Competitive pressure, audit quality and industry

specializationWieteke Numan

(KU Leuven)

Marleen Willekens(KU Leuven)

University of Oklahoma Price College of Business

Steed School of Accounting Conference

Regulation and the Audit Industry

18 May 2012

o Audit market has high level of concentration, a given demand for audits and high entry barriers → Regulators are concerned

o Former SEC Chairman Christopher Cox on the 2005 AICPA National Conference:

“… within the accounting profession and within the SEC, we are forced to ask ourselves: “Is this intense concentration in the market for large public company auditing good for America?” If you believe, as I do, that genuine competition is essential to the proper function of any market the answer is no.”

Motivation

o U.S. Government Accountability Office, Audits of public companies, 2008:

“Firms with significant market power have to potential to reduce the quality of their products… however the presence of high market shares does not necessarily mean anticompetitive behavior is occurring…

… competition in an oligopoly can also be intense and result in a market with competitive prices, innovation and high-quality-products”

Motivation

Motivation

o Green paper, European Commission, October 13, 2010:

“ The market appears to be too concentrated in certain segments and deny clients sufficient choice when deciding on their auditors.

Moreover, being an auditor of large listed companies seems to create a reputational endorsement”

o UK - House of Lords Inquiry July 2010 / report Economics Affairs Committee March 2011

“….. explore concerns that market domination by a small number of firms damages competition and reduces choice in the audit market as well as raising questions about the quality of audited accounts…”

o What are the effects of competition on audit quality?

o Is supplier concentration a good measure when assessing competition in the audit market?

o Which economic theory fits the audit market and competition within that market?

Intended contribution

Key findings

o Concentration per se does not seem to affect audit quality

o The pressure from a closely competing audit firm does. Higher pressure is associated with lower quality

o The likelihood of issuing a going concern opinion and the accruals based quality of earnings are affected by industry expertise dominance over the closest competitor.

Competition surely exists even in oligopolies….

The Herfindahl measure of competitionMarket A (city-

industry)Auditor Market

share1 60%2 10%3 10%4 10%5 10%

Herf = 40%

Market B (city-industry)

Auditor Market share

1 30%2 25%3 20%4 15%5 10%

Herf = 22.5%Which market is most competitive? And why?

Competitive Pressure?

- (Absolute difference between incumbent office’s market share and the market share of the office that has the closest market share to that of the incumbent office) Market A (city-

industry)Auditor Mkt - Abs.

diff1 60% -50%2 10% 0%3 10% 0%4 10% 0%5 10% 0%

Market B (city-industry)

Auditor Mkt - Abs. diff

1 30% -5%2 25% -5%3 20% -5%4 15% -5%5 10% -5%

Herf = 0.40 Herf = 0.225

Problems with the Herfindahl• The Herfindahl is derived theoretically from a model of

Cournot competition– Cournot may not be a good description of the

matching process between audit firms and clients

• The conclusion that concentrated markets are less competitive does not hold when: – there are search costs (Stiglitz, JPE 1987)– competition is on price rather than quantity

• Assumes that the level of competition is the same for all firms within a given market

o Previous studies examine the relationship between audit quality and market concentration, but find mixed results

o Kallapur et al. (2010) and Newton et al. (2011) – positive association

o Francis et al. (2012) – international studyTwo facets of concentration:

oB4 concentration level: positive association

oWithin B4 herfindex: negative association

Concentration and quality

o Differentiation through industry specialization (market share)

o Lim and Tan (2008): no main effect Only conditional on high joint supply of NAS

o Reichelt and Wang (2010): positive main effect

Prior empirical studiesaudit quality and differentiation?

→ Effects of Industry specialization are studied, but not the effect of competitive pressure?

o Based on theoretical front end in Simunic (1980) = neo-classical economics:

o Perfect competition vs. monopolistic pricing

o No strategic elements included

o However, is this the appropriate model to study competition in the audit market?

o Oligopolistic competition o Product differentiation

(horizontal /vertical)

Prior empirical studies about competition in the audit market

Competition through differentiation

o Oligopolistic competition through differentiation

o The audit market characterized by a large number of heterogeneous clients that pay unique fees: heterogeneity of consumer tastes → horizontal differentiation

o Clients are willing to pay higher fees for audits performed by different type of auditors → horizontal and vertical product differentiation

- auditors strategically choose to become ex post heterogeneous through specialization

- audit firms modeled as Bertrand oligopolists who

simultaneously choose specializations w.r.t. client

characteristics

- empirically tested by merger Price Waterhouse and Coopers & Lybrand in Australia

Chan et al. (2004): location model

Competition and audit pricing

.

o Numan and Willekens (2012) show that on a city level, audit fees not only increase in the level of industry expertise, but also in industry expertise dominance over the closest competitor

→ First aspect of market location through differentiation = Industry specialization as measure of auditor fit with client

→ Second aspect of market location through differentiation = Distance/closeness of closest competitor

o Hence: competitive pressure (market share distance) from closest competitor has negative (positive) effect on audit fees

Competition and quality?

