Real and accrual-based earnings management in the pre- and post- engagement partner signature requirement periods in the United Kingdom

Abstract

In order to address the concern about whether the Public Company Accounting Oversight Board’s recent requirement to disclose an EP’s name in the United States (US) may cause unwanted consequences such as an increase in real earnings management (EM), I investigate the effect of requiring an engagement partner (EP) to sign audit reports on real and accrual-based earnings management in the United Kingdom, a comparable test environment. I measure accrual-based earnings management as the signed abnormal discretionary accruals and real earnings management as abnormally low levels of operating cash flows and discretionary expenses, as well as abnormally excessive production. I do not find a significant change in the usage of accrual-based and/or real earnings management for firm-years suspected of beating/meeting the zero, last-year, or analyst forecast consensus earnings threshold from the pre- to post-signature period. The results are robust to alternative proxies for EM and the control sample approach. My results mitigate the concern that managers may resort to more economically costly real earnings management following the mandatory disclosure of an EP’s name in the US.

Appendix A: Variable descriptions

MV::

Market value of equity

A::

Total assets

BE::

The book value of equity

CFO::

Cash flows from operations

Accruals::

Earnings—cash flows from operations

COGS::

Cost of goods sold

Production costs (PROD)::

COGS + Change in inventory

S::

Sales

ΔS::

Change in sales

Inventory turnover ratio::

COGS/[Beginning inventory + Ending inventory)/2]

Receivables turnover ratio::

S/[(Beginning gross receivables + Ending gross receivables)/2]

PPE::

Property, plant and equipment

ABCFO::

Abnormal cash flows, measured as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{CFO_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{S_{t} }}{{A_{t - 1} }}} \right) + \beta_{2} \left( {\frac{{\Delta S_{t} }}{{A_{t - 1} }}} \right) + \varepsilon_{t}$$

ABPROD::

Abnormal production costs, measured as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{PROD_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{S_{t} }}{{A_{t - 1} }}} \right) + \beta_{2} \left( {\frac{{\Delta S_{t} }}{{A_{t - 1} }}} \right) + \beta_{3} \left( {\frac{{\Delta S_{t - 1} }}{{A_{t - 1} }}} \right) + \varepsilon_{t}$$

ABDisEXP::

Abnormal discretionary expenses, measured as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{DisEXP_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{S_{t} }}{{A_{t - 1} }}} \right) + \varepsilon_{t}$$

ABDisXRD::

Adapted from Vorst (2016), I measure abnormal discretionary research and development expenses as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{RD_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{RD_{t - 1} }}{{A_{t - 1} }}} \right) + \beta_{2} \left( {TobinQ} \right) + \beta_{3} \left( {\frac{{Infunds_{t} }}{{A_{t - 1} }}} \right) + \beta_{4} MV + \varepsilon_{t}$$

ABDisXSGA::

Adapted from Vorst (2016), I measure abnormal discretionary selling, general, and administrative expenses as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{SGA_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{\Delta Sale_{t} }}{{A_{t - 1} }}} \right) + \beta_{2} \left( {\frac{{\Delta Sale_{t} }}{{A_{t - 1} }}} \right)*NEGSALE_{t} + \beta_{3} \left( {TobinQ} \right) + \beta_{4} \left( {\frac{{Infunds_{t} }}{{A_{t - 1} }}} \right) + \beta_{5} MV + \varepsilon_{t}$$

ABDisXRD_K::

Adapted from Kothari et al. (2016), I measure abnormal discretionary research and development expenses as deviations from the predicted values of the corresponding industry-year regression

$$\frac{{RD_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{{\Delta RD_{t, t - 1} }}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{RD_{t - 1} }}{{A_{t - 1} }}} \right) + \beta_{2} \left( {\frac{{S_{t} }}{{A_{t - 1} }}} \right) + \varepsilon_{t}$$

ABAQ::

Abnormal accruals, measured as deviations from the predicted values of the corresponding industry-year regression,

$$\frac{{Accruals_{t} }}{{A_{t - 1} }} = \alpha_{0} + \alpha_{1} \left( {\frac{1}{{A_{t - 1} }}} \right) + \beta_{1} \left( {\frac{{\Delta S_{t} }}{{A_{t - 1} }}} \right) + \beta_{3} \left( {\frac{{PPE_{t} }}{{A_{t - 1} }}} \right) + \varepsilon_{t}$$

SI::

An indicator variable that is equal to one if a firm-year has: (1) scaled earnings greater than or equal to zero but less than 0.005; (2) changes, ranging from zero to 0.005, in scaled earnings between the current period and prior period or (3) a small number of earning forecast errors per share, ranging from zero to 0.02, equal to zero otherwise

IB::

Income before extraordinary items scaled by lagged total assets (A), expressed as the deviation from the corresponding industry-year mean

Size::

Logarithm of MV, expressed as the deviation from the corresponding industry-year mean

MB::

The ratio of MV to the book value (BV), expressed as the deviation from the corresponding industry-year mean

POST::

An indicator variable equaling one if a firm-year ends in or after April 2009, and zero otherwise

SI_POST::

An interaction term between suspect firm-years (SI) and the EP signature effect (POST)