We consider prices across the advance bookings as a multivariate variable, modelling the time effect by means of a structural vector autoregression. Its unit-specific fixed effect is taken as the dependent variable in a set of hedonic regressions allowing us to estimate the marginal contribution to price of the (time-invariant) attributes along the price trajectory. Results from our two-step approach indicate that although past occupancy rates, events, bank holidays, and seasonality differentiate pricing strategies across advance bookings, time-invariant services/attributes tend to increase in hedonic value as the advance booking period decreases, reducing the contribution of dynamic pricing.
K E Y W O R D Sbusiness tourism, dynamic pricing, fairs, Milan, seasonality, structural vector autoregression