- Industry experts can deliver higher quality as specialists earn fee premiums that they can invest further in industry knowledge of the client. In an oligopolistic competition setting, the pressure from close (read: similar) competitors puts fee premiums under pressure, which could lead to lower effort/quality.

- In a competitive market, an auditor is more likely to lose its clients to competitors. The auditor has stronger incentives to please management in order to keep the engagement.

+ Competitive pressure may force the auditor to distinguish himself on other factors than price and results in innovation, for example to avoid client switching (GAO report 2008)

The extent to which pressure from close competitor affects audit quality is an open, empirical question

Design Model

Degryse and Ongena (2005) and Numan and Willekens (2012):

Audit quality = P (going concern)

Accruals-based earnings quality

Competitive pressure = Industry market share ‘closeness’ to closest competitor

Industry expertise = Fit between the auditor and client in terms of industry specialization

Controls = i.e. Alternative competition variables such as the

Herfindahl Index and Big4

ControlsexpertiseIndustry pressure competive qualityAudit 13

321

iio aaaa

Industry specialization measuresRiley and Neal (2004):

Market share approach: “How well has an audit firm differentiated itself from its competitors in terms of market share within a particular industry”

• Possible to designate specialists to industries that are too small

• Possible to inadequately recognize large, highly competitive industries where most major accounting firms generate significant revenues and devote significant resources to develop industry audit technologies

Portfolio share approach: “Relative distribution of audit services and related audit fees across the various industries for each audit firm considered individually”

• Could be driven by size of the industry

Sample and datao Audit market m: Audit firms serving a 2-digit SIC industry in a Metropolitan Statistical Area (MSA /City level; Francis et al. 2005)

o Competitive pressure: ‘Closeness’ of closest competitor based on industry market share

o Industry expertise: Both portfolio share measures and market share measures (Neal and Riley 2004)

Going concern analysis

o Industry specialists have the expertise and independence to more often issue a going concern opinion (Lim and Tan 2008; Li 2009)

o Sample is restricted to financially distressed firms (Lim and Tan 2008; Li 2009).

o Sample period 2005-2008

o Data sources: Compustat, Audit Analytics

o 4,134 firm-year observations; about 20% received a going concern opinion

Descriptives Going-concern sample  N Mean StdDev Min P25 Median P75 Max                 

Dependent variable                Going concern 4134 0.204 0.403 0.000 0.000 0.000 0.000 1.000                 

Independent variables                

Competitive pressure closest competitor 4134 -0.140 0.219 -0.955 -0.167 -0.034 -0.004 0.000

Industry expertise (office) 4134 0.318 0.341 0.004 0.056 0.170 0.467 1.000

Industry expertise (national) 4134 0.165 0.211 0.000 0.056 0.089 0.172 1.000

Industry leader (office) 4134 0.250 0.433 0.000 0.000 0.000 1.000 1.000

Industry leader (national) 4134 0.138 0.345 0.000 0.000 0.000 0.000 1.000

Industry specialist (office) 4134 0.183 0.387 0.000 0.000 0.000 0.000 1.000

Industry specialist (national) 4134 0.118 0.322 0.000 0.000 0.000 0.000 1.000Herfindex_msa 4134 0.278 0.074 0.164 0.240 0.261 0.291 0.993

Total assets in million $ 4134 384 1147 0.050 15 56 201 11550Size 4134 3.974 2.119 -2.996 2.689 4.033 5.301 9.354Current ratio 4134 3.274 3.723 0.007 1.124 2.035 3.978 29.396Leverage 4134 0.176 0.304 0.000 0.000 0.022 0.253 2.745

Leverage change 4134 0.023 0.189 -1.087 -0.005 0.000 0.034 1.333Roa 4134 -0.679 1.962 -31.676 -0.567 -0.208 -0.049 0.299Loss_lag1 4134 0.753 0.432 0.000 1.000 1.000 1.000 1.000Cfo_lag1 4134 -0.325 0.944 -11.383 -0.308 -0.063 0.039 0.420Sales_turn 4134 0.885 0.863 0.000 0.213 0.663 1.293 4.820MTB 4134 2.615 8.803 -55.072 0.718 1.758 3.842 80.351Big4 4134 0.483 0.500 0.000 0.000 0.000 1.000 1.000Ln_tenure 4134 1.669 0.792 0.000 1.099 1.792 2.197 3.401

Earnings quality analysiso High quality auditors are more likely to detect questionable accounting practices and misrepresentations

o Industry leadership / expertise is associated with higher earnings quality (i.e. Reichelt and Wang 2010)

o Our measure of earnings quality: performance-adjusted abnormal accruals based on a cross-sectional Jones (1991) model.

