An assessment of the implications for competition of a cap on auditors' liability
July 2004
OFT741
1
1.1
EXECUTIVE SUMMARY
This report contains the OFT's advice to Government on the implications for competition in the audit market of proposals to permit auditors to limit their liability by way of negotiated caps.
The current liability position
1.2 At present auditors may not limit their liability for fault due to negligence or incompetence in audit work. Alongside regulation and reputation, liability acts as a discipline on audit quality in a context where shareholders and other third parties rely on information from an audit which is paid for by the company being audited. We are not aware of evidence suggesting that the courts in the UK have made, or are liable to make, excessive damages awards against auditors. Professional indemnity insurance is available, and LLP status – the chosen corporate form of many audit businesses – exists to protect partners' personal assets.
The competition arguments
1.3 It has been argued that a cap on auditors' liability would however be procompetitive because it would lead to: • • • 1.4 a reduction in the barriers to entry and growth facing smaller audit firms the maintenance of competition between larger audit firms, including for non-audit work, and less risk of collapse of one of the Big Four.
On the basis of the evidence available to us, however, none of these arguments appears compelling. First, the current liability position is a minor barrier to entry in comparison to reputation, third party perceptions, economies of scale, global networks, regulation, and various other impediments to the entry and growth of smaller audit firms. The liability position is symmetric as between all audit firms, so the introduction of a cap would not
1.5
Office of Fair Trading
1
appreciably alter the relative positions of the Big Four compared to other auditors. 1.6 Second, unlimited liability does not appear to be causing Big Four firms to withdraw audit services. Audit, and the non-audit services sold to audit clients, are valuable business that the Big Four firms appear keen to retain. Third, as in the case of Andersen, reputation damage rather than financial damage is likely to be the primary cause of collapse for a Big Four firm implicated in a major audit scandal, should that ever occur in future. If the claims are generated abroad, a cap on liability in the UK would offer little protection for an audit firm's global network.
1.7
Cap design
1.8 Lack of adequate safeguards and some forms of cap design – for example caps that afforded proportionately more protection to large auditors – could distort competition.
Conclusion
1.9 It is likely that allowing audit caps would be competitively neutral overall. Arguments that allowing caps would be pro-competitive are not compelling. Some forms of cap design could distort competition, so it will be important to ensure that there are no anti-competitive effects if scope for caps is allowed.
Office of Fair Trading
2
2
2.1
BACKGROUND
Following the Enron affair and subsequent collapse of Andersen, the OFT announced that it would keep the accountancy and audit market under review and would advise on any competition implications arising from any proposals for regulatory reform.1 The Department of Trade and Industry (DTI) recently consulted on the issue of permitting auditors to limit their liability (the Consultation),2 with a cap on auditor liability one of the suggestions put forward. Given the Government's objective of improving quality and competitive supply in the audit market, DTI asked the OFT to advise on whether a negotiated cap would significantly affect competition in the audit market, and, if so, how. The OFT was asked to consider the issue against the impact of the measures already taken and being considered to extend the coverage of the audit and improve audit quality. Given the limited time available, DTI recognised that the OFT's advice would be based on the responses to the Consultation and other readily available information. We reviewed the responses to the Consultation alongside fee data, market reports and the available annual reports of the Big Four, and compiled evidence to support or refute each potential impact on competition from the policy proposals. There is necessarily an element of judgment in our conclusions, given the time constraints, the lack of 'hard' evidence in the responses, the scarcity of historical data for the accounting firms and the requirement to consider the proposal alongside a package of forthcoming regulatory changes.3
2.2
2.3
OFT Press Release PN 79/02, 22 November 2002. Director and Auditor Liability: A Consultative Document (DTI, December 2003). 3 Factors such as individual regulations, reputation and competitive dynamics all interact and need to be considered as a whole.
1 2
Office of Fair Trading
3
3
3.1
THE CURRENT LIABILITY POSITION
At present auditors may not limit their liability for fault due to negligence or incompetence in audit work. Alongside regulation and reputation, liability acts as a discipline on audit quality in a context where shareholders and other third parties rely on information obtained from an audit that is paid for by the company being audited.
The context
3.2 The provision of audit services is subject to market failure. First, there is information asymmetry. Most shareholders are not in a position to gauge the quality of audit services and are unable to observe whether an individual audit firm has cut audit quality. Second, there are externalities. The managers who selected the audit firm may have done so on price rather than quality, or may have been implicated in financial wrongdoing. In the event of a poor audit there are significant negative repercussions for those 'external' to management: for example shareholders (including pension fund contributors), creditors, potential investors, potential creditors, suppliers, employees and customers. Consequently, there is a demand for independently verified company information and a statutory requirement for larger firms to have an annual independent audit. Regulatory and non-regulatory mechanisms (reputation and liability) act as disciplines on audit quality in the presence of this market failure. In terms of regulation on the 'input' side of quality, statutory audit is restricted exclusively to qualified members of a number of designated chartered accountancy bodies. There is a system of self-regulation supported by statute and by government involvement with accounting and auditing standards (on the input side), and investigation and disciplinary procedures (on the output side). Following a series of reviews of the regulatory framework for accountancy,4 the Financial Reporting Council (FRC) has recently been designated as the new single regulator for the accountancy profession (funded in equal parts by Government, firms and
3.3
For example: the EC Financial Services Action Plan looked at audit independence (May 2002); the OFT report Competition in professions (OFT 328, March 2001); the FSA review of the UK Listing Rules (looking at auditor independence, 2001); a Treasury Select Committee (July 2002); the Coordinating Group on Audit and Accounting Issues, January 2003); and a government review of accountancy regulation (January 2003).
