Dynamic capitalist economies are characterised by relentless innovation and novelty and hence exhibit an indeterminacy that cannot be reduced to measurable risk. How then do economic actors form expectations and decide how to act despite this uncertainty? This pa per focuses on the role played by imaginaries, narratives, and calculative technologies, and argues that the market impact of shared calculation devices, social narratives, and contingent imaginaries underlines the rationale for a new form of 'narrative economics' and a theory of fictional (rather than rational) expectations. When expectations cannot be anchored in objective probability functions, the future belongs to those with the market, political, or rhetorical power to make their models or stories count. The paper also explores the dangers of analytical monocultures and the discourse of best practice in conditions of uncertainty, and considers the link between uncertainty and some aspects of populism.