D’Amore-McKim School of Business Separating Auditor Provided Tax Compliance and Tax Planning Services: Audit Quality Implications James Chyz University of Tennessee Ronen Gal-Or Northeastern University Vic Naiker Monash University 2016 Deloitte/University of Kansas Symposium May 21, 2016 1
D’Amore-McKim School of Business Research Questions RQ1: Does knowledge spillover between tax and audit functions depend on the nature of tax service provided (i.e. planning or compliance) to audit clients? RQ2: Does knowledge spillover across tax and audit service lines arise between clients of the same office and does it depend on the nature of tax service provided? RQ3: Do office-level tax compliance and tax planning NAS capture audit offices’ allocation of resources and attention to the audit practice in ways that benefit or harm audit quality? 2
D’Amore-McKim School of Business Findings RQ1: Does knowledge spillover between tax and audit functions depend on the nature of tax service provided (i.e. planning or compliance) to audit clients? • Yes… Audit quality enhancing knowledge spillovers at the engagement level derive from the provision of tax compliance, but NOT tax planning non-audit services. RQ2: Does knowledge spillover across tax and audit service lines arise between clients of the same office and does it depend on the nature of tax service provided? • Yes… High levels of office level tax compliance (but not planning) NAS is associated with improved audit quality. RQ3: Do office-level tax compliance and tax planning NAS capture audit offices’ allocation of resources and attention to the audit practice in ways that benefit or harm audit quality? • High levels of office level tax planning NAS is associated with reduced audit quality. 3
D’Amore-McKim School of Business The Tax NAS Debate Continued regulatory/ congressional/ proxy advisor concerns surrounding tax NAS • PCAOB Rule 3552 - prohibits audit firms from providing tax services that are related to the marketing, p la nning or opining in favour of transactions that would be considered aggressive tax position transactions • Institutional Shareholder Services (ISS) - propose “withholding” auditor ratification and AC member vote if other non-audit fees (including fees paid for tax p la nning NAS) are excessive. • Congressional (SEC) investigations regarding the tax p la nning services provided by PWC (EY) to Caterpillar (HP). Practitioners contend that efficiencies and knowledge spillovers arise when both services are jointly provided. • “We believe that a strong and leading-edge consulting practice makes us smarter — not only smarter as consultants, but also more knowledgeable as an overall firm regarding risks and challenges that affect our clients —audit and non-audit alike.” – PWC 2014 Audit Quality Report 4
D’Amore-McKim School of Business Engagement Level Tax NAS AUDIT TAX TAX DIVISION OF CLIENT NAS AUDIT FIRM A Kinney, Palmrose, and Scholz 2004; Paterson and Valencia 2011; Seetharaman, Sun, and Wang 2011; Gleason and Mills 2011; Krishnan, Visvanathan, and Yu 2013; Choudhary, Koester, and Pawlewicz 2014
D’Amore-McKim School of Business Engagement Level Tax NAS T.P AUDIT TAX DIVISION OF NAS CLIENT AUDIT FIRM T.C A NAS Paterson and Valencia (2011) Recurring tax NAS (Tax Compliance NAS ) Lower likelihood of restatement Non-recurring tax NAS (Tax Planning NAS ) Higher likelihood of restatement
D’Amore-McKim School of Business Hypothesis 1 Engagement Level Tax compliance NAS More likely to yield knowledge spillovers 1. Provided by local audit office 2. Promotes continuous contact with client and audit team 3. Knowledge transfers are more likely to be unbiased, relevant, and high quality (cf. when knowledge is sourced from a third party) 4. Can better evaluate tax provisions, tax reserves, etc. but also… 5. Helps auditors to better understand corporate structure, state and federal NOLs, foreign divisions/ subsidiaries, foreign tax credits, intercompany relationships, and information about specific accounts (i.e. capital expenditures and intangible assets) H1: Engagement level tax compliance fees are positively associated with engagement level audit quality. 7
D’Amore-McKim School of Business Hypothesis 2 Engagement Level Tax planning NAS Could also lead to knowledge spillovers Higher margin work could increase the cost of an audit failure (Gleason and Mills 2011) Involved with less frequent interactions with clients Deal with specific projects Often outsourced from larger offices Providers less likely to have accounting backgrounds (e.g. tax lawyers, transfer pricing economists) Economic bonding concerns (if tax planning contracts more lucrative) H2: Engagement level tax planning fees are not associated with engagement level audit quality. 8