4
Office of Fair Trading
4
the profession itself). New regulatory initiatives aimed at enhancing audit quality (overseen by the FRC) include: • • audit partner rotation (a five year maximum for the audit engagement partner, seven years for other key audit partners) the creation of an Audit Inspection Unit (under the Professional Oversight Board for Accountancy, part of the FRC) to monitor the audit of major listed companies in terms of the audit process and decisions taken by the auditors clearer ethical standards, e.g. on non-audit services (one of the proposed ethical standards proposed by the Auditing Practices Board (APB) is to impose a 10 per cent threshold for any single client as a proportion of an audit firm's revenue before economic independence is considered compromised) an enhanced role for audit committees with guidance from the FRC on composition, terms of reference and resources a two year cooling-off period announced by the Institute of Chartered Accountants in England and Wales (ICAEW) for audit partners planning on leaving to join an audit client improved disclosure (e.g. of fees from a listed client when they exceed 5% of total fees received by an audit firm) published annual reports from audit firms (three of the Big Four have done so to date) EU adoption of International Accounting Standards by 2005, and collaboration between the Accounting Standards Board (ASB) and the International Accounting Standards Board (IASB) to further develop standards, and, a recommendation by the Coordinating Group on Audit and Accounting Issues (CGAA) that a proactive stance is taken by the Financial Reporting Review Panel in terms of enforcing standards.
•
• •
• • •
•
3.4
Two non-regulatory mechanisms also play a role in maintaining audit quality: reputation and liability for damages. A history of repeated business without problems creates a reputation that can help overcome the asymmetric information problem outlined above. Reputation is an important factor in the audit market because a statutory audit is a service bought regularly by a large client, where the purchasers (management) can measure 'output' (not necessarily the same as the 'quality'
Office of Fair Trading
5
shareholders desire) and the audit firm benefits from the client acting on this knowledge to make repeat purchases. A dispute involving shareholders over the quality of an audit damages reputation and the ability to win future business and retain existing business. 3.5 Liability also helps to maintain audit quality, given the prospect of severe financial penalties in the event of an auditor being found at fault due to negligence or incompetence leading to harm. Audit has an outcome that is relatively easily measured and linked to inputs, so the threat of being sued is likely to be an important discipline on quality alongside reputation and the regulatory factors already discussed.
The case for change
3.6 We are not aware of evidence suggesting that the UK courts have made, or are liable to make, excessive damages awards against auditors. The scope for damages claims against audit firms in the UK is already quite narrow (see, in particular, Caparo v Dickman),5 and is quite different to the situation in the USA. Under English law it is usually only the company which can properly act as plaintiff, with no automatic liability to individual shareholders, those making investment decisions or to creditors, for example. The DTI states in its Consultation that it is minded to accept the Company Law Review recommendation that there be no changes to the existing scope of auditors' liability. The majority of respondents to the Consultation agree with that position. The arguments in favour of a cap on auditors' liability focus on the potential pro-competitive impacts rather than any shortcomings in the way the courts handle claims.
3.7
5
Caparo Industries plc v Dickman, House of Lords (1990).
Office of Fair Trading
6
4
4.1
THE COMPETITION ARGUMENTS
It has been argued that a cap on auditors' liability would be procompetitive because it would lead to: • • • a reduction in the barriers to entry and growth facing smaller audit firms the maintenance of competition between larger audit firms, including for non-audit work, and less risk of collapse of one of the Big Four.
4.2
On the basis of the evidence available to us, however, none of these arguments appears compelling. The relevant market in this case is taken to be the provision of statutory audit services to quoted and large companies in the UK (see, for example, two recent EC competition cases involving large audit firms.)6 This market is characterised by a very high level of concentration, with just four incumbent accountancy firms (the 'Big Four'), and a significant size gap between the fourth and fifth largest firms (see Figure 1). The 20 largest firms in the second tier of auditors are known as 'Group A'.
4.3
EC cases M1016, Price Waterhouse / Coopers & Lybrand, 1998; M2810, Deloitte & Touche / Andersen (UK), 2002.
