Dynamic Programming. Macro 3: Lecture 2. Mark Huggett 2. 2 Georgetown. September, 2016

Size: px
Start display at page:

Download "Dynamic Programming. Macro 3: Lecture 2. Mark Huggett 2. 2 Georgetown. September, 2016"

Transcription

1 Macro 3: Lecture 2 Mark Huggett 2 2 Georgetown September, 2016

2 Three Maps: Review of Lecture 1 X = R 1 + and X grid = {x 1,..., x n } X where x i+1 > x i 1. T (v)(x) = max u(f (x) y) + βv(y) s.t. y Γ 1 (x) Γ 1 (x) = {y X : 0 y F (x)} 2. T (v)(x) = max u(f (x) y) + βv(y) s.t. y Γ2 (x) Γ 2 (x) = {y X grid : 0 y F (x)} 3. T next page Map 1: T is standard DP map Map 2: T is finite DP map Map 3: T is piecewise-linear DP map While Map 1 cannot quite be executed on a computer, both Map 2 and Map 3 can be at least approximately.

3 X grid = {x 1,..., x n } X α(x) (x x i )/(x i+1 x i ) for x i x x i+1 { T max u(f (x) y) + βv(y) s.t. y Γ1 (x) : x X (v)(x) = grid (1 α(x))t (v)(x i ) + α(x)t (v)(x i+1 ) : x X grid

4 One could prove the following: T maps B(X, R) into itself and is a contraction map. T maps B(Xgrid, R) into itself and is a contraction map. T maps P L(X, R) into itself and is a contraction map. Thus, one can apply the Contraction Mapping Theorem to get the existence of a unique fixed point for each map. Even though one might view both ( T, T ) as simply computational tools, they both can be analyzed at a theoretical level. At a minimum, this is potentially useful for discovering errors in ones code.

5 Homework 1-2: Message Cake Eating Problem: All methods (Map 2-3) poorly approximate the theoretical value function at low asset levels. Optimal Growth Problem: All methods (Map 2-3) seem to work well for a range of values of the state variable. Approximate convergence of all algorithms is rapid as β << 1. True optimal decision rule is well approximated (at least at grid points) by the computed decision rules in Maps 2-3. Map 3 produces a monotone decision rule. Could one prove that this must be true regardless of the precise choice of grid? If so, useful check on error(s) in your code.

6 Construct a New Map Based on Euler Eqn Note two things about the original problem: 1. one could prove that T maps increasing, concave and differentiable functions into the same function space. 2. could carry out the T map by using first-order conditions v j+1 (x) = T (v j )(x) = max u(f (x) y) + βv j (y) s.t. y Γ 1 (x) Sufficient condition for y j+1 (x) to solve RHS of T map: u (F (x) y j+1 (x)) βv j(y j+1 (x)) = when y j+1 (x) > 0 Useful Recursion: v j+1 (x) = u (F (x) y j+1 (x))f (x)

7 Algorithm Based on Euler Eqn 1. Guess v 1 (x) decreasing for x X grid. 2. Given v j, solve for y j+1(x) = y for x X grid. Use ( ) or ( ) ( ) u (F (x) y) βv j(y) and = if y > 0 ( ) G(x, y) u (F (x) y) βv j(y) 0 and = if y > 0 3. Set v j+1 (x) = u (F (x) y j+1 (x))f (x) for x X grid. 4. Repeat 2-3 until approx convergence. Get y(x) for x X grid. 5. Approximate v(x), where y is the computed decision rule, by interating on v j+1 (x) = u(f (x) y(x)) + βv j (y(x)).