Earnings quality analysiso Sample period 2005-2008

o Data sources: Compustat, Audit Analytics

o 7,071 firm-year observations and 2,640 unique clients

Descriptives abnormal accrualssample  N Mean StdDev Min P25 Median P75 Max

                 

Dependent variable                Abs_abn_accruals 7071 0.111 0.119 0.000 0.037 0.079 0.144 1.253                 

Independent variables                Competitive pressure closest competitor 7071 -0.186 0.248 -0.970 -0.246 -0.070 -0.017 0.000Industry expertise (office) 7071 0.243 0.295 0.004 0.039 0.112 0.320 1.000

Industry expertise (national) 7071 0.119 0.172 0.003 0.033 0.074 0.113 1.000Industry leader (office) 7071 0.339 0.473 0.000 0.000 0.000 1.000 1.000

Industry leader (national) 7071 0.202 0.402 0.000 0.000 0.000 0.000 1.000Industry specialist (office) 7071 0.242 0.429 0.000 0.000 0.000 0.000 1.000Industry specialist (national) 7071 0.176 0.381 0.000 0.000 0.000 0.000 1.000

Herfindex_msa 7071 0.281 0.074 0.164 0.240 0.262 0.296 0.993Total assets in million $ 7071 1444 3702 0.625 53 218 950 30734Size 7071 5.423 2.048 -0.470 3.962 5.383 6.856 10.333

Std_cfo 7071 0.111 0.183 0.008 0.033 0.058 0.116 2.816Cfo 7071 -0.009 0.282 -2.927 -0.024 0.066 0.125 0.327

Leverage 7071 0.154 0.210 0.000 0.000 0.061 0.249 1.487Loss 7071 0.411 0.492 0.000 0.000 0.000 1.000 1.000MTB 7071 2.860 4.209 -24.011 1.258 2.143 3.661 37.904

Litigation 7071 0.338 0.473 0.000 0.000 0.000 1.000 1.000Current ratio 7071 3.014 2.617 0.082 1.401 2.166 3.645 19.901Roa 7071 -0.102 0.386 -5.283 -0.119 0.026 0.075 0.298

Sales_turn 7071 1.026 0.740 0.000 0.504 0.880 1.374 4.253Total_accruals_lag 7071 -0.081 0.184 -2.553 -0.094 -0.046 -0.012 0.266

Ln_tenure 7071 1.866 0.838 0.000 1.386 1.946 2.485 3.526Big4 7071 0.673 0.469 0.000 0.000 1.000 1.000 1.000

Dependent variable: P (going concern)

Expertise = Portfolio share

Parameter Estimate Prob Estimate Prob

     

Intercept 0.591 0.0805 0.668 0.0502

Competitive pressure -0.506 0.0440

Industry expertise (office) 0.499 0.0094 0.450 0.0205

Herfindex_msa -0.909 0.2286 -1.211 0.1169

Other control variables

Dependent variable: P (going concern)

Expertise = Leader

Parameter Estimate Prob Estimate Prob

     

Intercept 0.769 0.0198 0.816 0.0140

Competitive pressure -0.479 0.0781

Industry expertise (office) 0.274 0.0853 0.153 0.3802

Herfindex_msa -0.501 0.4880 -0.783 0.2925

Other control variables

Dependent variable: Abnormal accruals

Expertise = Portfolio share

Parameter Estimate Prob Estimate Prob

     

Intercept 0.135 0.0002 0.131 0.0003

Competitive pressure 0.022 0.0135

Industry expertise (office) 0.008 0.4316 0.012 0.2973

Herfindex_msa -0.023 0.2783 -0.012 0.66111

Other control variables

Dependent variable: Abnormal accruals

Expertise = Leader

Parameter Estimate Prob Estimate Prob

     

Intercept 0.143 0.0000 0.138 0.0000

Competitive pressure 0.028 0.0007

Industry expertise (office) 0.002 0.5404 0.009 0.0632

Herfindex_msa -0.014 0.5918 0.004 0.8873

Other control variables

Competition and audit quality

• Competitive pressure ∆↑ 10% - P(going concern) ↓ from 12.9% to 12.1% Earnings quality ↓ ∆ 2%

• Similar results Big 4 and non-Big 4 firms

• No effect of general concentration

Conclusions and interpretations

o The higher the pressure from a closely competing audit firm, the lower the delivered audit quality of the incumbent audit firm.

o The likelihood of issuing a going concern opinion or the accruals based quality of earnings are affected by industry expertise dominance over the closest competitor.

Some limitations

o Endogeneity

o Measure of ‘fit’dimensions of auditor differentiation

o Definition of market segments

o Vertical or horizontal differentiation?

Thank you for your comments and attention