6
Office of Fair Trading
7
Figure 1: Ranked share of total fee income for the top 50 audit firms (2003/04)
25%
20%
15%
10%
5%
0%
D el 1 P oi tte wC & Er 3 T LLP ns KP ou t & M ch 5 Y G e 6 BD Gra oun LLP O nt g L St Th LP oy or H nt 7 ay on Ba w 9 Sm ke ard rT i 10 th & ill 12 Te W 8 P y R SM 11 no illia KF n m R Mo Gr so ob or o n so e S up 14 n p R tep lc H ho he or w de n at h 13 s L s C la M LP 15 rk az N W ar 1 u hi s 17 6 H me teh 18 B ai rica ill Sa ent nes pl ffe ley W c 20 r J a C 19 y C enn tts ha K h nt in am ison re gs p n y Ve ton ess 22 lla Sm c U H 21 ott ith Y H Va DFK a n 24 23 cke tis p M Me r Y lc ac n ou 25 in zi n e 26 W tyre s M g S ilki H R 27 olo ns ud I Ar mo Ken son m nH n e 29 2 stro ar dy Jo 8 H ng e L hn ay W LP st sm ats on a o C cin n a t 32 31 3 rm yr R Re 0 B ich e ot e hm ve uz ael an s & zac 3 o 33 3= Pan Ney tt = Lo ta la C ve ll n oo w & pe el Co 35 r P l Bl Fr arr ake an y ci LL s P 37 Sc 3 Cla ot 6 rk tM W 39 38 on BS Li Pric cr t i 40 tlej e B eff La ohn aile rk F y in ra g se G r ow 42 e 44 Ha 41 n M zle DT er w E ce oo 46 r & ds H H LB o AV 45 le A CL 48 4 ud B 50 M 7 it Be 49 or Ha plc rg Bi ley sle Ka sh & rs pr op Sc ow Fl ot Le em t w ing is LL P 4
Source: Accountancy Age website (www.accountancyage.com/Top50).
4.4
2
However, for the purposes of market shares in the relevant market, audit fees received from firms in the FTSE-100 and FTSE-350 indices provide the most suitable categorisations (see Figures 2 and 3 below). In addition to the high concentration, choice is further limited by conflicts of interest and stricter independence rules (particularly in some specialist sectors such as banking and insurance). Audit firms have a captive market (demand is inelastic) due to the statutory obligation to have an audit, the small size of audit fees relative to the turnover of the largest firms and switching costs from changing auditors. Competition for audit contracts occurs at the time of a competitive tender (where two to four firms are invited to bid), and imposes a competitive constraint on long-term audit firms (of which there are many). The EC competition authorities found that market
4.5
4.6
Office of Fair Trading
8
Figure 2: Market shares by audit fees for the FTSE-100 (2002/03)
Deloitte 18.0%
PwC 42.4% E&Y 11.1%
KPMG 28.5%
Figure 3: Market shares by audit fees for the FTSE-350 (2002/03)
Other 0.5% Deloitte 18.4%
PwC 39.5%
E&Y 15.0%
KPMG 26.5%
Source (Figures 2 & 3): Financial Director, January 2004. Excludes investment trusts (apart from 3i) & two E&Y clients.
Office of Fair Trading
9
shares do not reflect the ability to win such tenders. Evidence provided to the EC (and also evidence from audit clients reviewed in the OFT's report on the professions)7 suggests that competition between incumbents is strong. The EC left open the question of whether or not there existed a situation of 'oligopolistic dominance'. 4.7 Potential competition for the relevant market is restricted due to significant barriers to entry, some of these barriers resulting from the same factors that help to discipline quality (regulation, reputation and liability, discussed earlier). We consider barriers to entry more fully in paragraphs 4.12 to 4.15 below. We now assess, in turn, the arguments advanced for the potential procompetitive impacts of a cap on auditor liability.
4.8
The impact on barriers to entry and growth facing smaller audit firms
4.9 If the current liability position significantly impedes potential new entrants to the relevant market, the introduction of a cap on the liability of auditors could be pro-competitive. Actual new entry, or the realistic threat of new entry by current Group A firms, would enhance competition in the provision of statutory audits for the largest firms.
4.10 Our analysis, however, does not support this argument. The reasons are as follows: • unlimited liability is, at most, a minor entry barrier in comparison to other impediments to new entry: for example reputation, third party perceptions, economies of scale, global networks and regulation. In the presence of these other significant barriers to entry, capping auditors' liability is not likely to have any impact on competition in the provision of audit services to the largest firms the liability position is symmetric as between all audit firms, so the introduction of a cap would not appreciably alter the relative positions of the Big Four compared to other auditors
•
7
Competition in professions (OFT328, March 2001).
Office of Fair Trading
10
•
the scope of auditors' liability is already limited by the courts (see paragraph 3.6 above) and professional indemnity insurance is available to cover the residual risks. Potential new entrants are also able to limit their liability by adopting the form of a Limited Liability Partnership (LLP), thereby protecting the personal assets of partners not implicated.8 Some Group A firms are already LLPs.
4.11 It appears, therefore, that a cap on auditor liability would not be a catalyst for new entry while other more significant barriers to entry remain in place. Most responses to the Consultation are fairly pessimistic about the willingness and/or ability of Group A firms to compete actively with the Big Four for the audits of listed companies. There is broad agreement that unlimited liability is but one – relatively insignificant barrier among the many other barriers to entry facing smaller audit firms.
OTHER BARRIERS TO ENTRY ARE MORE SIGNIFICANT
4.12 Two significant barriers to entry are reputation and a global network. An established reputation, built through a history of repeated business without problems, is undoubtedly important to an audit client. However, what appears to be of even greater significance is the reputation of audit firms with third parties such as investors, creditors and rating agencies. A recent EC merger investigation9 found that it was the 'norm' for audit appointments to be renewed, with firms reluctant to change auditors because of fears that this may damage their reputation or stock market rating since the investment community may suspect disputes over financial reporting. Indeed, a wide range of respondents to the Consultation, including Group A firms, accountancy professional bodies, investors and corporate audit clients emphasise the role that third party perceptions of reputation play in the appointment of auditors. 4.13 Multi-national firms (and their shareholders) often want audit work done consistently around the world by an audit firm with global reach. Some Group A firms have set up their own international networks in an attempt
LLPs were introduced in 2000 and provide the benefit of protection of the personal assets of innocent partners, whilst maintaining the favourable tax treatment of partnerships. 9 M1016, 'Price Waterhouse / Coopers & Lybrand', 1998.
