How dark is the dark side of diversification?

Purpose - This paper aims to illustrate, within the context of a well-known linear diversification model, that risk management as exerted by banks and regulators ultimately depends on how risk is assessed and conceptualized. The two risk metrics used are the probability of bank failure and value at risk (VaR). The paper also extends the results of the model by incorporating an explicit analysis of correlation of the bank's portfolios. Design/methodology/approach - The paper is based on a well-known model of linear diversification of two banking institutions developed by Wagner (2010) in the Journal of Financial Intermediation. The authors added considerations that were unexplored by Wagner and derived the corresponding logical and practical implications. Findings - The authors found that depending on which of the two risk metrics being used, the way diversification is perceived and risk is managed may differ. This situation may very well end-up generating different incentives for banks and regulators. The authors suggest a general rationale for considering how to think about the apparent dilemma and the challenges faced by regulators. The authors also offer an explicit analysis of correlation for the bank's portfolios. Research limitations/implications - The results are dependent on the particular aspects of the model, so the research results may lack generality in other contexts. Practical implications - Despite the limitations already mentioned, the paper illustrates some relevant points within the open debate about risk measurement and diversification. Originality/value - This paper contributes to the open discussion of diversification, risk perception and systemic crisis.