Risk Reporting Incentives: A Cross-Country Study

Abstract

The current study aims to investigate empirically the main incentives for mandatory and voluntary risk reporting (MRR and VRR) across the USA, the UK and Germany, each of which has a unique approach towards risk reporting. While the UK approach encourages more voluntary risk reporting above imposing risk rules, the German approach formally requires firms to provide risk information in a certain place in their annual report narratives. The US approach is a compromise between these two approaches; it obligates and encourages firms to provide more information about their risks mandatorily and/or voluntarily, respectively. Investigating the incentives for risk reporting in such set of countries answers the calls of some prior research (e.g., Linsley and Shrives, 2006; Dobler, 2008; Dobler, Lajili and Zeghal, 2011) to deepen our understanding of what motivates firms to disclose their risks. To this end, computerised content analysis and multilevel analysis (MLA) on a large scale (compared with previous work e.g., Linsley and Shrives, 2005, 2006; Abraham and Cox, 2007) are utilised. The results are produced in four cumulative contexts through Chapters Six to Nine. These results are consistent with managers’ incentives theories (discussed in Chapter Two) and prior risk reporting literature (discussed in Chapter Three and Chapter Four).

Based on 15 firms in each country during 2007 and 2008, multivariate analysis of variance (MANOVA) results reveal significant differences between a firm’s risk levels and its risk disclosure levels across the USA, the UK and Germany. The correlation results indicate that these differences are statistically correlated, supporting the main argument of the current study that differences in a firm’s risk levels should be reflected in their risk reporting practices (Chapter Six).

Based on 1160 firm-years of non-financial firms of the FTSE all share index over 2005-2008, linear mixed model (LMM) results document that firms with higher levels of systematic and financing risks are likely to exhibit significantly higher levels of aggregated and voluntary risk reporting, whereas firms with high variability of stock returns or lower levels of liquidity are likely to exhibit significantly lower levels of aggregated and voluntary risk reporting. The current study also finds, however, that MRR is associated significantly and positively with firm size rather than with risk levels. The results also indicate that managers of firms exhibiting greater compliance with UK risk reporting regulations have greater incentives to disclose voluntary risk information (Chapter Seven).

When the study extends the scope to the other two countries, different patterns of relations are found. Based on 1270, 1410 and 1005 firm-year observations over 2005 to 2009 in the USA, the UK and Germany, respectively, repeated measures multilevel analysis (RMMLA) results suggest that, in the USA, MRR is more sensitive to firm risk levels (total, systematic and liquidity risks) than is VRR, which is more correlated to other firm characteristics. The UK results suggest that VRR is more sensitive to firm risk levels (systematic and liquidity risks) than is MRR, which is dominated by firm size, among other firm characteristics. In Germany, however, both MRR and VRR are significantly related to risk levels (total, systematic, un-systematic, financing and liquidity risks) (Chapter Eight).

Based on 3685 firm-year observations during the period between 2005 and 2009, and concerning both firm- and country-level analyses, repeated measures multilevel analysis (RMMLA) results support that variations in MRR can be attributed to differences in the legal systems (country characteristics) and in firm size (firm characteristics). The variations in VRR are more associated with firm characteristics, especially a firm’s risk levels across the USA, the UK and Germany (Chapter Nine).

These results have many implications and support the respective regulatory approach adopted within each country by interpreting the extent to which either MRR or VRR is more or less sensitive to underlying risks.