ABSTRACT Managing sector risk within factor portfolios has traditionally been viewed as a binary decision problem: to neutralise sector risk or not. Challenging this view, we introduce a novel conceptual framework that allows a continuum of choices, thus reframing the problem as one of optimising sector risk. Our approach begins by decomposing a factor portfolio into sector-neutral and sector-specific components, then recombining them using a mean-variance framework to create a sector-optimised factor portfolio. By dynamically optimising over time, the framework can exploit time- and state-variation in the component portfolios, resulting in substantial performance improvements over both standard and sector-neutral factor portfolios. Unlike the traditional formulation, where the merits of sector-neutralisation have been shown to be factor-specific, empirical analysis indicates that the benefits of our optimisation framework are not confined to any specific factors or methodologies.