Auditors: Market concentration and their role - Economic Affairs Committee Contents

Further supplementary memorandum by Mr Timothy Bush (ADT 17)


Dear Lord McGregor

  I am pleased to be asked to give my remarks. I have watched, or read the transcripts of, every session including this session with Mr Davey, Mr Carter and Mr Hoban.

  I am happy to stand behind all of my evidence to date. I treat the Committee as I would a Court, the whole truth, not selective bits or half truths. I took all of my evidence from prime information and the primacy of law rather than hearsay. In this submission I have included prime law in appendices.

  I noted with great interest that Mr Hoban did say that regulation should start from the accounting numbers Q531. He is correct, and that is precisely what happened. The FSA regulated from IFRS numbers.

  The FSA took (imprudent and unreliable) IFRS statutory accounting numbers from banks and made regulatory adjustments (Basle) to that. The problem was that those adjustments were also imprudent and therefore did not negate the impact of IFRS where the banks were in difficulty (running faulty/risky business models). I believe that these adjustments in some cases merely amplified the problem.

  The French banks did not use IFRS, they used already prudent French GAAP, and then the regulator made regulatory adjustments (Basle) to it. However, those adjustments were only imprudent to the extent there was already a problem in those banks, but there was no problem to amplify.

  The difference between French GAAP+Basle and IFRS+Basle is that the poor regulation in itself does not necessarily mess up the running of the business. The regulator may regulate less efficiently, whatever that means, but the business itself is running on the correct ratios/KPIs if it is using prudent GAAP in first instance. Basically the directors, managers and shareholders can see what is properly going on in terms of proper business economics, even if the regulator then puts himself in the dark a bit. However, if you use IFRS numbers to run the business you may well get lost, and then run the wrong model, but appear profitable, with all the wrong ratios and KPIs.

  In maths, as an analogy, the regulation and the accounting are not commutative, like division and subtraction you get a different answer depending on the order of things. If an IFRS using business has a "softness factor" in it due to imprudent accounting, regulation cannot re-harden it.

  Those banks that failed due to bad debts (rather than the investment banking/Lehman crisis) tended to be former Buildings Societies (B&B, NR, HBOS and divisions of it such as Birmingham Midshires). I presume, and I did observe this from some analyst meetings, they seemed to be too dependent on IFRS, ie they were only running one set of books. Those lending banks that survived, HSBC-UK, Barclays-UK, NatWest and Lloyds-TSB, the former clearers, seem to have used a form of UK GAAP before applying IFRS (ie two sets of books, with the right ratios/KPIs). So did Abbey/Santander which essentially ignored IFRS. It is fairly easy to see why there was a systemic problem.

31 January, 2011


  Neither the BIS nor the Treasury minister addressed in Q540-547 (or elsewhere) that the law requires statutory accounts for the purpose of conducting the business of the AGM properly. That is a prudential purpose, a discipline, so that shareholders and directors know where their business is truthfully at. For banks at risk the audited IFRS numbers were unreliable for that purpose; profits were exaggerated and dividends and bonuses were paid out of unreal or reversible soft profits, for the banks to then collapse shortly afterwards, depleted of capital.

  Mr Carter hits the nail on the head without perhaps realising it (Q547), "the losses were immensely greater than all the bonuses and dividends that the banks had paid", this is precisely due to the gearing of a bank. False profits may be paid out or retained. Any false retained profit is false capital, and that multiplied >20 times—in gearing up—gives (false) over-lending capacity. Over-lending will be to more marginal borrowers. Eventual losses will be large, more than merely the already false capital will emerge as losses. Legally safe dividends are a function of safe reserves left behind ex-dividend.

  I dispute that there is no evidence on IFRS versus UK GAAP and banking problems (Mr Davey Q546). Lord Turner highlighted IFRS as a problem (a whole speech to ICAEW in February 2010 about "illusory profits"). There is also counterfactual evidence. No French bank came anywhere near collapsing, they kept prudent French GAAP for the accounts of banking companies, meaning normal financial governance (self-control) operated properly in first instance alongside prudential regulation (state intervention). The problem is not merely regulators getting it wrong after the fact, they do not run the business. The problem is a business doing the wrong thing but thinking it is on the right track. Banks are run on key ratios from the accounts.

