Dr Sylvia Gottschalk and Dr Edward Bace explore how gender diversity in the banking sector can relate to a reduction of risk taking.
The 2008 financial crisis has been recorded as the biggest disruption of the global financial sector since the Great Depression of 1929. Although it originated in the United States and quickly spread to the banking systems of many large European countries, Canada was noticeably unaffected by this crisis.
Following the failure of Lehman Brothers, the International Monetary Fund (IMF) Head, Christine Lagarde, quipped that, “Had it been Lehman Sisters, the world would be different”. She was alluding to IMF’s empirical research showing that a higher proportion of women on boards of banks and financial supervision agencies is associated with greater stability.
A well-documented literature on gender differences with regard to risk-taking has shown that, on average, women tend to be more risk-averse than men. Based on this evidence, it has been suggested that a greater representation of women on executive boards may improve the risk-taking behaviour of male bankers.
How valid is this so-called “Lehman Sisters hypothesis”?
We investigate the relationship between risk-taking in the financial services industry and the female-to-male ratio in financial institutions in Canada and the US between 2008 and 2019.
Our initial research looked at the simplest form of risk-taking measure, leverage, which is defined as the ratio of total debt to total assets.
Figure 1 and Figure 2 show leverage in Canada and the U.S. between 2008 and 2019, against the female to male ratio on executive boards of financial institutions (FIs), for highly indebted financial institutions (leverage ratios between 100% and 300%).
Figure 1: Leverage and gender diversity on Canadian executive boards 2008-2019
Figure 1 clearly shows that, between 2008 and 2014, most Canadian FIs that have low gender diversity also have leverage ratios above 300%. Low gender diversity covers female to male ratios between 0 to 30%. Figure 1 also shows that most financial institutions have low leverage (0 to 50%) and no gender diversity.
This suggests that a lack of gender diversity does not necessarily imply high leverage. From 2014 on-wards, we see an increase in the number of women on executive boards compared to previous years. On Figure 1, this translates as a much wider range of female to male ratios, and we now see executive boards where half of the members are female. Over the whole decade following the 2008 Financial Crisis, it is clear that very few financial institutions have leverage above 300%. Most companies have debt levels below 100%.
Our findings also clearly indicate that having more female directors does not necessarily lead to lower leverage. In 2019, some Canadian financial institutions with more than 30% females on board also have a leverage ratio above 300%.
Figure 2: Leverage and gender diversity on US executive boards 2008-2019
In Figure 2 we see that many more US financial institutions have very high leverage (above 300%) than their Canadian counterparts.
In the year of the financial crisis, most companies that have very high leverage also have no gender diversification on executive board. However, there is a wide range of leverage levels among the financial institutions with no female director, and a significant number have leverage ratios below 100%.
Figure 2 shows a similar development as in Canada as far as female to male ratio is concerned. We see a steady increase in the number of women over the decade between 2008 and 2019. In 2019, the representation of women on the executive boards of many US financial institutions reaches 50%. As in the Canadian case, the correlation between high female representation and low leverage is not immediate. Some companies with leverage above 300% also have female to male ratios above 30%, and in one case, even 50%.
What do the findings show?
Our research shows a clear improvement of female representation in financial decision-making in Canadian and US financial institutions in the decade following the 2008 financial crisis. Nonetheless, the evidence that risk-taking has decreased as a result is less clear.
We can account for this result by noting that the “Lehman Sisters hypothesis” relies crucially on the assumption that women who join banks will conform to certain female stereotypes. Although women are –arguably-on average more risk-averse than men, women choosing a career in the financial services industry may not be so. Gender stereotypes apply to the average individual, but not to all, and there may be a wide range of risk-taking personalities among women. This suggests that having a woman on the executive board need not lead to more risk-averse decision-making.
Our results can also be explained by our choice of risk measure. The ratio of debt to total assets may not be the best indicator of risk-taking behaviour. We intend to follow this up in future research by looking at a wider range of risk-taking measures.