I did not meant to imply that auditors have been accepting compensation for doing nothing. In the past, the loyalty of the auditor was to the public and auditing firms showed a decided willingness to resign from an engagement rather than give a "clean opinion" on financial statements that were based on "aggressive" interpretations of GAAP. Over time, I noted that this loyalty had shifted. I saw auditors help their client company justify an aggressive accounting position. (Indeed, a significant portion of audit revenues came from such "consultation" on the "proper" way to account for certain transactions.) This is not to say the more aggressive accounting positions were not based on GAAP, it just the auditors put more emphasis on how they could help a client push the envelope than they put on keeping the financial statements well within standard practices.

In the end, financial accountants became more like tax accountants -- they focused more on finding loopholes than on preserving the pristine nature of the financial statements.

One big question, I think, would be whether such insurance is mandatory. Isn't there a case to be made that such insurance, held by stock owners against actions of managers and auditors, makes sense even without being manditory? The size of the market will be dictated by whether such insurance is mandatory, and the larger pool implied by mandatory insurance would probably keep costs down. It also strikes me that there must be some impediment to voluntary insurance, or there would be a larger market already. The point that insurance is an added cost is correct. It is also an added cost in the mortgage market, but I suspect most mortgage holders would probably buy some level of insurance even it it weren't required. There are benefits to insurance - risks are mitigated. Stock holders might also prefer to mitigate risks. Even with manditory insurance, some firms might chose to insure past the required point, to attract risk-averse investors.

I would think the illegality issue is can be taken care of since the insurance could be held by owners against the action of employees, like liability insurance. That seems to be the kind of activity that audit insurance would cover. Insurance would cover not one's own illegal acts, but the acts of another.

I think that insurance does have a role, but the proposed scheme isn't it.

I think that the appropriate vehicle is something that looks a lot like a mutual fund but it provides certain guarantees on return.

Such a fund will have the right incentives and will protect the correct parties. And, it's optional - if someone wants to go naked with their money, they can.

The tricky part is ensuring that these guaranteed mutual funds keep their promises. However, that problem appears to be solvable - insurance company failure is a managed problem.

Yes, I know that I've re-invented stock-backed insured annuities.

Naive question here.

Why not simply a law that says "You can be a licensed auditor, or you can be in the consulting business, but you can't be both and neither can you have a partnership or interlocking directorate with a consultant. Furthermore, if you're a publicly held company, you must have a licensed auditor do your audits"?

That shouldn't be such an onerous regulation.
After all, we don't (at least in CA) allow licensed termite inspectors also to engage in or partner in the building construction contracting business. The conflict of interest is readily apparent.

But the problem remains: the auditors get their revenue from the company, which gives them an inherent conflict of interest.

I was just midway through posting a long comment on how it's the shareholders and not ever ever ever company management who should hire the auditors... when I thought I might scroll up and actually read what the other folks here had written.

Baruch Lev stole my idea!

Responding to why not forbid the same business from doing both audit and consulting,
Jane Galt wrote:

"But the problem remains: the auditors get their revenue from the company, which gives them an inherent conflict of
interest."

Pressing the termite inspector / building contractor analogy further --

The property owner doesn't want termites found if he's selling either, but that doesn't stop the termite inspector from finding them.

That's where the licensing comes in. The licensed inspector can lose the license if he lies, even just once, at least theoretically.

Presumably an auditor's license would be something like a CPA license, a neutral examination of technical skills and swearing to some ethics code.

Dear curmudgeon: Auditors are already subject to the licensing and regulation you suggest. To issue an audit opinion on a public company, an auditor must be licensed by the state and is regulated by the SEC. (For all intents and purposes, this means the auditor must be a CPA, having passed a 4 day exam, and must meet continuing education requirements.)

