Bayesian Computations for DSGE Models

Transcription

1 Bayesian Computations for DSGE Models Frank Schorfheide University of Pennsylvania, PIER, CEPR, and NBER October 23, 2017

2 This Lecture is Based on Bayesian Estimation of DSGE Models Edward P. Herbst & Frank Schorfheide companion-web-site-bayesian-estimation-of-dsge-models

3 Some Background DSGE model: dynamic model of the macroeconomy, indexed by θ vector of preference and technology parameters. Used for forecasting, policy experiments, interpreting past events. Ingredients of Bayesian Analysis: Likelihood function p(y θ) Prior density p(θ) Marginal data density p(y ) = p(y θ)p(θ)dφ Bayes Theorem: p(θ Y ) = p(y θ)p(θ) p(y ) p(y θ)p(θ)

4 Some Background Implementation: usually by generating a sequence of draws (not necessarily iid) from posterior θ i p(θ Y ), i = 1,..., N Algorithms: direct sampling, accept/reject sampling, importance sampling, Markov chain Monte Carlo sampling, sequential Monte Carlo sampling... Draws can then be transformed into objects of interest, h(θ i ), and under suitable conditions a Monte Carlo average of the form h N = 1 N N i=1 h(θ i ) E π [h] h(θ)p(θ Y )dθ. Strong law of large numbers (SLLN), central limit theorem (CLT)...

5 Bayesian Inference Decision Making The posterior expected loss of decision δ( ): ρ ( δ( ) Y ) = L ( θ, δ(y ) ) p(θ Y )dθ. Θ Bayes decision minimizes the posterior expected loss: δ (Y ) = argmin d ρ ( δ( ) Y ). Approximate ρ ( δ( ) Y ) by a Monte Carlo average ρ N ( δ( ) Y ) = 1 N Then compute N L ( θ i, δ( ) ). i=1 δ N(Y ) = argmin d ρ N ( δ( ) Y ).

6 Bayesian Inference Point estimation: Quadratic loss: posterior mean Absolute error loss: posterior median Interval/Set estimation P π {θ C(Y )} = 1 α: highest posterior density sets equal-tail-probability intervals

7 Computational Challenges Numerical solution of model leads to state-space representation = likelihood approximation = posterior sampler. Standard approach for (linearized) models Model solution: log-linearize and use linear rational expectations system solver. Evaluation of p(y θ): Kalman filter Posterior draws θ i : MCMC Book reviews the standard approach, but also studies more recently developed sequential Monte Carlo (SMC) techniques.

8 Sequential Monte Carlo (SMC) Methods SMC can help to Approximate the likelihood function (particle filtering): Gordon, Salmond, and Smith (1993)... Fernandez-Villaverde and Rubio-Ramirez (2007) approximate the posterior of θ: Chopin (2002)... Durham and Geweke (2013)... Creal (2007), Herbst and Schorfheide (2014) or both: SMC 2 : Chopin, Jacob, and Papaspiliopoulos (2012)... Herbst and Schorfheide (2015)

9 Review Importance Sampling

10 Importance Sampling Approximate π( ) by using a different, tractable density g(θ) that is easy to sample from. For more general problems, posterior density may be unnormalized. So we write π(θ) = p(y θ)p(θ) p(y ) = f (θ) f (θ)dθ. Importance sampling is based on the identity E π [h(θ)] = h(θ)π(θ)dθ = f (θ) h(θ) Θ Θ g(θ) g(θ)dθ f (θ) g(θ) g(θ)dθ. (Unnormalized) importance weight: w(θ) = f (θ) g(θ).

11 Importance Sampling 1 For i = 1 to N, draw θ i iid g(θ) and compute the unnormalized importance weights w i = w(θ i ) = f (θi ) g(θ i ). 2 Compute the normalized importance weights W i = 1 N w i N i=1 w i. An approximation of E π [h(θ)] is given by h N = 1 N N W i h(θ i ). i=1

12 Illustration If θ i s are draws from g( ) then E π [h] 1 N N i=1 h(θi )w(θ i ) 1 N N i=1 w(θi ), w(θ) = f (θ) g(θ) f 0.30 g g weights f/g 1 f/g

13 Accuracy Since we are generating iid draws from g(θ), it s fairly straightforward to derive a CLT: N( h N E π [h]) = N ( 0, Ω(h) ), where Ω(h) = V g [(π/g)(h E π [h])]. Using a crude approximation (see, e.g., Liu (2008)), we can factorize Ω(h) as follows: Ω(h) V π [h] ( 1 + V g [π/g] ). The approximation highlights that the larger the variance of the importance weights, the less accurate the Monte Carlo approximation relative to the accuracy that could be achieved with an iid sample from the posterior. Users often monitor ESS = N V π[h] Ω(h) N 1 + V g [π/g].