8
Office of Fair Trading
11
to overcome this barrier to entry, but coverage is not as extensive as that of the Big Four. 4.14 Additional barriers to entry relate to regulation and economies of scale and scope. The performance of statutory audits is a regulated activity restricted exclusively to accountants registered with certain (but not all) chartered accountancy bodies (including the Association of Chartered Certified Accountants (ACCA) and ICAEW). In addition to this demarcation, regulation creates various other barriers to entry. For example, the so called '50 per cent rule'10 limits statutory audit work to firms in the majority control of qualified auditors, a significant restriction on a multi-disciplinary partnership (MDP) entering the market (preventing, for instance, a large non-accountancy firm – perhaps with complementary skills such as a law firm - entering the market through the acquisition of small audit firms). Moreover, proposed new regulations (see paragraph 3.3 above) may act as unintended barriers to entry – such as the APB's proposed ethical guideline on economic independence, the 10 per cent limit on any one client's audit and non-audit fees as a proportion of the audit firm's total fees. 4.15 Economies of scale are another important barrier to entry – audit firms need to have available audit staff with in-depth expertise and a proven track record in the relevant sector, and in sufficient number to service the largest firms.11 Economies of scope may also impede new entry, as many large audit clients appear to prefer a firm that can offer a range of additional non-audit services. Figure 4 shows the extent to which total fees paid to the audit firm are for non-audit services.
Companies Act 1989 (implementing a requirement of the EU 8th Company Law Directive). 11 M1016, 'Price Waterhouse / Coopers & Lybrand', 1998.
10
Office of Fair Trading
12
Figure 4: Audit and other fees from FTSE-100 firms
800 706 700 659
600 524 £ million (2003 prices) 500 454 425 400
300 225 200
290
297 226 201 203 193 245 234
100
0 1997 1998 1999 2000 Total Audit fees Total Other fees 2001 2002 2003
Sources: Financial Director, January 2004 / January 2003; ONS Economic Trends.
INSURANCE AND LLP STATUS ALREADY LIMIT AUDITORS' LIABILITY
4.16 Two factors reduce the significance of liability as a barrier to entry: the availability of professional indemnity insurance and LLP status. LLP status is the chosen corporate form of many audit businesses – and exists to protect partners' personal assets. Professional indemnity insurance, despite recent cost increases, remains available and should become more affordable and extensive in the medium term. 4.17 Steep, and often sudden, increases in the cost of liability insurance cover (including professional indemnity insurance) have affected a number of sectors in recent years. The Big Four firms appear to have been subject to some of the highest cost increases, although a lack of specific data has made it difficult to corroborate this. Datamonitor report average premium rate increases of 35-40 per cent in 2002 and 25-30 per cent in 2003, but confirm that the Big Four, along with surveyors and solicitors, have
Office of Fair Trading
13
experienced higher and more variable increases due to poor claims records.12 4.18 Insurance firms responding to the Consultation point to a lack of capacity, with only a limited number of insurers (around ten) willing to underwrite the professional indemnity exposure of the Big Four. Capacity is scarce because of the recent claims history for international accounting firms, several years of falling premiums in real terms,13 underwriting losses in the past ten years combined with falls in equity markets (so insurers are unable to rely on their investment returns), and the specialist expertise required to assess the appropriate premiums for uncertain claims. Capacity loss has led to increased retentions and premiums, decreased levels of cover (by 50-75 per cent in the last few years according to the insurance firm Aon), restricted policy terms and an exit from some classes of risk. 4.19 The current situation is not, however, expected to last. A study of liability insurance by the OFT in August 200314 concluded that the forces responsible for restrictions on capacity and premium increases are likely to subside. The suggestion by the British Insurance Brokers' Association that most 'corrective action' should have been taken by autumn 2004, and the fact that 'safety-net' agreements15 between insurers and bodies like the ICAEW are under-utilised (a situation confirmed by Datamonitor in April 2004), led the OFT to conclude that capacity should begin to return to the market. Indeed, Aon point to some recovery in capacity at the beginning of 2004, a trend confirmed later in 2004 by Datamonitor. Insurers are also reporting more profitable underwriting results for professional indemnity business, and Datamonitor estimate that the overall market has returned to profitability. They predict average premium increases for professional indemnity insurance of around 10 per cent in 2004 and 2005 and then decreasing premiums from 2006 onwards. The OFT is due to look again at the liability insurance market in late 2004.
UK Commercial Liability Insurance 2004 (Datamonitor, April 2004). According to the OFT report on Liability insurance (OFT659b, August 2003), liability premiums fell in real terms by 14% between 1995 and 2001. 14 Liability insurance (OFT659b, August 2003). 15 Accountants have access to an assigned risk pool (ARP) where risks can be placed if no cover can be found on the open market (risks are shared proportionately across insurers).
