Essays on reference dependence in macroeconomics : theory and evidence

There are indications that psychological factors and biases are gaining the centre stage of economics and finance once again, as they did during the thirties with the spreading of Keynesian ideas. Behavioural economics and finance is a booming field, if judged by the number of contributions in recent years and by its success in explaining phenomena which cannot easily find and explanation according to conventional models. Against this background, the main objective of this dissertation is to assess whether the psychological factors modelled in the behavioural finance literature, and in particular by prospect theory (Kahnerman and Tversky, 1979), can have a bearing on macroeconomic outcomes, be it prices set in large and competitive financial markets or real economic developments. In particular, in Chapter 1 I review the behavioural finance literature to see whether the 'anomalies' that this literature has identified could affect market prices. I find many examples in which it could be the case. This is particularly true for anomalies which are widespread among economic agents, for instance because of a social bias. In Chapter 2, I look at how a central bank which is characterised by reference dependence and diminishing sensitivity with respect to deviations of inflation from its target, and which weighs probabilities of uncertain events non-linearly. I find that it can behave in a way which is different from what standard models suggest, in particular by violating the certainly equivalence principle. In Chapter 3, I see how it is optimal, under quite general conditions, for a prospect theory agent to concentrate, rather than diversify risks. Finally, in Chapter 4, I show how loss and disappointment aversion may explain the equity premium puzzle if different assumptions are maintained on the representative agent's investment time horizon and reference point.