The problem lies not with the skill or integrity of the auditors, its all in the question of to whose benefit will that skill and integrity be used? If an auditor's loyalty is to the audited company, the auditor will help the company groom its financial statements so the company is put in the best light. (Note: This does NOT mean the auditors are trying to help company's break the rules. But the economy is very complex and the financial accounting rules match that complexity -- meaning there a lots of areas where there is no clear cut answer as what the "right" accounting treatment may be. This gives auditors a lot of room to help a company look good with out clearly violating the rules.) On the other hand, if an auditor's loyalty is to the public, the auditor is liable to require a company PROVE its proposed accounting treatment is proper -- rather than allowing things to pass if it cannot be shown to be improper.

I like the idea of having shareholders hire the audit firm, but this is, in large measure is already being done. The audit committee, made up of outside directors, hires most of the auditors. Many companies require the selection of auditor to be approved by shareholder vote.

What's to be done? We need to get auditors to return to being more loyal to the public than to the company's they audit. That requires protecting auditors when they do their jobs properly. The current system punishes auditors whether or not the audit was done properly.

Why not require publicly traded companies to change audit companies every 5 years, as a precondition for being listed on the stock exchange? Seems to me that if an audit firm's competitors were guaranteed to be reviewing their work a few years down the road, that knowledge would be a pretty good incentive for probity.

Um. You'd still need some way to avoid collusion between supposedly-rival audit firms. Are there other downsides I've missed?

having worked for these firms can answer why legal restrictions are a bad idea...

for large firms, auditors are (almost) never given real consulting work (now)... maybe some tax stuff (and formerly IT), but no real strategy stuff... why the tax? cause they already did the audit and know what the hell is going on... every time you use an accountant for your own taxes he is essentially auditing you (try to get him to evade the law, he won't as its his own ass).. you'd need to spend incredible amounts of money to get corporate taxes done... which especially screws small companies that are or hope to go public (pre-ipo companies abide by all the public rules just for simplification and readiness)

small companies lean on their auditors for all sorts of stuff (along ith their lawyers) because they're the people they need to hire (do taxes, comply ith securities law...) and they'll give you a better rate (and the time of day) than other firms will, and you already know them (managing these relationships is expensive)

so yeah this regulation would be fine for large companies, but it screws the small guys who can't afford it... yet one more regulation that favours big business (and you wonder why lots of big biz votes dem?? Jon Corzine?? )

and as for why partnerships... well initially it was because they were unlimited liability... the auditors' houses were the insurance (we've been down this insurance road before, just that earlier we cut out the middleman) so that made sure that no one screwed around!

as partnerships had to get bigger, this caused problems, and LLPs are pretty much corporations... they're in the midst of restructuring how they work... but their's a lot more central control....

unlimited liability is also why lawyers and doctors were in partnerships... but again this is changing...

I have greatly enjoyed the discussion here and would like to add a couple of thoughts. A little history a few arcane rules and some personal thoughts. Sorry for the use of the pseudonym but I would like to be candid. In all the following I speak for my self only and not my firm.

Also I am a second generation CPA but am not a partner. I have had long talks with my father about what went wrong and when. So I hope to provide a little insight into this discussion.

First, in theory, the auditor is hired by the shareholders through the audit committee of the board of directors and not by management. Since the net assets of a company in theory belong to the shareholders, it is the shareholders who pay the auditors.

One of the problems of the past 30 years is that as corporations became very large, the overall influence of any one investor becomes smaller and smaller. While many years ago, the board was comprised of individuals with large direct interest in the companies the deteriorated over time. In the end, at Enron the board and audit committee consisted of physicians supported by contributions controlled by management, etc. Often the dot.coms boards consisted of venture capitalists and others who did not have a long term vested interest in the health of the companies but were interested in maximizing the IPO price of the companies.

As to the issue of why the government requires auditors to be a partnership, the original idea was that the individual members of the partnership would have joint and several unlimited liability for the torts of the organization. However, today most of the accounting firms are Limited Liability Partnerships. This means that unless a partner is directly responsible for a particular failure, the partner only has his or her individual capital in the firm at stake not their house, car, etc.

As to the issue of whether the loose structure and independence of the individual partners permits those partners to take excessive audit risks or to fail to execute their responsibilities, each of the big 5 firms has their own personality. Some of them are run like large corporations with very tight central control while others were true partnerships of equals.

