Staff Working Paper No. 1,197
Jamie Lenney and Biagio Rosso
This paper proposes a flexible approach to departing from full information rational expectations (FIRE) in DSGE models using the sequence space framework. We implement a reduced-form behavioural expectations process in which agents can simultaneously overreact or underreact to current economic conditions and underreact to news, and we derive an associated behavioural expectations solver deployable to a wide class of DSGE and HANK models. The approach nests several different expectations models as special cases while remaining agnostic on the precise source of belief frictions. We apply it to a medium-scale two-asset HANK model and jointly estimate the model’s dynamic and behavioural parameters on US business cycle data, including inflation expectation data. Behavioural expectations quantitatively improve the empirical fit of the model and the qualitative properties of its impulse response functions.
A flexible deviation from FIRE in the sequence space