  I can understand why BIS is defensive (and I can see this whenever "prudence" or "law" is raised). BIS legislated to allow the use of IFRS for banking companies. This was in part because the FSA, which had used UK GAAP for prudential banking regulation, signed off on IFRS for use for prudential banking regulation, whilst appearing to miss some of the major changes with IFRS, most notably:

    — the impact of the IFRS mark-up-to-market or model without the BBA SORP moderating it (Mr Carter does seem to acknowledge/allude to this omission in Q544 where he says "with the benefit of hindsight" when referring to the loss of the BBA SORP),

    — the impact of IFRS on bank profit and loss accounts. The FSA only adjusted the balance sheet for its identified IFRS impact, not profits, but IFRS has a disproportionate effect on profits due to gearing, this can easily increase "profits" by >20%, even more where there is increased risk, and

    — the impact of the IFRS bad debt provisioning model on balance sheets, particularly where there was very high risk lending (eg HBOS, Northern Rock, Bradford & Bingley). The FSA reversed out the deficient IFRS defined "incurred provision" to substitute it with a one-year—expected loss in the balance sheet. But, one-year-expected loss itself was also deficient, as shareholders in any limited liability company may walk away when total foreseeable losses exceed capital. One year expected loss was not much better than "incurred loss". It too rode with the boom, as it was also evidence based, rather than commonsense economics.

  In short, the FSA missed that prudential banking regulation did not work with IFRS. The FRC (under BIS) seems to have missed the dysfunctional impact on governance, pay and profits, of IFRS. The Treasury allowed IFRS for Building Societies. The Big-4 sought a global model (Mr Halliday-E&Y said so).

  Q540-547—To give some structure to my remarks, as it all interlinks, I have ordered this under four subjects that were covered in Q540-547. The subjects are:

    — prudence/losses/realisation, "UK GAAP" versus IFRS;

    — law/legality of IFRS as a preparation method (which the IASB/EU decided);

    — audit purpose and the output standard (which Parliament decides); and

    — going concern/expectation gap/"snap shot" balance sheets.

  All of the subjects interlink with going concern. Going concern requires reliable accounts, and vice versa. The law is structured so that AGM's can function (essentially on the nod) with a one year ahead view based on a truthful and reliable audited account of true condition, statutory accounts.

1.   Prudence and IFRS and UK GAAP (Mr Carter, Mr Davey and Mr Hoban's answers Q540)

  I see a good deal of implying equivalence of IFRS and pre-2005 UK GAAP. However, I have attached the Accounting Rules ("the Rules") from the Companies Act statutory instrument ("UK GAAP") as Appendix A for the benefit of the Committee. The Rules give prudence a statutory bearing.

  The law requiring prudence, Rule 19, is unequivocal, it is a must. I agree that one might not agree on the precise amount of it, calculating most things is subjective. But prudence is directionally clear.

  Mr Carter's evidence, Q544 states correctly that prudence was not defined in statute. However, the required accounting outcome from applying prudence is set out in objective form in statute as Section 830-837 of the Companies Act which links the audited accounts to the distributability of reserves. As assets back all reserves, including those left behind ex-div, I think it self-evident that asset values should be prudent and reliable rather than soft or speculative. Rule 32 (1) specifically requires bank loans to be carried at net realisable value where lower than cost. In other words its recoverable amount. Rule 19 sits over Rule 32, ie UK GAAP in law is a prudent estimate of net realisable value.

  Also, Rule 19 (b) requires negative post balance sheet changes to be adjusted. Mr Hoban's evidence though stated that balance sheets are a "snap shot". Rules 19, and 32 are forward looking beyond being a snap shot, they adjust things by requiring judgments. The same goes for inventory provisions. By definition any inventory left at the balance sheet date is goods not sold, inventory provisions are for goods not likely to be sold. That is a judgment.

  IFRS is more "snap shot" and evidential, and not the same as UK GAAP. I therefore wholly disagree with Mr Davey's comments and Mr Carter's then agreement with his minister (Q540). The FRC paper on auditor "scepticism" is in my opinion its way of resolving the prudence/IFRS problem without overtly admitting it. It is worthy of note that that BIS (nor the FRC) does not refer to scepticism in the context of the FRC's (excellent) auditing standard on going concern (see later).

