The trouble with Audit (very briefly)

?You may not care too much about the audit of big companies, but you probably should. Chances are, their shares are where a fair chunk of your pension pot is. It’s been a hardy perennial of competition and regulatory concern, and there have been a suite of reports and recommendations recently, with more to come. This note highlights why ensuring good auditing is quite so hard.

Unreliable agents

James Bond: Do you expect me to talk?

Auric Goldfinger: No, Mr Bond, I expect you to die!

Bond: You’re forgetting one thing. If I fail to report, 008 replaces me.

Goldfinger: I trust he will be more successful. [1]

Audit is an elaborate and anguished example of the agent-principal problem. Put simply, if you ask someone to do a job for you – how can you be confident they are doing it well, and are not profiting at your expense? Problems become more acute where the agent knows more than you do, and their interests don't align with yours. Big company audit has the most agents outside a Le Carré novel. The original principal is the shareholder, and the original agent the company director. Then auditors were sent in to check on the director. And then company Audit Committees were set up to check on the auditor, and then regulators to set standards and check the auditors too, and perhaps now also the Audit Committees. But then, if you can’t trust your first agent, why should you be able to trust any subsequent agents sent in to check on the first one? The information asymmetries that cause the first principal-agent problem are the same for all agents. The little old lady ended up swallowing a horse, but it didn’t resolve her issue with the fly, or at least not in a good way.[2]

Are we being audited now, or what?[3]

Unlike the death of Mr Goldfinger, whether or not a company has been audited well is hard for the outsider to determine. There will always be a good-looking audit report, written in careful technical language, but whether there has been a thorough, impartial and wise appraisal by experts in the field, or a bleary box-ticking exercise by exhausted, indifferent juniors, can’t easily be seen.[4] It might make no difference, if the company has itself produced a good set of accounts. The only way to see if an audit has been done well is to send in another team of auditors. See above.

No one really cares

You may have no exposure to the fortunes of big listed companies. You may have some, but only remotely via a pension fund. You may have more, via a shares ISA or personal investments in funds, or you might be one of the hardy types who pick your own shares. Along that scale you will have an increasing interest in ensuring that the accounts of any one company are an accurate reflection of the state of its business, so that you can decide whether to invest. But likely, you will also have an increasing awareness that public accounts might not tell the whole story, and the instances of companies collapsing unexpectedly are sufficiently common that you may have priced it in to your investment strategy. This indifference, combined with the lack of visibility of quality (above), could allow the bad to drive out the good and unreliable company accounts to proliferate.[5]

Accurate accounts matter hugely for our economy and fortunes collectively, but not very much for any one individual. If you do exert yourself to ensure a good audit, then others benefit quite as much as you do. To address this free-rider problem and ensure there is a bell on the cat, audit is mandatory and regulated, so it must and does happen. But does it happen well?

Oligopoly

Oligopoly is a well-known hard problem for competition authorities to deal with. Without any trips to Switzerland or the golf course (the traditional places for cartel meetings), oligopolies can gently raise their prices/reduce their quality to monopoly levels and compete less intensely. But most competition lawyers would probably fancy their chances of persuading the authorities to allow a merger down to 4, which is the current number of auditors offering big company audit. There are additional issues to interest the competition folk, such as high switching costs, barriers to entry, and learning curves which can dampen rivalry. 

Hey DJ! I can’t dance to that music that you’re playing![6]

Or, he who pays the piper, calls the tune. The huge conflict of interest, that has been inherent to corporate audit since its inception, is that companies choose, instruct, and pay for their own auditor. Shareholders can vote against an auditor appointment at the AGM - but they almost never do [7]. So, just how much of a thorough investigation do companies (or even their audit committees) really, really want? The more thorough the investigation, the more expensive, intrusive and disruptive it will be. Companies’ core businesses are not supplying company accounts: they don’t and can’t compete on the quality of the audits they commission. They might benefit from an extra-thorough review of their accounts by an external auditing firm (particularly if they want to expose rather than conceal wrong-doing or overly optimistic reporting). But their consistent and pervasive incentives are to minimise costs and so to seek an auditor who will do what they have to do (which at its most basic is to meet the minimum standards set by regulation), as efficiently and discreetly as possible. Auditors may scrupulously (enough) follow the rules (which are necessarily subject to interpretation and judgment), have their reputations to think of, and are meant to be 'sceptical'. Yet they too have hungry mouths to feed and mortgages to pay…

Despite the occasional tiger-sighting, this elephant has been grazing contentedly in the corporate audit field since 1844. [8] 

Disclaimer: all views are strictly personal.

Disclosure: I was Inquiry Director on the Competition Commission’s 2013 investigation. https://assets.publishing.service.gov.uk/media/5329db35ed915d0e5d00001f/131016_final_report.pdf

[1] https://www.youtube.com/watch?v=wzwPI1zJ9K0  Go on – treat yourself!

[2] https://www.youtube.com/watch?v=V6Mzqxxio_w This is really quite disturbing – and apparently aimed at children.

[3] Another stolen title, from an interesting article on definitional difficulties in a quite different area, by Greta Christina, available in various places including ‘The Philosophy of Sex’ edited by Soble and Power, 5th edition, page 23.

[4] It is a ‘credence good’ – i.e. one whose true quality is unknown at the point of use. A former colleague thought face cream was a credence good, too.

[5] The classic account of how bad second-hand cars (‘lemons’) drive good ones out of the market is here, although its vocabulary is, and attitudes are, seriously dated: https://wwwdata.unibg.it/dati/corsi/8906/37702-Akerlof%20-%20Market%20for%20lemmons.pdf

[6] https://www.youtube.com/watch?v=aKvq8V29dE0 

[7] https://www.ft.com/content/a8c97aa2-5469-11e8-b24e-cad6aa67e23e Shareholders clearly have most to lose, most directly, from inaccurate accounts. Why so disengaged?

[8] https://assets.publishing.service.gov.uk/media/5329db3740f0b60a73000027/131016_final_report_appendices_glossary.pdf