8 Algorithm Based on Euler Eqn Comments 1. Algorithm uses Euler eqn to do maximization. Thus, CMT applies to this algorithm. Fixed point is w/in class of differentiable functions v instead of piecewise linear v. 2. To solve the non-linear eqn need to interpolate v j because we only calculate it at x X grid. Use linear interpolation... or a continuous interpolant consistent with v j decreasing. 3. Useful to graph G to see what solving the non-linear eqn involves... and see downside to some interpolants. 4. Instead of solving for choice given state, could pick y and solve for the state x. Method of endogenous grid points

9 Endogenous Grid Method 1. Specify Y grid = {y 1,..., y n }, where y i < y i+1, and specify arbitrary {v 1 (y i)} decreasing. 2. Solve for state x i, given choice y i Y grid using u (F (x i ) y i ) βv 1 (y i) and = if y i > 0. Get {(x i, y i )} n i=1. 3. Set v 2 (x i) = u (F (x i ) y i )F (x i ) for x i X grid, which is the ENDOGENOUS grid constructed in step Repeat 2-3 until approximate convergence. Interpolate at each iteration to get v 2 on Y grid rather than X grid. 5. Result: Set {(x i, y(x i ))} n i=1 = {(x i, y i )} n i=1 based on last iteration.

10 Endogenous Grid Method: Comments 1. Roughly put, the algorithm solves for x given y using F (x) y = (u ) 1 [βv j (y)]. 2. This is much faster than the previous method. Instead of solving a nonlinear equation by several trial calculations or by your favorite non-linear eqn solver, you solve the equation above by one function evaluation. 3. The method is the Euler equation method but with a faster implementation of solving the Euler equation. 4. To keep the Y grid fixed across iterations on the previous slide you need to interpolate v 2 to figure out values on Y grid. 5. Method can be extended to handle shocks and several control variables for some cases.

11 How to Approximate a Markov process Issue: Choose finite Markov chain to approx a VAR or an AR-1 process Motivation: VAR is an empirical summary of a time series process. Use a VAR (or AR-1) as an input to an economic model as a driving process. Specific Paper: Tauchen (1986) - simple method 1. EX 1: x F ( ) and E[x t ] = 0 iid process on R 1 2. EX 2: x t+1 = ρx t + ɛ t+1 and E[ɛ t+1 ] = 0 AR-1

12 Tauchen: EX 1 iid shocks 1. Set {x 1,..., x n } 1. (increasing) x i+1 > x i, i 2. (synmmetry) x 1 = x n 3. (equispaced) x i+1 x i = w > Set {p 1,..., p n } 1. p 1 = F ((x 1 + x 2 )/2) 2. p i = F ((x i + x i+1 )/2) F ((x i 1 + x i )/2) 3. p n = 1 F ((x n 1 + x n )/2)

13 Tauchen Method: Let F be defined on an interval [a, b]. Then the distribution function F n produced by an n-point Tauchen procedure on [a, b] converges weakly to F. Proof by picture. Comments: 1. Weak convergence is equivalent to lim n gdfn = gdf for all g continuous. Thus, in the conext of DP, the discretized process will in the limit get the expectation term in Bellman s equation right for continuous objectives. 2. While Tauchen is simple(r), quadrature methods will have smaller integration errors.. In some applications, may want discretized process to integrate well AND cover a wide range of shock values. Quadrature does former but not the latter.

14 EX 2: x t+1 = ρx t + ɛ t+1 and E[ɛ t+1 ] = 0, ɛ F 1. Set {x 1,..., x n } V ar(ɛ 2 ) = σ 2 ɛ and σ 2 σ 2 ɛ /(1 ρ) 2 1. (symmetry) x 1 = x n = 3σ 2. (equi spaced) x i+1 x i = w > Set p i,j 1. p i,1 = F (x 1 + w/2 ρx i ) 2. p i,j = F ((x i + x i+1 )/2) F ((x i 1 + x i )/2) for 1 < j < n 3. p i,n = 1 F (x n w/2 ρx i ) NOTE: p i,1 = P r{x t+1 = ρx t + ɛ t+1 x 1 + w/2 x t = x i } p i,1 = P r{ɛ : ɛ x 1 + w/2 ρx i } = F (x 1 + w/2 ρx i )

15 Numerical Quadrature Issue: Given (f, w), compute ( ) or ( ) by choosing N pairs (x k, w k ): f(x)dx = N k=1 f(x k)w k ( ) f(x)w(x)dx = N k=1 f(x k)w k ( ) f - function to be integrated w - weighing function (density) (x k, w k ) - quadrature points and weights Newton Quadrature: Given {x k }, choose {w k } in a smart way. Gaussian Quadrature: Choose both {x k } and {w k } in a smart way.