1.1  Architecture of IFRS standards and the force of law (Mr Carter and Mr Davey's answers on IFRS and prudence, Q544)

  The answers here were incomplete or muddled. Each EU-IAS (EU endorsed IFRS) Standard is an EU legal instrument, ie the law. IAS 1 is the general "mother" standard of IFRS. However, its principles/rules, are excluded for areas where the subject matter is covered in another standard. For bank loans IAS 39 is sovereign.

  Furthermore, IAS 1 does not contain the word prudence at all, nor a synonym. What Mr Carter describes in his evidence as prudence in the "the Framework" (Q544) is not only irrelevant as IAS 39 is sovereign, but "the Framework" is not endorsed by the EU. The EU has quite remarkably incorporated something into law that is definitively cross referencing to something that is outside of the law. (Further, the IASB is currently changing its "Framework" to exclude mention of prudence from even "the Framework", ie neutrality instead).

  But more simply than all of that, IAS 39 has to exclude prudence. It is impossible to have mark-up-to-market—as in IAS 39—with prudence operative in IAS 39 itself, or from anywhere else in IFRS, whether IAS 1 or "the Framework", as prudence would counteract it.

  IAS 39 (para 59) states "losses expected as a result of future events, no matter how likely are not recognised". It then goes on to give in AG 90, as an example in practice, that even substantial credit risk from the death of borrowers, who have not died before the balance sheet date does not require provisioning. That exclusion of inevitable loss cannot possibly encompass the "net realisable value" test of Rule 32 in the Companies Act Accounting Rules (UK GAAP).

  Further, IAS 39, para 59, has the rather weasel words, "observable data that comes to the attention of the holder of the asset". "Coming to the attention of" is a qualifying term, a safe harbour sanctioning accounting that is missing things, but it is also a perverse incentive not to even look. The wording "holder of the asset" also seems to exclude things coming to the attention of the auditors. That to me is a limitation of scope, "I see no ships". Company Law is worded to avoid any definitive frame of reference. Truth is truth. The auditing issue should be digging sufficiently to get at it.

1.2  Mark to market (again on UK GAAP and IFRS having been similar Q544)

  The Company Law Accounting Rules, Para 32, is drafted such that "net realisable value" might be the lower of cost or market value, where there is a market.

  But the term "realisable value" is not defined restrictively to be a market value (as with IAS 39). UK GAAP is able to deal with the paradox of "price" and "value" and take a recoverable amount rather than a distressed price. IFRS was not able to, IFRS required "market" or a model of a market when markets were distressed. The EU then had to suspend part of EU-IAS 39 in Q4 2008. That is not an equality with UK GAAP in my view. Different rules, different authority/sovereignty.

1.3  "UK GAAP was an incurred loss model" (various evidence)

  Several bits of evidence have claimed an equality of IFRS and UK GAAP for bank bad debts, due to UK GAAP also being "an incurred loss model". I think that this is a well rehearsed wordplay on the indefinite article "an". The Companies Act Accounting Rules are not incurred loss, it is prudent net realisable value.

  The BBA SORP, is not saying what IAS 39 does, nor is it is a legal instrument able to override the Companies Act Rules. My understanding is that the BBA SORP was worded to exclude recessionary forecast in provisioning, such as Spanish economic cycle provisioning, the ultimate end of the spectrum of "expected" loss. The BBA SORP was not excluding such losses as might be expected by making 120% mortgage loans to people of poor credit standing. If there is a spectrum from incurred loss to expected loss, IFRS is at one extreme, Spanish economic cycle provisioning at the other, and "UK GAAP" (Companies Act Rules, including the SORP) was somewhere between.

  Indeed, Old Mutual plc, which has banking business (audited by KPMG), stated in its IFRS conversion statement that it used UK GAAP's expected loss model.

    Additionally, under IFRS, the Group has moved to an "incurred loss" provisioning model within its banking segment. Under UK GAAP, the Group utilised an "expected loss" provisioning model.

  Another example of the clear difference on transition to IFRS, the Nationwide Building Society, Annual Report and Accounts, 2006 stated:

    "52j The net impact of more stringent evidence testing required by IFRS has resulted in a decrease in the carrying value of loan provisions of £86.0 million."

