Purpose
Due to stakeholders’ concerns on the contribution of corporate governance in monitoring insurance companies during financial crisis, this study aims to investigate whether and how various corporate governance practices would have affected firm performance of listed and non-listed insurance firms in the UK during financial crisis.

Design/methodology/approach
This study uses a unique manually collected data set from listed and non-listed insurance firms in the UK and applies different regressions models to test the hypotheses and to address the endogeneity problem.

Findings
The findings show that board non-duality and the presence of a majority shareholder improve firm performance in insurance companies. Furthermore, the findings for the sub-samples indicate a stronger positive association between board of directors and firm performance in listed insurance companies after the financial crisis, while a positive impact has been found between large shareholders and external audit firms in non-listed insurance companies before and during the crisis.

Practical implications
The results offer important practical implications for the government, management, shareholders and policymakers. For example, regulators and policymakers should benefit from these results to revise the recommendations for corporate governance mechanisms that prove to be effective on firm performance, as well as those mechanisms that have different or unexpected effects among listed or non-listed firms and/or during the turbulent periods. Investors should be aware of those specific corporate governance mechanisms that would have higher effect on performance of UK insurance firms in which they are considering to invest in.

Originality/value
This study contributes to the current literature by exploring the effect of corporate governance on financial performance by comparing between listed and non-listed insurance companies during financial crisis. Further, to the best of the authors’ knowledge, this is the first study to use two new insurance-related performance measures, the revenue growth ratio and the adjusted combined ratio, as performance proxies to explore whether these new variables create any insights.