In reading on the Enron situation, it appears that the audit partner was permited under Andersen rules to overrule the recommended accounting by the national experts. Based on what I have read, this practice was not permitted at the other 4 firms and I know that it is not permitted at mine.

One of the interesting bits of history is that prior to the mid-70s, the AICPA had a rule that essentially said, "thou shalt not poach thy fellow member's client" as a result changes of auditor were very rare, there was little price competition and the firms tended to minimize the risk of an audit failure. Then the FTC noted the lack of price competion and force the AICPA to change the rule. As fee pressure rose, the firms could no longer focus on minimizing the risk and also sought more profitable areas of service. Audits were reduced to a commodity and no longer completely covered the costs of the associated risks.

How do we fix this? It's a question that I have pondered. I don't think that insurance is a workable solution. How many insurance companies can afford to underwrite the risk of Microsoft? Given its level of market capitalization above liquidation value the risks would be enormous.

Similarly, mandatory rotation would bring a new set of problems. It often takes a couple years to learn the full picture of a copmany and to develop the expertise to understand their business. By rotating audit firms every few years there is a frequent period in which the auditor is just learning the business of the client. Frequent changes will also disrupt industry expertise and may result in more unintentional oversight.

I think that the Sarbanes bill is actually not a bad start. The auditor must work for the board and the board needs to be truly independent of management. The board has to take greater responsibility for the the company. Mr. Warren Buffet is probably the best example of an independent board member so we may have to relax the cloning rules in this instance.

One way that the profession could make audits more transparent is to require the audit firm to explicitly state what financial statement assertions were tested (e.g., the existence of off balance sheet financing arrangements) the nature of the tests performed (e.g., confirmation with underwriters of off balance sheet arrangements, contract terms, recalculation of residual risk)and the results of those tests. This document could then be filed with the company's 10k. This would permit financial statement users to evaluate the extent of audit testing and whether such testing was acceptable to the investor. While this may seem impractical, is currently performed for a special type of internal control audit commonly referred to as a SA 70.

Finally, I have one thing to get off my chest. If GAAP results in misleading financial statements, one of the AICPA's ethics rules (I think its either 501 or 503 but I don't have my professional standard here at home with me) requires the auditor to ask the client to vary from GAAP and the auditor must explain the variance in his or her opinion. So GAAP is not an adequate defence of misleading financial statements.

In the end, the auditors must retain high ethical standards and moral values or any changes to the system will fail.

>> After all, we don't (at least in CA) allow licensed termite inspectors also to engage in or partner in the building construction contracting business. The conflict of interest is readily apparent.

Giggle.

Surely no one believes that CA termite inspectors aren't in cahoots with CA construction folk, regardless of the law.

David Walser:

Thanks for clearing that up. That's a much more plausible scenario, although not much more palatable.

curmudgeon:

The key difference between the auditor/business relationship and the termite inspector/homeowner relationship is the termite inspector's got very little hope of repeat business next year from the same homeowner, except perhaps in rare instances.

Of course, you could always enforce auditor rotation, but that would likely raise howls on the business side and might result in other problems I'm too lazy to consider right now.

Another thing to consider is there are still tricks that can fail to register on an auditor's radar. Let's say a company books revenue on an item in one quarter and takes a return on that same item the next, without said item ever moving from its place on the shelf. This kind of activity really only works over the space of a quarter, but an auditor isn't going to catch it. If you know next quarter's going to be better (or have some indication it will) you can effectively borrow against future revenue and make this quarter's numbers look good, without anyone outside the company being the wiser.

Thanks to all who explained the issue(s) patiently,

I had mistakenly assumed only two interested parties. Opportunities for mischief dramatically with the number of interested parties.

Management not actually pursuing interests of shareholders?
A five Claude Shock rating, duh!

Explanations were excellent and clear. Thanks again.

Very interesting conversation. Thanks everybody.

Comments are Closed.