2.   The legality/conformity of IFRS with the purpose of accounts (Mr Davey' specific answer Q544)

  Mr Davey, was asked specifically about my evidence. But he seems to misunderstand the question around "conformity of law". I was not saying IFRS is "illegal" which is how Mr Davey has answered the question.

  Law is law and IFRS is law. But one part of law can clash with other law. Statute can clash with other statute or common law. The issue is not IFRS being "illegal" the issue is law not working. Tax law is notorious for inconsistency, and hence creating overlap or loopholes.

  The purpose of audited accounts set out in IFRS and also the purpose against which the EU assesses an IFRS for adoption does not accord with the purpose of accounts set out in UK Company Law, which is corporate financial governance ("stewardship"). The purpose in English Law is set out in the Caparo case which describes the statutory position (see Appendix 3):

    (i) audited accounts protect the company (such detecting frauds and not paying dividends out of capital, Companies Act 2006 Section 830-837, which is a statutory articulation of the common law), and

    (ii) audited accounts inform the members at the AGM for the AGM to properly hold the directors to account and pay them properly for performance (even though this is retrospective, it takes reliable profits (and ratios) to do it).

  This law "concatenates-pyramids", ie if a if company owned by another company pays a dividend, then the above model applies. eg HBOS plc's board (as shareholder) could rely on the audited accounts of Bank of Scotland when Bank of Scotland declared its 2007 dividend, and paid it up to HBOS plc in February 2008.

  Bank of Scotland has now had had >£28 billion injected into it but the y/e 2007 accounts had justified a final dividend of £1.2 billion off capital and reserves of £18 billion. The accounts were not at all reliable for the AGM (nor the acquisition by Lloyds), nor had they been for expanding lending. Total losses since the 2007 Accounts were signed off have been £30 billion.

  IFRS has a very vague stated purpose (see Appendix 4). IFRS is concerned with ill-defined "users" for an ill-defined purpose which is "to be useful for users". It is an inconsequential definition. It came from the litigation conscious USA. (See going concern later).

  Usually if statute clashes with common law statute wins, or if statute clashes with statute the latter statute wins by the doctrine of constructive repeal. However, the purpose of accounts in English law, is expressly not overriden by the EU Regulation under Article 5. The EU and the Department of Business confirmed this prior to the adoption of IFRS. However, IFRS in practice frustrates delivery of Section 830-837. This is evidenced clearly by the bad debt provisioning model not being prudent or mark to market accounting not addressing net realisable amounts. It is directionally wrong to be prudent.

  In summary there was no functional mechanism to deliver the law, as the law is at cross purposes. This incompatibility can be seen in the ICAEW/ICAS legal advice on applying IFRS with Company Law. From 1982 to 2005, the ICAEW guidance on distributable reserves was three to four pages. Under IFRS it is >120 pages and increasingly confused. It is clear from reading it, that one leg of the law (Sections 830-837) requires prudence, whilst IAS 39 is imprudent. I question whether the BIS Minister was made aware of that. I am happy to supply it to the Committee.

  However, there is a legal problem with UK-Eire FRS 26 as part of "UK GAAP" (ie not part of IFRS). ASB standards are not law, and must conform with the law or state their deviation from it so that the impact is shown nevertheless. I cannot see how the paragraphs in FRS 26 which have been copied straight from IAS 39 can possibly conform with the Companies Act Accounting Rules. In that case, this is not a clash of law, the inferior body (the Accounting Standards Board) is contravening the law of Rules 19 and 32.

3.   Going concern and "expectation gap"

  Lord Hollick had asked about auditors needing to be more sceptical on going concern (Q531) and Lord Forsyth on the risk of carrying 40 times gearing (Q537).

  I was perturbed by what Mr Davey said about there being an inevitable "expectation gap" with auditors when he was asked about going concern. I would also refer him and the Committee to the Barings plc and Baring Futures Singapore cases. They tend to betray any expectation gap (both were settled under English Law). I do not know what "literature" he refers to on "expectation gap" but if any of it is from the USA, it may well not accord with English law. US auditors contract essentially for reporting for the market, not for the company. The "market" can be ambivalent on going concern.