16 Example: Trapezoidal Rule [x 0,x n] f(x)dx h 2 [f(x 0) + 2f(x 1 ) + + 2f(x n 1 ) + f(x n )] h x k x k 1, k Graph it to be clear. The Trapezoidal Rule is like Newton quadrature since weights are chosen but not the quadrature points.

17 Claim: N point Guassian quadrature rule integrates polynomials of order 2N 1 exactly and weights satisfy w k > 0 and k w k = 1. Points and weights satisfy: (i) 1w(x)dx = k 1w k (ii) (iii) xw(x)dx = k x 2 w(x)dx = k x k w k x 2 k w k Idea: Know LHS of (i)-(iii) for specific w functions. Choose RHS variables to solve system of equations.

18 In practice, for many choices of densities w( ), the N quadrature points and weights have already been solved for you. You just need to look them up in textbook quadrature tables. Could choose low N value as N = 6 implies that you would integrate exactly all polynomials of degree 11. So if the only point of quadrature in applications is to integrate the expectation term in Bellman s equation accurately and you conjecture that v is smooth, then not many quadrature points are needed.

19 Application: Based on Gauss-Hermite f(x) exp( x2 )dx v(x) = max u(x a ) + βe[v(x )] s.t. 0 a x x = exp(e ) + a (1 + r) and e N(µ, σ 2 ) Apply quadrature to get (z (e µ)/( 2σ)). E[v(x )] =. 1 N v(a(1 + r) + exp( 2σz i + µ)w i π i=1 Some steps behind this are on the next slide.

20 Application: Warm up Step 1: E[f(y)] = 1 2πσ Step 2: x = (y µ) 2σ Step 3: Step 4: E[f(y)] = 1 2πσ E[f(y)] = 1 π (y µ)2 f(y) exp( )dy 2σ 2 and y = 2σx + µ and dy = 2σdx f( 2σx + µ)exp( x 2 ) 2σdx f( 2σx + µ)exp( x 2 )dx E[f(y)]. = 1 π N f( 2σx i + µ)w i i=1

21 E[v(exp(e ) + a (1 + r))] = 1 2π v() exp( (e µ) 2 2σ 2 )de z (e µ) 2σ z N(0, 1), e = 2σz + µ de /dz = 2σ E[v(exp(e ) + a (1 + r))] = v(a (1 + r) + exp( 2σz + µ)) exp( z 2 )dz 1 π E[v(exp(e ) + a (1 + r))] = 1 π N v(a (1 + r) + exp( 2σz i + µ))w i k=1

22 How to Simulate a Markov Process? Context: Suppose you have a computed decision rule a(a t, e t ) solving the income fluctuation problem. Earnings are Markov. Create draws {e t } T t=0 and {a(a t, e t )} T t=0 starting from a known initial state (a 0, e 0 ). Suppose e t takes discrete values in E = {1, 2, 3} with transition probabilities π i,j. 1. Create draws {u t } T t=1 on [0, 1] from a random number generator. 2. Suppose e 0 = i. Set e 1 = 1 if u 1 [0, π i1 ), e 1 = 2 if u 1 [π i1, π i1 + π i2 ), and e 1 = 3 if u 1 [π i1 + π i2, 1). 3. Repeat step 2 for t = 1, 2,.. using the intervals [0, π j1 ), [π j1, π j1 + π j2 ) and [π j1 + π j2, 1) if e t = j. 4. Calculate {a t } T t=0 using {a(a t, e t )} T t=0, based on the computed decision rule and the shock draw from steps 2-3

23 Psuedo Code 1. oldindex=1 2. y(1) = e(oldindex) 3. do it =2, nt u = randomnumber(0,1) index = 1 sum = pi(oldindex, index) do while (u.ge. sum) index = index + 1 sum = sum + pi(oldindex, index) end do y(it)=e(index) oldindex = index 4. end do

Macro 1: Dynamic Programming 1

Macro 1: Dynamic Programming 1 Macro 1: Dynamic Programming 1 Mark Huggett 2 2 Georgetown September, 2016 DP Warm up: Cake eating problem ( ) max f 1 (y 1 ) + f 2 (y 2 ) s.t. y 1 + y 2 100, y 1 0, y 2 0 1. v 1 (x) max f 1(y 1 ) + f

More information

Macro 1: Dynamic Programming 2

Macro 1: Dynamic Programming 2 Macro 1: Dynamic Programming 2 Mark Huggett 2 2 Georgetown September, 2016 DP Problems with Risk Strategy: Consider three classic problems: income fluctuation, optimal (stochastic) growth and search. Learn

More information

LECTURE 16 GAUSS QUADRATURE In general for Newton-Cotes (equispaced interpolation points/ data points/ integration points/ nodes).