14 Likelihood Approximation

15 State-Space Representation and Likelihood Measurement Equation: y t = Ψ(s t ; θ) +u }{{} t optional State transition: s t = Φ(s t 1, ɛ t ; θ) Joint density for the observations and latent states: T T p(y 1:T, S 1:T θ) = p(y t, s t Y 1:t 1, S 1:t 1, θ) = p(y t s t, θ)p(s t s t 1, θ). t=1 Need to compute the marginal p(y 1:T θ). t=1

16 Filtering - General Idea State-space representation of nonlinear DSGE model Measurement Eq. : y t = Ψ(s t, t; θ) + u t, u t F u ( ; θ) State Transition : s t = Φ(s t 1, ɛ t ; θ), ɛ t F ɛ ( ; θ). Likelihood function: p(y 1:T θ) = T p(y t Y 1:t 1, θ) t=1 A filter generates a sequence of conditional distributions s t Y 1:t. Iterations: Initialization at time t 1: p(s t 1 Y 1:t 1, θ) Forecasting t given t 1: 1 Transition equation: p(s t Y 1:t 1, θ) = p(s t s t 1, Y 1:t 1, θ)p(s t 1 Y 1:t 1, θ)ds t 1 2 Measurement equation: p(y t Y 1:t 1, θ) = p(y t s t, Y 1:t 1, θ)p(s t Y 1:t 1, θ)ds t Updating with Bayes theorem. Once y t becomes available: p(s t Y 1:t, θ) = p(s t y t, Y 1:t 1, θ) = p(yt st, Y1:t 1, θ)p(st Y1:t 1, θ) p(y t Y 1:t 1, θ)

17 Conditional Distributions for Kalman Filter (Linear Gaussian State-Space Model) Distribution s t 1 (Y 1:t 1, θ) N ( s ) t 1 t 1, P t 1 t 1 s t (Y 1:t 1, θ) N ( s t t 1, P t t 1 ) y t (Y 1:t 1, θ) N ( ȳ t t 1, F t t 1 ) s t (Y 1:t, θ) N ( s t t, P t t ) Mean and Variance Given from Iteration t 1 s t t 1 = Φ 1 s t 1 t 1 P t t 1 = Φ 1 P t 1 t 1 Φ 1 + Φ ɛσ ɛ Φ ɛ ȳ t t 1 = Ψ 0 + Ψ 1 t + Ψ 2 s t t 1 F t t 1 = Ψ 2 P t t 1 Ψ 2 + Σ u s t t = s t t 1 + P t t 1 Ψ 2 F 1 t t 1 (y t ȳ t t 1 ) P t t = P t t 1 P t t 1 Ψ 2 F 1 t t 1 Ψ 2P t t 1

18 Bootstrap Particle Filter 1 Initialization. Draw the initial particles from the distribution s j iid 0 p(s0 ) and set W j 0 = 1, j = 1,..., M. 2 Recursion. For t = 1,..., T : 1 Forecasting s t. Propagate the period t 1 particles {s j t 1, W j t 1 } by iterating the state-transition equation forward: s j t = Φ(s j t 1, ɛj t; θ), ɛ j t F ɛ( ; θ). (1) An approximation of E[h(s t) Y 1:t 1, θ] is given by ĥ t,m = 1 M M j=1 h( s j t )W j t 1. (2)

19 Bootstrap Particle Filter 1 Initialization. 2 Recursion. For t = 1,..., T : 1 Forecasting s t. 2 Forecasting y t. Define the incremental weights w j t = p(y t s j t, θ). (3) The predictive density p(y t Y 1:t 1, θ) can be approximated by ˆp(y t Y 1:t 1, θ) = 1 M M j=1 w j t W j t 1. (4) If the measurement errors are N(0, Σ u) then the incremental weights take the form { w j t = (2π) n/2 Σ u 1/2 exp 1 ( yt Ψ( s j t, t; θ) ) Σ 1 ( u yt Ψ( s j t, t; θ) )}, (5) 2 where n here denotes the dimension of y t.