12 13
Office of Fair Trading
14
4.20 This evidence suggests that audit firms, particularly the Big Four, have seen big increases in their insurance costs and restrictions on the cover available. All, however, appeared to have obtained adequate cover, albeit at higher cost. Looking ahead, some further, but more modest, cost increases are expected in the next 12-18 months but the situation is then forecast to improve as higher profitability attracts further new capacity. 4.21 In our view, given that the current liability position is common to all auditors, the significance of other barriers to entry and the availability of existing means of limiting liability, a cap on auditor liability would not stimulate new entry into the relevant market.
Unlimited liability and existing competition
4.22 The current liability position could induce (and it is alleged to have done so in the past) Big Four firms to withdraw audit services from risky companies or, in the extreme, from the whole audit market for listed firms.16 The same factors, it is argued, are causing problems with the recruitment and retention of audit partners and staff. (We consider this issue separately in paragraphs 5.5 to 5.7 below.) Partial or full withdrawal from the audit market would lead to reduced choice, dampened competition, and government intervention to ensure firms can comply with their statutory audit requirement. 4.23 We have not seen any evidence to support the linkage between unlimited liability and the existing level of competition between Big Four firms, or market concentration. The current liability position does not appear to be causing Big Four incumbents to withdraw from the market. Our analysis suggests two reasons for this:
Some respondents would consider a selective withdrawal from the riskiest firms a positive development, given the valuable signal that this sends to shareholders (particularly if the auditor files a meaningful statement on resigning). In the view of one bank, it would be perverse if limiting liability resulted in audit firms taking on high-risk companies. They suggest it might be better if the auditor withdraws, thereby encouraging internal reforms sufficient to attract back the services of an audit firm and comply with the statutory audit requirement.
16
Office of Fair Trading
15
• •
the value of audit, and the related business of supplying non-audit services to audit clients, and the scope for significant increases in audit fees and provisions for expected future claims to help offset the risks of carrying out audit work.
4.24 Reform of the current liability position does not appear to be justified by considerations of existing competition between Big Four firms.
AUDIT AND RELATED SERVICES ARE VALUABLE BUSINESS
4.25 Withdrawal from the provision of audit services to listed firms on a significant scale would seem unlikely given the value of the market – worth £234 million in 2003 for FTSE-100 firms alone (see Figure 4 above). Moreover, it is a market that is 'captive' as a result of the statutory audit requirement placed on larger firms (see paragraph 4.6 above) and high barriers to entry (see paragraphs 4.12 to 4.15 above). The EC also cite17 the insignificance of the audit fee relative to the turnover of the largest firms, and switching costs – as the result of a learning curve - as further factors behind the inelastic demand for audit services to large firms. 4.26 Turning to the indirect value of the audit market to the Big Four, audit clients are major purchasers of non-audit services from their auditors, worth a further £454 million in 2003 for FTSE-100 clients alone (see Figure 4 above). An indication of the current importance to the Big Four of non-audit revenue generated from audit clients is shown in Figure 5.
17
M1016, 'Price Waterhouse / Coopers & Lybrand', 1998.
Office of Fair Trading
16
Figure 5: Fee ratio for FTSE-100 firms (2002/03)
300% 284%
250% 218% 200%
Fee ratio
156% 150% 147%
100%
50%
0% Deloitte E&Y KPMG PwC
Source: Financial Director, January 2004. Fee ratio = total non-audit fee/audit fee.
4.27 Non-audit fees are between 156 per cent and 284 per cent of the audit fee for FTSE-100 firms (tax, 'other' audit and M&A being the most important sell-on services). Figure 4 showed the position over time – and the even greater importance of non-audit work in earlier years.18 4.28 The combined direct and indirect value of the audit market for listed firms suggests that Big Four auditors will not easily relinquish their established positions. Indeed, we are not aware of evidence pointing to the actual withdrawal by Big Four firms from the provision of audit services to listed firms. On the contrary, the very high concentration levels in the relevant market have been increasing steadily, and auditors are rarely changed (see Figures 6 and 7 respectively).
18
This is likely to be the result of disposals of management consultancy and outsource businesses owned by audit firms, and the impact of the Smith Report on the role of audit committees.
Office of Fair Trading
17
Figure 6: Big Five/Big Four shares of the audit and accountancy sector
100%
90%
80%
70% 59.8% 62.2% 64.6%
66.0%
67.2%
60%
50%
40%
30%
20%
10%
0% 1998 1999 2000 2001 2002
Source: KeyNote Accountancy Market Report, September 2003.
4.29 According to the 2002/03 accounts of listed firms, the Big Four now control 100 per cent of FTSE-100 audits and 99% of FTSE-350 audits (i.e. all but eight firms) - see Figures 2 and 3 above. An examination of FTSE-100 auditors in 1997 (Figure 7) reveals that only 12 per cent of the FTSE-100 firms still listed in that index in 2002 had changed auditor over the five year period (and FTSE-100 firms are likely to be the greatest liability risk). Of the 95 FTSE-100 companies in January 2004 that were also listed in that index in January 2003, only three had switched audit firm in that year.
Office of Fair Trading
18
Figure 7: Audit rotation – 1997 v 2002
Different auditor 12% Retained one auditor from joint audit 1% Same/legacy auditor as one premerger PLC 7%
Same/legacy auditor 80%
Source: Financial Director, January 2003. Base: 69 companies that were in the FTSE100 in 1997 and in 2002.