LECTURE 16 GAUSS QUADRATURE In general for Newton-Cotes (equispaced interpolation points/ data points/ integration points/ nodes). CE 025 - Lecture 6 LECTURE 6 GAUSS QUADRATURE In general for ewton-cotes (equispaced interpolation points/ data points/ integration points/ nodes). x E x S fx dx hw' o f o + w' f + + w' f + E 84 f 0 f

More information

Lecture 4: Dynamic Programming

Lecture 4: Dynamic Programming Lecture 4: Dynamic Programming Fatih Guvenen January 10, 2016 Fatih Guvenen Lecture 4: Dynamic Programming January 10, 2016 1 / 30 Goal Solve V (k, z) =max c,k 0 u(c)+ E(V (k 0, z 0 ) z) c + k 0 =(1 +

More information

Higher-Order Dynamics in Asset-Pricing Models with Recursive Preferences

Higher-Order Dynamics in Asset-Pricing Models with Recursive Preferences Higher-Order Dynamics in Asset-Pricing Models with Recursive Preferences Walt Pohl Karl Schmedders Ole Wilms Dept. of Business Administration, University of Zurich Becker Friedman Institute Computational

More information

Discrete State Space Methods for Dynamic Economies

Discrete State Space Methods for Dynamic Economies Discrete State Space Methods for Dynamic Economies A Brief Introduction Craig Burnside Duke University September 2006 Craig Burnside (Duke University) Discrete State Space Methods September 2006 1 / 42

More information

Lecture 1: Dynamic Programming

Lecture 1: Dynamic Programming Lecture 1: Dynamic Programming Fatih Guvenen November 2, 2016 Fatih Guvenen Lecture 1: Dynamic Programming November 2, 2016 1 / 32 Goal Solve V (k, z) =max c,k 0 u(c)+ E(V (k 0, z 0 ) z) c + k 0 =(1 +

More information

Heterogeneous Agent Models: I

Heterogeneous Agent Models: I Heterogeneous Agent Models: I Mark Huggett 2 2 Georgetown September, 2017 Introduction Early heterogeneous-agent models integrated the income-fluctuation problem into general equilibrium models. A key

More information

DYNAMIC LECTURE 5: DISCRETE TIME INTERTEMPORAL OPTIMIZATION

DYNAMIC LECTURE 5: DISCRETE TIME INTERTEMPORAL OPTIMIZATION DYNAMIC LECTURE 5: DISCRETE TIME INTERTEMPORAL OPTIMIZATION UNIVERSITY OF MARYLAND: ECON 600. Alternative Methods of Discrete Time Intertemporal Optimization We will start by solving a discrete time intertemporal

More information

Value Function Iteration

Value Function Iteration Value Function Iteration (Lectures on Solution Methods for Economists II) Jesús Fernández-Villaverde 1 and Pablo Guerrón 2 February 26, 2018 1 University of Pennsylvania 2 Boston College Theoretical Background

More information

Dynamic Programming with Hermite Interpolation

Dynamic Programming with Hermite Interpolation Dynamic Programming with Hermite Interpolation Yongyang Cai Hoover Institution, 424 Galvez Mall, Stanford University, Stanford, CA, 94305 Kenneth L. Judd Hoover Institution, 424 Galvez Mall, Stanford University,

More information

Exam in TMA4215 December 7th 2012

Exam in TMA4215 December 7th 2012 Norwegian University of Science and Technology Department of Mathematical Sciences Page of 9 Contact during the exam: Elena Celledoni, tlf. 7359354, cell phone 48238584 Exam in TMA425 December 7th 22 Allowed

More information

Integration. Topic: Trapezoidal Rule. Major: General Engineering. Author: Autar Kaw, Charlie Barker.