20 Bootstrap Particle Filter 1 Initialization. 2 Recursion. For t = 1,..., T : 1 Forecasting s t. 2 Forecasting y t. Define the incremental weights w j t = p(y t s j t, θ). (6) 3 Updating. Define the normalized weights W j t = 1 M w j t W j t 1 M j=1 w. j t W j t 1 (7) An approximation of E[h(s t) Y 1:t, θ] is given by h t,m = 1 M M j=1 h( s j t ) W j t. (8)

21 Bootstrap Particle Filter 1 Initialization. 2 Recursion. For t = 1,..., T : 1 Forecasting s t. 2 Forecasting y t. 3 Updating. 4 Selection (Optional). Resample the particles via multinomial resampling. Let {s j t } M j=1 denote M iid draws from a multinomial distribution characterized by support points and weights { s j t, W j t } and set W j t = 1 for j =, 1..., M. An approximation of E[h(s t) Y 1:t, θ] is given by h t,m = 1 M M j=1 h(s j t )W j t. (9)

22 Likelihood Approximation The approximation of the log likelihood function is given by T ln ˆp(Y 1:T θ) = ln 1 M w t j W j t 1. (10) M t=1 j=1 One can show that the approximation of the likelihood function is unbiased. This implies that the approximation of the log likelihood function is downward biased.

23 The Role of Measurement Errors Measurement errors may not be intrinsic to DSGE model. Bootstrap filter needs non-degenerate p(y t s t, θ) for incremental weights to be well defined. Decreasing the measurement error variance Σ u, holding everything else fixed, increases the variance of the particle weights, and reduces the accuracy of Monte Carlo approximation.

24 Generic Particle Filter 1 Initialization. Same as BS PF 2 Recursion. For t = 1,..., T : 1 Forecasting s t. Draw s j t from density g t( s t s j t 1, θ) and define ω j t = p( sj t s j t 1, θ) g t( s j t s j t 1, θ). (11) An approximation of E[h(s t) Y 1:t 1, θ] is given by ĥ t,m = 1 M h( s j t )ω j M tw j t 1. (12) j=1 2 Forecasting y t. Define the incremental weights w j t = p(y t s j t, θ)ω j t. (13) The predictive density p(y t Y 1:t 1, θ) can be approximated by 3 (...) ˆp(y t Y 1:t 1, θ) = 1 M M j=1 w j t W j t 1. (14)

25 Adapting the Generic PF Conditionally-optimal importance distribution: g t ( s t s j t 1 ) = p( s t y t, s j t 1 ). This is the posterior of s t given s j t 1. Typically infeasible, but a good benchmark. Approximately conditionally-optimal distributions: from linearize version of DSGE model or approximate nonlinear filters. Conditionally-linear models: do Kalman filter updating on a subvector of s t. Example: y t = Ψ 0 (m t ) + Ψ 1 (m t )t + Ψ 2 (m t )s t + u t, u t N(0, Σ u ), s t = Φ 0 (m t ) + Φ 1 (m t )s t 1 + Φ ɛ (m t )ɛ t, ɛ t N(0, Σ ɛ ), where m t follows a discrete Markov-switching process.

26 Distribution of Log-Likelihood Approximation Errors Bootstrap PF: θ m vs. θ l Notes: Density estimate of ˆ 1 = ln ˆp(Y 1:T θ) ln p(y 1:T θ) based on N run = 100 runs of the PF. Solid line is θ = θ m ; dashed line is θ = θ l (M = 40, 000).

27 Distribution of Log-Likelihood Approximation Errors θ m : Bootstrap vs. Cond. Opt. PF Notes: Density estimate of ˆ 1 = ln ˆp(Y 1:T θ) ln p(y 1:T θ) based on N run = 100 runs of the PF. Solid line is bootstrap particle filter (M = 40, 000); dotted line is conditionally optimal particle filter (M = 400).

28 Great Recession and Beyond Log Standard Dev of Log-Likelihood Increments Notes: Solid lines represent results from Kalman filter. Dashed lines correspond to bootstrap particle filter (M = 40, 000) and dotted lines correspond to conditionally-optimal particle filter (M = 400). Results are based on N run = 100 runs of the filters.

29 Posterior Sampling

30 Implementation: Sampling from Posterior DSGE model posteriors are often non-elliptical, e.g., multimodal posteriors may arise because it is difficult to disentangle internal and external propagation mechanisms; disentangle the relative importance of shocks. θ θ Economic Example: is wage growth persistent because 1 wage setters find it very costly to adjust wages? 2 exogenous shocks affect the substitutability of labor inputs and hence markups?

31 Sampling from Posterior If posterior distributions are irregular, standard MCMC methods can be inaccurate (examples will follow). SMC samplers often generate more precise approximations of posteriors in the same amount of time. SMC can be parallelized. SMC = importance sampling on steroids For now we assume that the likelihood evaluation is exact.

32 From Importance Sampling to Sequential Importance Sampling In general, it s hard to construct a good proposal density g(θ), especially if the posterior has several peaks and valleys. Idea - Part 1: it might be easier to find a proposal density for π n (θ) = [p(y θ)]φn p(θ) = f n(θ). [p(y θ)] φ np(θ)dθ Z n at least if φ n is close to zero. Idea - Part 2: We can try to turn a proposal density for π n into a proposal density for π n+1 and iterate, letting φ n φ N = 1.

