Auditor tenure, non-audit services and earnings conservatism: Evidence from Malaysia

This paper examines the effect of auditor tenure and provision of non-audit services on earnings conservatism. Following Basu (1997), we measure earnings conservatism by using the asymmetric timeliness of earnings model and the time-series test of timeliness in loss recognition model. Using a sample of 2,235 company-year observations from the period 2005-2008, we present evidence that auditor tenure and non-audit services are positively associated with earnings conservatism. These results are robust even after controlling for company size, leverage and growth. We further examine the influence of government-linked companies on the relationship between (1) auditor tenure, (2) provision of non-audit services, and earnings conservatism. We find that our main results hold for non-government linked companies. For government linked companies, we find no association between auditor tenure and earnings conservatism, while the provision of non-audit services is negatively associated with earnings conservatism. This result implies that high provision of non-audit services to government-linked companies lead to impairment of auditor independence. Our additional analysis also shows that auditor tenure and provision of non-audit services are associated with higher earnings conservatism, regardless of whether the companies are audited by industry specialist or non-industry specialist.