4.30 Many respondents to the Consultation take a sceptical view about the withdrawal claims, citing, in addition to the factors listed above, the speed with which Andersen UK's clients were acquired by the Big Four following its collapse. Several audit clients and Group A firms maintain that the Big Four appear to be competing strongly for any audit business when it arises, with no evidence of withdrawal from either a complete audit or from certain specific audit tasks. The Financial Services Authority (FSA) is not aware of any evidence to suggest that liability risk contributes to concentration and does not believe that firms will withdraw from audit work in the absence of reforms of liability.
ARE THE RISKS SUFFICIENT TO PROMPT WITHDRAWAL FROM AUDIT?
4.31 If the market were operating correctly, we would expect to see audit fees and provisions for expected claims increase in response to increased liability risk. We would expect this to occur in advance of any withdrawal from the provision of audit services. Neither appears to have occurred to any significant degree. This could be because the risks (or lack of
Office of Fair Trading
19
insurance cover) are exaggerated, and/or because the existing Big Four business model relies on low audit fees to generate a stream of more profitable non-audit business. 4.32 Figure 8 shows the real average audit fee for the FTSE-100. Even allowing for the downturn in the economic cycle, the fact that there was no real increase in audit fees between 2002 and 2003 does not appear to be consistent with the claim that liability risk has reached a level where Big Four firms are actively withdrawing audit services from certain risky clients. Figure 8: Average UK audit fees for FTSE-100 firms
3.0
2.5 2.24 2.03 2.0 £ million (2003 prices) 2.07 2.01 2.31
2.45
2.45
1.5
1.0
0.5
0.0 1997 1998 1999 2000 2001 2002 2003
Sources: Financial Director, January 2004; ONS Economic Trends.
4.33 It might also be expected that the average time taken to sign off a completed audit would increase in the face of higher risks (and pressures to improve audit quality), but Figure 9 shows that the steady decline in sign-off times in recent years has yet to be reversed.
Office of Fair Trading
20
Figure 9: Average sign-off times for FTSE-100 firms
80
70
67 64 64 61 60 59 59
60
50
Days
40
30
20
10
0 1997 1998 1999 2000 2001 2002 2003
Source: Financial Director, January 2004.
4.34 There are likely to be efficiencies in the joint provision of audit and nonaudit services and this appears to be the basis for the established business model for the Big Four (see Figure 4 above). This may help to explain the behaviour of audit fees relative to the risks of carrying out audits on listed firms. Big Four auditors face the discipline of competitive tenders (or the threat of one), and have to fight off low competing bids aimed at dislodging an incumbent who has already moved up the learning curve. The incumbent audit firm therefore keeps audit prices low, increasing their overall returns from the audit client with sales of nonaudit services. These joint returns are being squeezed, however, by higher insurance costs and declining revenues from the sale of non-audit services (Figure 4 shows the fall in non-audit services from a peak in 2001).19 In the absence of government action permitting a cap on liability, we would expect the market to re-balance risk and reward with audit fee increases. 4.35 Turning to the level of provisions put aside by the Big Four, a review of the 2003 audited Annual Reports for three of the Big Four (Deloitte's has
19
This is likely to be the result of disposals of management consultancy and outsource businesses owned by audit firms, and the impact of the Smith Report on the role of audit committees.
Office of Fair Trading
21
not yet been published) casts further doubt on the claim that liability risk is inhibiting competition for audit business and prompting withdrawal from the market.20 All three Reports provide details of apparently rigorous risk management procedures, in terms of client acceptance, renewal and monitoring. None of the firms disclose details of their insurance cover but all include statements to the effect that insurance cover is held and provisions have been made against the possibility of claim settlements. However, the balance sheets show that the scale of provision for future expected claims is fairly modest (see the Table below). Table: Big Four provisions and capital KPMG PwC 2003 Provisions Capital
21
E&Y 2002 £67m 2003 £20.3m 2002 £22.5m
2002 None
2003 £66m
None £272m
£270m £372m
£348m £269.8m £238.6m
Source: 2003 Annual Reports. Deloitte's Annual Report yet to be published.
4.36 KPMG appear to make no provision at all for future expected claims in 2001, 2002 and 2003. Both PwC and E&Y provisions fell in the year to 2003. No quantification is given of the expected scale of potential claims. All three, however, show increased capital reserves. 4.37 Overall, the evidence does not indicate to us that the existing liability position is causing Big Four firms to compete less keenly for audit business, still less a partial or full withdrawal from the provision of audit services.
See Professor Gwilliam's response to the Consultation: Auditor Liability – Are the Big Firms Crying Wolf? (University of Aberystwyth, February 2004). 21 Members' capital and other reserves for E&Y, PwC (consolidated) and KPMG (consolidated).
20
Office of Fair Trading
22
5
5.1
THE RISK OF COLLAPSE
It has been widely argued that there is a small but real risk of a massive claim for financial damages causing the collapse of one or more of the remaining Big Four firms, leading to even higher levels of concentration in the audit market and reduced competition between the remaining three large auditors. On the basis of the evidence we have seen, we do not consider that an auditors' liability cap would prevent such a collapse. As in the case of Andersen UK, reputation damage rather than financial damage is likely to be the primary cause of collapse for a Big Four firm implicated in a major audit scandal. Many respondents to the Consultation consider Big Four firms to be highly vulnerable to reputation damage regardless of their liability position. The FSA, for example, suspects that loss of client confidence presents a greater risk than liability in terms of collapse. Moreover, a global network (a vital asset required to compete for the audit business of the largest firms) under threat from a claim in the US, for example, would be unlikely to be protected by a cap on liability in the UK. The EC, after examining the proposed acquisition of Andersen UK by Deloitte,22 did not hold liability responsible for the firm's collapse, but rather loss of reputation (with the brand asset devalued), dissolution of the network and the loss of US SEC accreditation. We think it unlikely, therefore, that a cap on liability will offer much protection to a Big Four firm in the event of a successful damages claim, on a massive scale, for auditor negligence or incompetence.