D’Amore-McKim School of Business Office Level Tax NAS Client B T.P T.C Client C T.P T.C T.P AUDIT NAS TAX DIVISION OF CLIENT AUDIT FIRM T.P A T.C NAS T.C T.P T.C Client E
D’Amore-McKim School of Business Office Level Tax NAS Client B T.P T.C Client C T.P T.C T.P AUDIT NAS TAX DIVISION OF CLIENT AUDIT FIRM T.P A T.C NAS T.C T.P T.C Client E
D’Amore-McKim School of Business Office Level Tax NAS Client B T.P T.C Client C T.P T.C T.P AUDIT NAS TAX DIVISION OF CLIENT AUDIT FIRM T.P A T.C NAS T.C T.P T.C Client E Knowledge Transfer
D’Amore-McKim School of Business Office Level Tax NAS Client B T.P T.C Client C T.P T.C T.P AUDIT NAS TAX DIVISION OF CLIENT AUDIT FIRM T.P A T.C NAS T.C T.P T.C Client E Change in Strategic Focus
D’Amore-McKim School of Business Office Level Tax NAS Client B Client T.C T.P C T.P T.C T.P AUDIT NAS TAX DIVISION OF CLIENT AUDIT FIRM T.P A T.C NAS T.C T.P T.C Client E Audit Firm/ Office Culture Changes
D’Amore-McKim School of Business Hypothesis 3 & 4 Office Knowledge sharing/ spillover (O’Keefe et al. 1994, McGuire et al. 2012, Bianchi et al. 2014) Shifts in strategic focus/ culture Higher margin work (e.g., tax planning work) could lead to distractions from the audit and the audit firms “core values” – Former SEC Chief accountant (Paul Beswick), PCAOB Member (Steven Harris) Beardsley et al. 2015 –Increased focus on NAS associated with an increased likelihood of misstatements (in the presence of fee pressure). H3: Audit office level tax compliance NAS fees are positively associated with audit quality at the engagement level. H4: Audit office level tax planning NAS fees are not associated with audit quality at the 14 engagement level.
D’Amore-McKim School of Business Sample Selection Table 1 - Sample Selection Non-Misstatement a Misstatement Total Initial firm-year observations from Audit Analytics and Compustat 2,058 24,332 26,390 Firm-year observations with missing variables (410) (6,642) (7,052) MSA requirement (49) (482) (531) 1,599 17,208 18,807 Base Sample (2007-2011) Firm-year observations with positive Tax NAS 3) Firm-year 1) No Disclosure of Tax 2) Disclose Tax Fee observations with no Fee Breakdown Breakdown Tax NAS 9,260 2,892 6,655 Final Sample 9,547 a - We require that non-misstatements firms do not restate their 2006-2012 financial statements 15
D’Amore-McKim School of Business Research Design RESTATE = β 0 + β 1 ENGAGE_TAXCOMP_RATIO + β 2 OFFICE_TAXCOMP_RATIO + β 3 ENGAGE_TAXPLAN_RATIO + β 4 OFFICE_TAXPLAN_RATIO + β N Controls + Industry F.E. + Year F.E. + ε H1 β 1 (-) H2 β 2 (?) H3 β 3 (-) H4 β 4 (?) 16
D’Amore-McKim School of Business Table 3 – Primary Regressions Hypothesized (1) (2) (3) (4) DV = RESTATE Directional Expectation A One Stddev increase in -0.789 * -1.182 * ENG_TAXFEES - ENG_TAXCOMP is (-1.95) (-1.71) associated with an -1.911 ** -1.927 ** ENG_TAXCOMP - 8.73% reduction in (-1.78) (-1.80) misstatement likelihood ENG_TAXPLAN ? -0.089 -0.151 (-0.07) (-0.11) -0.115 *** OFF_TAXCOMP - (-3.75) 0.108 ** OFF_TAXPLAN ? A One Stddev increase in (2.56) OFF_TAXCOMP 0.174 ** ENG_LOG_AUDFEE + 0.052 0.053 0.059 ( OFF_TAXPLAN ) is (2.39) (0.56) (0.57) (0.62) associated with an 1.076 *** 1.327 *** 1.338 *** 1.372 *** ENG_OTH_NAF + (3.41) (3.30) (3.33) (3.41) 16.78% reduction of OFF_LOG_AUDFEE + -0.017 (11.55% increase in) (-0.41) misstatement likelihood. OFF_OTH_NAF + -0.010 (-1.03) Other Controls Yes Yes Yes Yes Industry and Year FE Yes Yes Yes Yes -3.776 *** -2.396 ** -2.406 ** -2.144 * Constant (-4.04) (-2.09) (-2.09) (-1.78) 17 Observations 18807 9547 9547 9547 Pseudo R 2 0.023 0.023 0.023 0.028
D’Amore-McKim School of Business Other Findings 1. The office-level results persist even for firms purchasing no tax NAS. 2. Results persist after controlling for the office-level propensity to disclose the tax fee breakdown. 3. Results persist after controlling for the decision to purchase tax NAS (Heckman two-stage analysis). 4. Changes in office-level tax compliance NAS immediately manifest in a lower likelihood of misstatement; Changes in office-level tax planning NAS gradually increase misstatement likelihood. 5. High levels of national-level tax planning NAS positively associated with misstatement likelihood. 6. The office level results are more prominent at the larger offices. 7. High office-level tax compliance is more strongly associated with errors rather than irregularities. 18
Recommend
More recommend