Integration. Topic: Trapezoidal Rule. Major: General Engineering. Author: Autar Kaw, Charlie Barker. Integration Topic: Trapezoidal Rule Major: General Engineering Author: Autar Kaw, Charlie Barker 1 What is Integration Integration: The process of measuring the area under a function plotted on a graph.

More information

minimize x subject to (x 2)(x 4) u,

minimize x subject to (x 2)(x 4) u, Math 6366/6367: Optimization and Variational Methods Sample Preliminary Exam Questions 1. Suppose that f : [, L] R is a C 2 -function with f () on (, L) and that you have explicit formulae for

More information

COURSE Numerical integration of functions (continuation) 3.3. The Romberg s iterative generation method

COURSE Numerical integration of functions (continuation) 3.3. The Romberg s iterative generation method COURSE 7 3. Numerical integration of functions (continuation) 3.3. The Romberg s iterative generation method The presence of derivatives in the remainder difficulties in applicability to practical problems

More information

Stat 451 Lecture Notes Numerical Integration

Stat 451 Lecture Notes Numerical Integration Stat 451 Lecture Notes 03 12 Numerical Integration Ryan Martin UIC www.math.uic.edu/~rgmartin 1 Based on Chapter 5 in Givens & Hoeting, and Chapters 4 & 18 of Lange 2 Updated: February 11, 2016 1 / 29

More information

Numerical Analysis Solution of Algebraic Equation (non-linear equation) 1- Trial and Error. 2- Fixed point

Numerical Analysis Solution of Algebraic Equation (non-linear equation) 1- Trial and Error. 2- Fixed point Numerical Analysis Solution of Algebraic Equation (non-linear equation) 1- Trial and Error In this method we assume initial value of x, and substitute in the equation. Then modify x and continue till we

More information

Numerical Methods in Economics

Numerical Methods in Economics Numerical Methods in Economics MIT Press, 1998 Chapter 12 Notes Numerical Dynamic Programming Kenneth L. Judd Hoover Institution November 15, 2002 1 Discrete-Time Dynamic Programming Objective: X: set

More information

1: PROBABILITY REVIEW

1: PROBABILITY REVIEW 1: PROBABILITY REVIEW Marek Rutkowski School of Mathematics and Statistics University of Sydney Semester 2, 2016 M. Rutkowski (USydney) Slides 1: Probability Review 1 / 56 Outline We will review the following

More information

Higher-order ordinary differential equations

Higher-order ordinary differential equations Higher-order ordinary differential equations 1 A linear ODE of general order n has the form a n (x) dn y dx n +a n 1(x) dn 1 y dx n 1 + +a 1(x) dy dx +a 0(x)y = f(x). If f(x) = 0 then the equation is called

More information

0.1 Tauchen s method to approximate a continuous income process

0.1 Tauchen s method to approximate a continuous income process University of Minnesota 8107 Macroeconomic Theory, Spring 2008, Mini2 Fabrizio Perri Notes on numerical methods for income fluctuations problems 0.1 Tauchen s method to approximate a continuous income

More information

Lecture 10 Polynomial interpolation

Lecture 10 Polynomial interpolation Lecture 10 Polynomial interpolation Weinan E 1,2 and Tiejun Li 2 1 Department of Mathematics, Princeton University, weinan@princeton.edu 2 School of Mathematical Sciences, Peking University, tieli@pku.edu.cn

More information

12.0 Properties of orthogonal polynomials

12.0 Properties of orthogonal polynomials 12.0 Properties of orthogonal polynomials In this section we study orthogonal polynomials to use them for the construction of quadrature formulas investigate projections on polynomial spaces and their

More information

NUMERICAL DYNAMIC PROGRAMMING

NUMERICAL DYNAMIC PROGRAMMING NUMERICAL DYNAMIC PROGRAMMING Kenneth L. Judd Hoover Institution and NBER July 29, 2008 1 Dynamic Programming Foundation of dynamic economic modelling Individual decisionmaking Social planners problems,