5.2
5.3
5.4
The impact on recruitment and retention
5.5 It is claimed that the current liability position is prompting some existing audit partners and staff to leave the auditing profession, inhibiting the Big Four's ability to compete for audit business and conduct high quality audits. In our view, the value of audit and related non-audit business, the adoption of LLP status, and the availability of insurance (see paragraphs
22
M2810 'Deloitte & Touche / Andersen (UK)' 2002.
Office of Fair Trading
23
4.22 to 4.37 above) means that audit will continue to be seen as an attractive and rewarding career. Moreover, as discussed above (paragraphs 4.1 to 4.4) a liability cap will not provide much protection to an audit firm in the event of a massive claim, and so would be unlikely significantly to enhance the recruitment and retention of audit staff. Figure 10: ICAEW Members in Accountancy Practice
40,000
35,000
33,187
33,775
30,000
29,947
25,000
20,000
15,000
10,000
5,000
0 2002 2003 2004
Sources: ICAEW; KeyNote Accountancy Market Report, September 2003.
5.6
Data were not available to conclusively support or refute the conflicting anecdotal accounts on this issue. Figure 10 shows that the numbers of ICAEW accountants working for an accountancy firm fell by 11 per cent between 2003 and 2004. Only three years of data were available, however, and so it is not possible to gauge the significance of this fall alongside the context of possible earlier fluctuations. Other data from the ICAEW, however, shows a positive year-on-year growth in overall membership numbers, of nearly 2 per cent per annum since the mid1990s (see Figure 11). The ICAEW data in Figures 10 and 11 should be interpreted with care, however.23
23
The ICAEW is only one of the audit professional bodies, and so changes in the relative attractiveness of alternative bodies may account for at least part of the trends shown in both Figure 10 and 11. Moreover, it would be wrong to draw firm conclusions
Office of Fair Trading
24
Figure 11: Growth in ICAEW Membership
10% 9% 120,000 8% 7% Annual % change 6% 5% 4% 3% 2% 1% 0%
19 83 19 84 19 86 19 87 19 85 19 88 19 89 19 90 19 91 19 93 19 94 19 92 19 95 19 96 19 97 19 98 20 00 19 99 20 01 20 02 20 03
2.8% 2.9% 2.6% 2.3% 3.2% 2.7% 2.4% 2.1% 1.5% 1.5% 1.8% 1.5% 2.0% 1.6%
140,000
100,000
80,000
4.7%
60,000
3.5% 2.9% 3.0%
40,000
2.2% 1.9% 1.6%
20,000
0
% change
Total Members
Source: ICAEW.
5.7
Our view, based on the partial evidence available, is that unlimited liability is not currently inhibiting the recruitment and retention of audit staff in any significant way. As described earlier, we believe that the audit market will remain attractive to audit firms, their partners and employees. Insurance cover is available to protect against claims of all but the highest magnitude, and insurance cover should become more extensive and affordable in the medium term. Finally, LLP status provides protection for the personal assets of partners – many of whom would be likely to find alternative employment with other audit firms in the event of a collapse (as in the case of Andersen UK).
on the role of liability exposure without evidence on other relevant factors, such as recruitment of other professionals or accounting technicians, or automation in the audit process. Allowance would also have to be made for the positive attractions offered by employers (such as banks) competing for accountants.
Office of Fair Trading
Total Members
25
6
6.1
THE CAP MECHANISM
Arguments that allowing auditors to cap their liability would be procompetitive are not compelling. Moreover, inadequate safeguards and some forms of cap design could actually distort competition (or quality incentives). There are, however, cap designs that are broadly neutral in terms of competition.
The need for mandatory guidelines
6.2 In our view, management may not take adequate consideration of the interests of shareholders when negotiating a cap (perhaps focusing instead on the audit fee),24 and we doubt the ability of shareholders to directly influence the negotiations. Left to the market, therefore, incumbent firms may negotiate for liability caps below the level necessary to provide quality audits and adequate redress for shareholders (a view echoed widely amongst a range of respondents to the Consultation).25 The role of liability as a discipline on audit quality has already been discussed (paragraphs 3.2 to 3.5). Although the Big Four audit firms argue that reputation and regulation (including a whole new raft of regulations) are sufficient to maintain audit quality standards, many others see a clear link between the level of liability and audit quality.26 The European Commission, in a study of different national liability regimes in 2002,27 comes out against the capping of liability on the grounds that liability is a driver of audit quality. Many respondents to the Consultation (e.g. the FSA and many investors), worried about the impact of a cap on audit quality, argue in terms of 'trading' further limits on liability for further tangible improvements in audit quality.
