Conviction, diversification or something else: constructing optimal portfolios with additional attributes

Purpose The purpose of this paper is to provide theory for some popular models and strategies used by practitioners in constructing optimal portfolios. King (2007), for example, advocated adding a diversification term to mean-variance problems to create better portfolios and provided clear empirical evidence that this is beneficial. Design/methodology/approach The authors provide an analytical framework to help us understand different portfolio construction practices that may incorporate diversification and conviction strategies; this allows us to connect our analysis to ideas in psychophysics and behavioural finance. The critical psychological ideas are cognitive dissonance and entropy; the economics are based on expected utility theory. The empirical section uses the theory outlined and provides the basis for constructing such portfolios. Findings The model presented allows the incorporation of different strategies within a mean-variance framework, ranging from diversification and conviction strategies to more ESG-oriented ones. The empirical analysis provides a practical application. Originality/value To the best of the authors’ knowledge, this model is the first to bridge the gap between portfolio optimisation and the psychological ideas mentioned in a coherent analytical framework.