More information

Introduction to Probability and Stocastic Processes - Part I

Introduction to Probability and Stocastic Processes - Part I Introduction to Probability and Stocastic Processes - Part I Lecture 2 Henrik Vie Christensen vie@control.auc.dk Department of Control Engineering Institute of Electronic Systems Aalborg University Denmark

More information

Iowa State University. Instructor: Alex Roitershtein Summer Homework #5. Solutions

Iowa State University. Instructor: Alex Roitershtein Summer Homework #5. Solutions Math 50 Iowa State University Introduction to Real Analysis Department of Mathematics Instructor: Alex Roitershtein Summer 205 Homework #5 Solutions. Let α and c be real numbers, c > 0, and f is defined

More information

f (x) = k=0 f (0) = k=0 k=0 a k k(0) k 1 = a 1 a 1 = f (0). a k k(k 1)x k 2, k=2 a k k(k 1)(0) k 2 = 2a 2 a 2 = f (0) 2 a k k(k 1)(k 2)x k 3, k=3

f (x) = k=0 f (0) = k=0 k=0 a k k(0) k 1 = a 1 a 1 = f (0). a k k(k 1)x k 2, k=2 a k k(k 1)(0) k 2 = 2a 2 a 2 = f (0) 2 a k k(k 1)(k 2)x k 3, k=3 1 M 13-Lecture Contents: 1) Taylor Polynomials 2) Taylor Series Centered at x a 3) Applications of Taylor Polynomials Taylor Series The previous section served as motivation and gave some useful expansion.

More information

Slides II - Dynamic Programming

Slides II - Dynamic Programming Slides II - Dynamic Programming Julio Garín University of Georgia Macroeconomic Theory II (Ph.D.) Spring 2017 Macroeconomic Theory II Slides II - Dynamic Programming Spring 2017 1 / 32 Outline 1. Lagrangian

More information

Integration, differentiation, and root finding. Phys 420/580 Lecture 7

Integration, differentiation, and root finding. Phys 420/580 Lecture 7 Integration, differentiation, and root finding Phys 420/580 Lecture 7 Numerical integration Compute an approximation to the definite integral I = b Find area under the curve in the interval Trapezoid Rule:

More information

Exam in Numerical Methods (MA2501)

Exam in Numerical Methods (MA2501) Norwegian University of Science and Technology Department of Mathematical Sciences Page 1 of 7 MA251 Numeriske Metoder Olivier Verdier (contact: 48 95 2 66) Exam in Numerical Methods (MA251) 211-5-25,

More information

MA2501 Numerical Methods Spring 2015

MA2501 Numerical Methods Spring 2015 Norwegian University of Science and Technology Department of Mathematics MA5 Numerical Methods Spring 5 Solutions to exercise set 9 Find approximate values of the following integrals using the adaptive

More information

Lecture 25: Review. Statistics 104. April 23, Colin Rundel

Lecture 25: Review. Statistics 104. April 23, Colin Rundel Lecture 25: Review Statistics 104 Colin Rundel April 23, 2012 Joint CDF F (x, y) = P [X x, Y y] = P [(X, Y ) lies south-west of the point (x, y)] Y (x,y) X Statistics 104 (Colin Rundel) Lecture 25 April

More information

1 Numerical Methods to Solve the Income-Fluctuation Problem

1 Numerical Methods to Solve the Income-Fluctuation Problem 1 Numerical Methods to Solve the Income-Fluctuation Problem 1.1 Tauchen s method to approximate a continuous income process In many of the problems we studied in the previous chapter, we postulated that

More information

COURSE Numerical integration of functions

COURSE Numerical integration of functions COURSE 6 3. Numerical integration of functions The need: for evaluating definite integrals of functions that has no explicit antiderivatives or whose antiderivatives are not easy to obtain. Let f : [a,

More information

Lecture 19: Solving linear ODEs + separable techniques for nonlinear ODE s

Lecture 19: Solving linear ODEs + separable techniques for nonlinear ODE s Lecture 19: Solving linear ODEs + separable techniques for nonlinear ODE s Geoffrey Cowles Department of Fisheries Oceanography School for Marine Science and Technology University of Massachusetts-Dartmouth