6.3
The original reason for the curtailment of limits to liability agreed between auditors and managers prior to 1929. 25 One firm of solicitors claims that 'unacceptably low' caps on liability are already routinely negotiated by Big Four firms for non-audit work. 26 Certain of the accountancy professional bodies, under questioning from the Treasury Select Committee, admitted that the potentially catastrophic consequences of poor quality work act as a very powerful discipline on partners. Paragraph 58, p18 The Financial Regulation of Public Limited Companies (House of Commons Treasury Committee Sixth Report of Session 2001-02 Vol.1, 18 July 2002). 27 A study on systems of civil liability of statutory auditors in the context of a Single Market for audit services in the European Union (EC, January 2002).
24
Office of Fair Trading
26
6.4
In addition to impacts on audit quality, low-cap/low-price deals could facilitate market foreclosure tactics (aimed at potential competitors in both the audit and, particularly, the non-audit markets). The joint provision of audit and non-audit services appears to be the Big Four's established business model (see paragraph 4.34 above), and so a plausible strategy consistent with this business model might be to offer very low audit fees in return for agreement on a low cap and non-audit sell-on business. This strategy could serve to raise entry barriers to both the audit and non-audit markets (important sell-on services for audit clients include tax, non-statutory audit, merger & acquisitions advice and other consultancy). If audit firms are to be allowed to cap their liability, our view is that there should be enforceable mandatory guidelines to prevent caps being set too low – with detrimental implications for audit quality, shareholder redress and competition (particularly in the non-audit markets, as the audit market is already subject to significant barriers to entry). This is consistent with the views expressed by a majority of respondents to the Consultation who, as a result of concerns over the low level of cap if left to the market, recommend that the Government set the cap (rather than delegating oversight to the FRC, for example).28
6.5
Cap design
6.6 We have already discussed the need for mandatory and enforceable rules to guard against the risk of abuse and an outcome where the cap is set too low. But the form of any rule used to set a minimum liability cap may also have implications for competition. For example, if the minimum cap level is linked to the size of the audit firm, competition may be distorted in favour of the Big Four.
28
The FRC itself said that it was not comfortable with taking responsibility for consulting on guideline cap levels (FRC response to the Consultation).
Office of Fair Trading
27
6.7
The Consultation offered four options for setting an appropriate minimum liability cap:29 • • • • as a multiple of the audit fee as a multiple of total fees paid by the company to the auditor in a given year as a multiple of the turnover of the auditor, and at one or more arbitrary levels depending on the size of the audit firm (reviewed periodically).
6.8
A cap linked to the total revenue of the audit firm could lead to a distortion of competition in favour of the larger audit firms (with higher liability caps offering shareholders more scope for redress). A cap linked to the total fees (audit and non-audit) generated from a client could result in a bias towards those firms offering a range of audit and non-audit services (i.e. the larger audit firms)30 and would run counter to the existing and proposed regulatory measures to ensure auditor independence. Two or more arbitrary minimum caps – set according to the size of audit firm – are also likely to distort competition in favour of the Big Four (e.g. large audit clients could be deterred from switching from a Big Four audit firm subject to a high cap to a Group A firm subject to a lower cap).
6.9
6.10 A single arbitrary minimum cap, however, if set at a sufficiently high level should be competitively neutral. The aim of such a cap would be to provide 'backstop' protection only – for those firms facing catastrophic claims. The minimum cap would need to be set at a high level in order to maintain quality incentives and to avoid serving as a substitute for insurance cover and provisions (otherwise, by reducing their costs, this
The adoption of a system of proportionate liability for economic loss generally – an option ruled out in the Consultation – attracted wide support. Many expressed the view that this should be the long-term solution to the liability issue. 30 According to academic research based on 2002 fee data, the fee ratio (non-audit fees / audit fees) is highest for those firms served mainly by the Big Four: FTSE-100 firms (271 per cent), and FTSE-250 firms (174 per cent). Small and Fledgling firms, served by a wider range of auditors, had fee ratios of 129 per cent and 116 per cent respectively. Table 4, p22 And then there were four: A study of UK audit market concentration – causes, consequences and the scope for market adjustment, Beattie, Goodacre & Fearnley.
29
Office of Fair Trading
28
mechanism would favour the Big Four). With the cap set at a high level there would be no reason, based on the level of redress available to shareholders, for audit clients to favour large auditors any more than they do at present. 6.11 Linking the minimum cap with the audit fee the client is charged is 'size neutral' and so should not distort competition in favour of larger audit firms. If, however, a low audit fee reflects a poor quality audit this mechanism would deliver a low cap rather than the high cap required for adequate shareholder redress. This mechanism may also serve to reinforce the practice of low audit pricing (with potential implications for audit quality) and correspondingly greater pressures to sell-on non-audit work. 6.12 Two options are relatively neutral in terms of competition: the single arbitrary minimum cap set at a high level, and the minimum cap linked to the audit fee. The latter, however, may distort incentives to maintain or improve audit quality (and may run counter to independence initiatives).
Office of Fair Trading
29
7
7.1
CONCLUSION
It is likely that allowing audit caps would be competitively neutral overall. The evidence available to us does not support the claim that a cap on auditor liability would be pro-competitive in terms of stimulating new entry, promoting existing competition for audit business, preventing the collapse of a Big Four audit firm, or as an aid to the recruitment and retention of audit staff. Lack of safeguards, and some forms of cap design, could distort competition and incentives to maintain audit quality. If caps are allowed, it will be important to ensure that they do not have anti-competitive effects.
Office of Fair Trading
30 