33 Illustration: Tempered Posteriors of θ n θ π n (θ) = [p(y p(θ) θ)]φn = f ( ) λ n(θ) n, φ [p(y θ)] φ np(θ)dθ n = Z n N φ

34 SMC Algorithm: A Graphical Illustration C S M C S M C S M φ 0 φ 1 φ 2 φ 3 π n (θ) is represented by a swarm of particles {θ i n, W i n} N i=1 : h n,n = 1 N N i=1 W i nh(θ i n) a.s. E πn [h(θ n )]. C is Correction; S is Selection; and M is Mutation.

35 Remarks Correction Step: reweight particles from iteration n 1 to create importance sampling approximation of E πn [h(θ)] Selection Step: the resampling of the particles (good) equalizes the particle weights and thereby increases accuracy of subsequent importance sampling approximations; (not good) adds a bit of noise to the MC approximation. Mutation Step: changes particle values adapts particles to posterior π n(θ); imagine we don t do it: then we would be using draws from prior p(θ) to approximate posterior π(θ), which can t be good! n θ1

36 More on Transition Kernel in Mutation Step Transition kernel K n (θ ˆθ n 1 ; ζ n ): generated by running M steps of a Metropolis-Hastings algorithm. Lessons from DSGE model MCMC: blocking of parameters can reduces persistence of Markov chain; mixture proposal density avoids getting stuck. Blocking: Partition the parameter vector θ n into N blocks equally sized blocks, denoted by θ n,b, b = 1,..., N blocks. (We generate the blocks for n = 1,..., N φ randomly prior to running the SMC algorithm.) Example: random walk proposal density: ( ) ϑ b (θn,b,m 1, i θn, b,m, i Σ n,b) N θn,b,m 1, i cnσ 2 n,b.

37 Application: Estimation of Smets and Wouters (2007) Model Benchmark macro model, has been estimated many (many) times. Core of many larger-scale models. 36 estimated parameters. RWMH: 10 million draws (5 million discarded); SMC: 500 stages with 12,000 particles. We run the RWM (using a particular version of a parallelized MCMC) and the SMC algorithm on 24 processors for the same amount of time. We estimate the SW model twenty times using RWM and SMC and get essentially identical results.

38 Application: Estimation of Smets and Wouters (2007) Model More interesting question: how does quality of posterior simulators change as one makes the priors more diffuse? Replace Beta by Uniform distributions; increase variances of parameters with Gamma and Normal prior by factor of 3.

39 SW Model with DIFFUSE Prior: Estimation stability RWH (black) versus SMC (red) l ι w µ p µ w ρ w ξ w r π

40 A Measure of Effective Number of Draws Suppose we could generate iid N eff ( Ê π [θ] approx N E π [θ], 1 N eff V π [θ] draws from posterior, then ). We can measure the variance of Ê π [θ] by running SMC and RWM algorithm repeatedly. Then, N eff V π[θ] V [ Ê π [θ] ]

41 Effective Number of Draws SMC RWMH Parameter Mean STD(Mean) N eff Mean STD(Mean) N eff σ l l ι p h Φ r π ρ b ϕ σ p ξ p ι w µ p ρ w µ w ξ w

42 A Closer Look at the Posterior: Two Modes Parameter Mode 1 Mode 2 ξ w ι w ρ w µ w Log Posterior Mode 1 implies that wage persistence is driven by extremely exogenous persistent wage markup shocks. Mode 2 implies that wage persistence is driven by endogenous amplification of shocks through the wage Calvo and indexation parameter. SMC is able to capture the two modes.

43 A Closer Look at the Posterior: Internal ξ w versus External ρ w Propagation ρ w ξ w

44 Stability of Posterior Computations: RWH (black) versus SMC (red) P (ξ w > ρ w) P (ρ w > µ w) P (ξ w > µ w) P (ξ p > ρ p) P (ρ p > µ p) P (ξ p > µ p)

45 Embedding PF Likelihoods into Posterior Samplers Particle MCMC, SMC 2. Distinguish between: {p(y θ), p(θ Y ), p(y )}, which are related according to: p(y θ)p(θ) p(θ Y ) =, p(y ) = p(y θ)p(θ)dθ p(y ) {ˆp(Y θ), ˆp(θ Y ), ˆp(Y )}, which are related according to: ˆp(Y θ)p(θ) ˆp(θ Y ) =, ˆp(Y ) = ˆp(Y θ)p(θ)dθ. ˆp(Y ) Surprising result (Andrieu, Docet, and Holenstein, 2010): under certain conditions we can replace p(y θ) by ˆp(Y θ) and still obtain draws from p(θ Y ).