  Going concern is the economic phenomenon of the company being able to stand on its own two feet, ie shareholders are funding it, are standing behind it and will continue to do so. Members have the prerogative to wind up a solvent company, and administrators insolvent ones. New equity will only be forthcoming when it is economically beneficial to put new money in. Shareholders will not rationally do if their new subscription is firstly funding excessive losses.

  IFRS leaves out losses and included unrealised and unrealisable profits. (Q 546, Lord Lawson, hits the nail on the head, so does Mr Carter, Q547). IFRS accounts in my opinion can be unsuitable for assessing whether a company is a going concern or for raising new capital.

  Further I note that KPMG's evidence on going concern was not consistent with the UK and Irish auditing standard ISA 570 that it claimed to be quoting from. ISA 570 does not allow a general presumption about the availability of capital whether debt or equity regardless of source or circumstances (KPMG supplementary evidence) when an auditor signs off the accounts. That would be illogical given that the audited accounts are supposed to be reliable enough for shareholders to take money out, as a final dividend. Shareholders vote on final dividends, based on the audited accounts after all.

  Further, the KPMG statement "regardless of source" is at odds with the pre-emption regime in this country. Shareholders have first refusal, and to exercise that right without being misled they require accounts that are true and fair, ie they are not being kept in the dark about hidden losses and/or loss making trends.

  Requiring new capital to be injected is fair to satisfy the going concern presumption if it is flagged to shareholders—the members—that new capital is required, such as with a co-terminous underwritten rights issue. In such a case the auditor opinion, and also the prospectus, should still look at the sufficiency of these additional capital resources at least a year ahead. (Or, for a subsidiary that requires support, binding parent company support).

  The FRC/APB auditing standard ISA 570 on going concern is strong, and is aiming to correct IFRS. It sets out quite clearly that [probable] future losses/asset write downs are a going concern risk, they are a loss of capital. But, there is no operative accounting standard with IFRS (or FRS 26) to pick that up.

  For a group holding company to be a going concern, the subsidiaries need to be going concerns, but some subsidiaries may rely on parent support. The only way to properly assess whether a group holding company is a going concern (and not being stressed by some subsidiaries) is if the group and the subsidiaries prepare prudent accounts. IFRS has upset that model, even in respect of the mandatory EU requirement for consolidated group accounts.


  I also submit, as it has not come out in evidence, that UK GAAP (the Accounting Standards Board element) was able to react to new issues within weeks (following the Dearing Review of 1988 any issue could be fast tracked). The IASB takes up to three years even when there is a known problem eg incurred loss and the EU can take two or more years on top of that after that.


  In particular I note your question to Mr Davey Q544 which referred to my evidence. I believe that the minister took your question about "conformity with the law" to mean "was IFRS illegal?"

  However that matter is bad law/clash of laws.

  IFRS as an EU legally sanctioned preparation method (under the option to use imprudent IFRS), versus the capital maintenance/solvency clauses of the Companies Act as an output requirement dependent on prudence. A clash of law in itself is not illegal, and that is how Mr Davey correctly responded. But he did not address the conflict of laws.

  Further, he did not answer your actual question which was the conformity of accounts prepared under the "statutory instrument", ie the detailed Accounting Rules for Companies Act accounts "UK GAAP", with the capital maintenance/solvency clauses of the Companies Act.

  However, the specific problem with "UK GAAP" in your question is that FRS 26 (the Accounting Standards Board copy of IAS 39) does not accord with the Accounting Rules. The ASB is supposed to set accounting standards in accordance with the Companies Act in both senses, complying with the Accounting Rules (input) and the capital maintenance/solvency clauses (output).

  This is not a matter of UK law clashing with EU law but is a matter of FRS 26 as a UK standard not conforming with UK law; neither the Accounting Rules of UK GAAP nor the capital maintenance/solvency clauses. The ASB thus unnecessarily copied the EU/UK law mismatch to create a UK/UK mismatch (and Eire/Eire).

  Bank directors applying IFRS (IFRS using companies) or FRS 26 (non-IFRS using companies) may not comply with Sections 830-837, the capital maintenance/solvency requirements. If that is so they will not have discharged their solvency obligations, but they may very well think that they have, as may their auditors and everyone else.

  Rather than being a counter to IFRS, like French GAAP was, FRS 26 made UK GAAP equally bad, Hobson's choice.

previous page contents next page

© Parliamentary copyright 2011