Confidence Accounting - A Bold Proposal from Long Finance, ACCA and CISI
Confidence Accounting is a proposal to use distributions, rather than discrete values, where appropriate in auditing and accounting. In a world of Confidence Accounting, the end results of audits would be presentations of distributions for major entries in the profit & loss, balance sheet and cashflow statements. The proposed benefits of Confidence Accounting include a fairer representation of financial results, reduced footnotes, more measurable audit quality and a mitigation of mark-to-market perturbations. Today sees the publication of a landmark, free-to-download report:
Confidence Accounting: A Proposal
by Ian Harris, Michael Mainelli and Jan-Peter Onstwedder of Z/Yen Group
published by Association of Chartered Certified Accountants (ACCA), Long Finance and the Chartered Institute for Securities & Investment (CISI) July 2012, 63 pages
Andy Haldane, Executive Director for Financial Stability at the Bank of England welcomes the proposal and writes in the foreword, “My hope is that this proposal moves our thinking a step closer towards a set of accounting standards for major entities that put systemic stability centre stage. In the light of the crisis, anything less than a radical re-think would be negligent.”
For a quick introduction to Confidence Accounting, particularly its application to banking, in 2011 the CISI published “Accounting for Confidence” which provides a short overview. Confidence Accounting was introduced for the non-professional in “The Price of Fish: A New Approach to Wicked Economics and Better Decisions”, winner of the 2012 Independent Publisher Book Awards Finance, Investment & Economics Gold Prize.
The report’s lead author, Professor Michael Mainelli, Executive Chairman of Z/Yen Group, said: “We wrote this report in order to suggest that accountants and auditors could provide greater social benefit by moving towards better measurement science. Already, the intense pre-publication reaction shows that professionals realise the need for debate, if not all the reforms we’re suggesting.”
ACCA, CISI and the authors have established a consultation period till the end of 2012. Long Finance intends to host a number of autumn discussion events during the consultation period where people will be welcome to put forward their thoughts, criticisms and suggestions – www.longfinance.net.
http://www.financialdirector.co.uk/financial-director/opinion/2219142/fd-guest-column-confident-accounting
Some questions as a non-accountant (forgive me if the answers are trivial, but please tell me them anyway)
1) How do we make responsibility for the errors “stick” to the account’s authors?
Imagine an accountancy firm with 100 sets of accounts. Each has “On the basis of the valuation of assets and liabilities the directors estimate that the uncertainty over the income during the next 12 months is within a range of £ +/- x at 90% confidence level” written in.
A year later the accountants are faced with the fact that only 40 of their forecasts are within range rather than the 90 expected. They will say “Ah, but these accounts don’t take into account unexpected microeconomic changes in the last year, so we can’t be blamed that the actual profits didn’t fall in the predicted profit ranges from last year.” They will almost certainly be right in this statement. But if this is the case, what is the point of point of publishing accounts with over-optimistic confidence bars if they can't be tested?
2) What will the management think about with accounts from firms with wide errors? Some firms will have genuine wide errors (eg you mentioned firms owning Russian oilfields). Indeed the whole point of confidence accounting is to point this out to investors. The problem is:
a) Why bother to pay the accountants? If the accountant says “Well I predict a profit of between -20 and +100 million pounds next year…. by the way my bill for telling you that is fifty thousand pounds”. The managers will say “Well I could have told you that myself, what was the point of paying you to tell me the bleeding obvious?”
b) Wide error bars make it look as if the managers are not in control of their companies. They want everyone to see a firm hand on the tiller. “No”, they want to say “employ me and I will increase profits by 30m from last year, you can be sure of that.”
3) People are poor at estimating errors and lose status for making wide estimates. As Michael quoted from Bertrand Russell “The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser men so full of doubts.” This will also push to overly narrow error bars. This is linked to the overconfidence quiz I mentioned in a previous email.
All this means that I predict that confidence accounts will tend to have optimistically narrow error bars, and accountants will be able to slip out of responsibility for this by blaming macroeconomic conditions.
So what do we do about this?
If you have not yet had time to pass on your feedback, or were unable to attend yesterday’s event, we would be grateful if you could take a moment to complete the online questionnaire so that we have as much information as possible to include in the final report on the consultation period - http://zyen.info/Qv3/QPyIn.py?qrid=6068440619018418&state=new/
In the document we didn't have the time to cover the many ways this could go. A book on the subject (perhaps if momentum builds) would have examined in detail Bayesian versus frequentist thinking, developed more types of distributions (especially binomial distributions), had more case studies (manufacturing, energy), explored the implications of discrete versus continuous distributions, and commented more on account users' utility functions. And that's for starters. The book is there if people believe this path is worth following.
The consultation period is until the end of the year and the publishers, Long Finance, ACCA and CISI, intend to host a few discussion events - but let's start some of that discussion here.
This site is part of the Z/Yen CommunityZ