The Main Determinants of External Audit Fees: Theory and Empirical Evidence (2014-2016) from (sectoral sets of) FTSE 250 companies

Research Background -The fundamental focus of this thesis is the External Audit service/function (hereafter, audit, or auditing). In so doing, it examines the main determinants of Audit Fees (AF)within the identified sectoral sets of the FTSE250 and explores relevant theories that underpin AF modelling and those that may help to explain the movements in AF under an era of Mandatory Audit Rotation (MAR). Audit and the trust that it engenders is critical to sound Corporate Governance (CG). Hence the thesis is within the domain of CG. Because knowledge is packaged in theories, the theories that could be relevant to AF modelling were first explored in Objective (OBJ)1 of this thesis, which is partly motivated by the absence of an accepted set of theories that help to explain how AF are determined. While statistical confirming literature, regarding the influence of Non-Corporate Governance Variables (NCGVs) and Corporate Governance Variables (CGVs) on AF in several studies exist, equivalent studies were not identified, which specifically focused on their determination within the FTSE 250 index. Additionally, the thesis considers whether the directions of influence and approximate levels of significance of selected AF determinants within the cases audited by the Big4 in this research exhibit a shared commonality across the Big4 in this thesis. This is a response to the common practice of treating the Big4 audit firms as one homogenous unit. Research Setting - The research focuses on theory and empirical testing. In terms of theory, it is mainly facilitated by Agency Theory (AT) and Stakeholder Theory (ST). However, some theories which were not previously associated with AF also appear to align with some qualitative/quantitative explanations for the behaviour of AF, including in relation to MAR and empirical testing in connection with company size (i.e., Dynamic Capability Theory (DCT) and Knowledge Based Theory (KBT)). Empirical testing is based on 83 FTSE 250 companies and their four sectors for three years (2014-2016). The timeline relates to the years 2014 and 2016, in which the Audit Regulation & Directive (2014) (ARD, 2014) came into force and later took effect; respectively. The index is incredibly significant since it serves as a barometer for measuring the performance of the UK listed market. Research Purpose - The purpose of this research is principally to identify the main determinants of AF in the identified cases and their business sectors within the FTSE 250 companies and to identify /explain associated AF modelling theories due to perceived absence of literature that focuses directly on the latter aspects of AF modelling. It also investigates the appropriateness of treating the Big4 auditors as a homogenous unit, as well as the theories that could explain the likely influence of ARD regulation (MAR) on AF. Research Design, Methodology & Approach - The methodological considerations commence by linking empirically focused set of Objectives (OBJs) 2, 3 & 4 with positivist philosophy and associated ontological, epistemological, and axiological issues. To these objectives, it assigns a deductive approach (mainly based on AT), a research strategy that is fundamentally archival, with the choice of research method as essentially mono-methodical and quantitative. Time (frames) horizons are both cross-sectional and longitudinal. OBJs 1 and 5 are theoretical, qualitative, and inductive in approach and hence interpretivist in philosophy, with associated ontological, epistemological, and axiological issues. Research data were culled from various sources including the annual reports of the identified UK FTSE250 companies themselves, information from different electronic databases and data registered at the Companies House and London Stock Exchange (LSE) in UK, for triangulation. Empirical analyses were enabled by multiple regression equations using Ordinary Least Squares regression (OLS) and implemented using IBM SPSS. Innovatively, the research also considers lagged data in empirical testing and develops unique proxies for AUQ and ACC. It also adopts sector-wise and auditor-wise regressions for relevant companies as well as for the overall identified 83 FTSE 250 companies using unlagged and lagged information. Research Findings - Because this research falls within the domain of CG it maintains a fair balance of qualitative (theoretical) objectives (OBJs1 & 5) and quantitative (empirical) objectives ( 2,3 & 4), and hence the nature of its findings/results. In terms of the qualitative (theoretical) aspects, the main ii findings suggest that AF modelling is principally underpinned by AT and ST (OBJ1), including some other theories (such as DCT, Signalling Theory that help explain the behaviour of AF during MAR (OBJ5). Empirically, results show that all models are highly significant (at 0.1%). OBJ2 indicates that the (primary) NCGV determinants of AF in the identified FTSE250 companies, using non-sector-wise regressions, are size, complexity and business sector factors. In terms of sector-wise regressions, complexity is the most influential determinant with some other factors such as size and auditor factors becoming significant determinants although not consistent across the sectors. The latter outcomes are essentially consistent with some relevant aspects of the seminal AF modelling study(Simunic,1980) and most subsequent studies. OBJ3 compares coefficients of NCGVs in OBJ2 with OBJ3 after addition of CGV (Audit Committee Competence (ACC) and Audit Quality (AUQ) to the regressions. It finds that only the predictive powers of Size, Risk, Complexity and one dummy each of the Location and Auditor variables are significantly enhanced. The indication is that the CGVs are not very impactful on the determinants of AF in the identified cases within the FTSE 250. It may suggest that accounting systems and internal controls in FTSE250 appear to be effective hence any extra investment (in the form of AF premia or) in audit committees beyond the minimum legal requirements (e.g., minimum number of Non-Executive Directors (NEDS)) including financial experience, may not generate commensurate returns in terms of certain variables. Finally, findings regarding OBJ4 do not support the regular practice of considering the Big4 as a homogenous unit in AF modelling based on clients’ characteristics. Research Limitations - Despite all models being highly significant, the research data is exclusively secondary. The dummy variables may not have sufficiently captured the effect of the factors (e.g., business sector) which they represent. The theories that mainly facilitate (AT and ST) the research, suffer from several limitations indicated with the thesis. Research Contribution to Knowledge – This thesis extends existing knowledge and contributes original contribution to AF modelling. These include narrowing of the gap arising from the absence of a generally accepted set of theories that help to explain AF determination and some of which help to explain the behaviour of AF under a MAR era.It reaffirms AT and ST as the main underpinning theories of AF modelling. Secondly, it highlights linkages between some additional theories and AF (e.g., DCT and KBT) which were not indicated in prior studies. Thirdly, it extends empirical knowledge to the primarily influence of NCGVs on AF in the most populated sectoral sets of the FTSE 250 (2014-2016). Fourthly, the lagged regression models contribute original knowledge by enabling statistical representation of real-world considerations relating to the importance of prior year variables on the current year’s AF. Fifthly, the thesis develops original proxies for ACC and AUQ. The indication that there is absence of shared commonality between levels of significance and directions of influence of some AF determinants across the Big4 in this research is unique. Consistent with a prior UK study on Big4 AF premia, the thesis also finds that such AF premia are not always justified by commensurate levels of audit quality. Detailed findings provided within the thesis offer some theoretical and empirical knowledge to stakeholders (e.g., Boards, auditors, audit committees, policy makers or regulators) generally, and within the identified cases within the FTSE250. Originality/Value – To the researchers knowledge, this thesis is the first direct quantitate research into AF in FTSE250 or its four most populated sectors. It is the first in UK, to consider the issue of shared commonality (or not) of the impact of a set of primary NCGVs and CGVs across the Big4 auditors. It develops a diagrammatic conceptual framework linking several aspects of the research. Other aspects of originality include linkage of additional theories to AF modeling, being the first to identify theories which help to explain the behaviour of AF under MAR era, developing new proxies of ACC and AUQ and using lagged data in regression models.