Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Advanced Macroeconomics II Summer semester 2016/2017 Warsaw School of Economics Instructor: Marcin Kolasa marcin.kolasa(at)sgh.waw.pl Assignment 1 Due date: 22 March 2017, 9.50 a.m. 1. Consider a standard real business cycle model. The equilibrium of such a model can be characterized by the following set of log-linearized equations: kt+1 = β −1 kt + b [At + (1 − α)lt − ct ] + δct α −θct = −θEt {ct+1 } + βbEt {At+1 + (α − 1)kt+1 + (1 − α)lt+1 } ϕlt + θct = At + αkt − αlt At+1 = ρAt + εt+1 where εt ∼ N (0, σ 2 ) and b = β −1 − 1 + δ. (a) Write the model in a matrix form as in formula (1) from the auxiliary material ”Solving DSGE models...”. Give matrices A1 , A0 and γ as a solution. Can this system be solved using the standard (i.e. based on the Jordan decomposition) Blanchard-Kahn algorithm? Why? (0.5p) (b) Use the third of the above four equations to eliminate lt . Write the thus obtained system of three equations in a matrix form, giving matrices A1 , A0 and γ. (0.5p) (c) Assume β = 0.99, α = 0.33, θ = 2, δ = 0.025, ϕ = 1, ρ = 0.95 and σ = 0.01. Solve numerically the model 0 obtained in (b), writing it as Wt = ΦWt−1 + ΨZt , where Wt = At kt ct . Give Φ and Ψ as a solution. HINT: You can use a modified version of the Matlab code given in the auxiliary material ”Implementing the Blanchard-Kahn algorithm...”. (1p) (d) Use the production function yt = At + αkt + (1 − α)lt to augment Wt with yt and calculate the new Φ and Ψ. Assuming that the economy at time t = 0 is in the steady state, calculate the response of output yt to a typical positive technology shock (i.e. ε1 = σ, εt = 0 for t > 1) for the first 100 periods. Print or sketch a chart as a solution, reporting the values at t = 1, 4, 8, 16. (1p) 1 2. Consider the standard small-scale log-linearized New Keynesian model given by the following equations: xt = Et {xt+1 } − 1 (Rt − Et {πt+1 } − rtn ) θ πt = βEt {πt+1 } + κxt where κ > 0, θ > 0, 0 < β < 1, while rtn is the natural interest rate that is a function of the exogenous disturbances and model parameters only. Consider a monetary policy rule of the form: Rt = rtn + φπ πt + φx xt Find all pairs of non-negative φπ and φx for which the model has a unique solution. What can you say about inflation and the output gap under this policy rule? (2p) 3. Consider the New Keynesian model in the following log-linearized form: xt = Et {xt+1 } − 1 (Rt − Et {πt+1 }) + εdt θ πt = βEt {πt+1 } + κxt + εst . The demand shock εdt and the supply (cost-push) shock where κ = (1−γ)(1−βγ)(θ+ϕ) γ εst follow independent AR(1) processes with autoregressive coefficients ρd and ρs , and standard deviations of innovations σd and σs , respectively. (a) Assume that the central bank acts such that inflation πt is zero at all times. How does the output gap xt and the nominal interest rate Rt respond to demand and supply shocks? As your answer derive the formulas in which both variables depend on εdt , εst and the model parameters only. (1p) (b) Do the same exercise for an alternative monetary policy variant that implies zero output gap xt at every period. (1p) (c) Can the formulas for Rt derived in (a) and (b) be considered as monetary policy rules that ensure perfect stabilization of, respectively, inflation and the output gap? If not, give examples of such rules that achieve these goals. (0.5p) (d) Based on the results obtained in (a) and (b), derive the formulas for ergodic variances of the output gap E(x2 ) and inflation E(π 2 ) for both monetary policy variants (such that these moments depend on the model parameters only). (1p) (e) Assume that the government subsidies production such that social welfare is proportional to −E(π 2 ) − φκ E(x2 ). Show for what values of κ strict stabilization of inflation is better for welfare than strict stabilization of the output gap. Which of these extreme policies will be preferred if the degree of price stickiness is low, and which if it is high? (0.5p) 2