A DSGE model is dynamic in the sense that it explains how the economy evolves
over time. It is stochastic because agents know only the distribution of future shocks thereafter their expected value is zero. Thus, only when these models are linearized to the first order do agents behave as if future shocks are equal to zero, which is the
certainty equivalence property. Finally, it is based on a general equilibrium framework as it depicts the macroeconomy as the sum of individual choices and decisions made by firms, households, the government, and the central bank, according to their own
preferences and views about the future.
This paper presents the initial specification of and results of the BSP’s DSGE model for the Philippine economy. The development of the model complements existing models used by the BSP for policy simulation. The basic model starts at the level of individuals and firms, which are assumed to make rational decisions on how much to save, spend or invest based on their preferences and available choices. In this approach, the macroeconomy is seen as the sum of individual choices and decisions (the
microeconomy). This basic framework is extended to an open economy setting and
embeds price and financial rigidities.
The paper employs Bayesian estimation, which works best with state of the art
theoretical models. In this method, one can formulate prior distributions about the structural parameters and not about reduced form coefficients. Posterior distributions of the parameters can be obtained by calculating the likelihood functions based on observed data. Since the method works with likelihood functions, it actually goes beyond
matching selected moments of artificially-generated data with actual data. With the advances in approximation methods, random number generators and sampling method like the Monte Carlo Markov Chain (MCMC), there is no need to restrict the parameter
distribution to normal distributions.