More information

Lecture Note 3: Interpolation and Polynomial Approximation. Xiaoqun Zhang Shanghai Jiao Tong University

Lecture Note 3: Interpolation and Polynomial Approximation. Xiaoqun Zhang Shanghai Jiao Tong University Lecture Note 3: Interpolation and Polynomial Approximation Xiaoqun Zhang Shanghai Jiao Tong University Last updated: October 10, 2015 2 Contents 1.1 Introduction................................ 3 1.1.1

More information

Numerical Methods in Physics and Astrophysics

Numerical Methods in Physics and Astrophysics Kostas Kokkotas 2 October 17, 2017 2 http://www.tat.physik.uni-tuebingen.de/ kokkotas Kostas Kokkotas 3 TOPICS 1. Solving nonlinear equations 2. Solving linear systems of equations 3. Interpolation, approximation

More information

Neoclassical Growth Model: I

Neoclassical Growth Model: I Neoclassical Growth Model: I Mark Huggett 2 2 Georgetown October, 2017 Growth Model: Introduction Neoclassical Growth Model is the workhorse model in macroeconomics. It comes in two main varieties: infinitely-lived

More information

We consider the problem of finding a polynomial that interpolates a given set of values:

We consider the problem of finding a polynomial that interpolates a given set of values: Chapter 5 Interpolation 5. Polynomial Interpolation We consider the problem of finding a polynomial that interpolates a given set of values: x x 0 x... x n y y 0 y... y n where the x i are all distinct.

More information

Kernel Density Estimation

Kernel Density Estimation EECS 598: Statistical Learning Theory, Winter 2014 Topic 19 Kernel Density Estimation Lecturer: Clayton Scott Scribe: Yun Wei, Yanzhen Deng Disclaimer: These notes have not been subjected to the usual

More information

1 Basic Analysis of Forward-Looking Decision Making

1 Basic Analysis of Forward-Looking Decision Making 1 Basic Analysis of Forward-Looking Decision Making Individuals and families make the key decisions that determine the future of the economy. The decisions involve balancing current sacrifice against future

More information

A nonlinear equation is any equation of the form. f(x) = 0. A nonlinear equation can have any number of solutions (finite, countable, uncountable)

A nonlinear equation is any equation of the form. f(x) = 0. A nonlinear equation can have any number of solutions (finite, countable, uncountable) Nonlinear equations Definition A nonlinear equation is any equation of the form where f is a nonlinear function. Nonlinear equations x 2 + x + 1 = 0 (f : R R) f(x) = 0 (x cos y, 2y sin x) = (0, 0) (f :

More information

Lecture XI. Approximating the Invariant Distribution

Lecture XI. Approximating the Invariant Distribution Lecture XI Approximating the Invariant Distribution Gianluca Violante New York University Quantitative Macroeconomics G. Violante, Invariant Distribution p. 1 /24 SS Equilibrium in the Aiyagari model G.

More information

Value and Policy Iteration

Value and Policy Iteration Chapter 7 Value and Policy Iteration 1 For infinite horizon problems, we need to replace our basic computational tool, the DP algorithm, which we used to compute the optimal cost and policy for finite

More information

Projection Methods. Felix Kubler 1. October 10, DBF, University of Zurich and Swiss Finance Institute

Projection Methods. Felix Kubler 1. October 10, DBF, University of Zurich and Swiss Finance Institute Projection Methods Felix Kubler 1 1 DBF, University of Zurich and Swiss Finance Institute October 10, 2017 Felix Kubler Comp.Econ. Gerzensee, Ch5 October 10, 2017 1 / 55 Motivation In many dynamic economic

More information

conditional cdf, conditional pdf, total probability theorem?

conditional cdf, conditional pdf, total probability theorem? 6 Multiple Random Variables 6.0 INTRODUCTION scalar vs. random variable cdf, pdf transformation of a random variable conditional cdf, conditional pdf, total probability theorem expectation of a random

More information

Preliminary Examination, Numerical Analysis, August 2016

Preliminary Examination, Numerical Analysis, August 2016 Preliminary Examination, Numerical Analysis, August 2016 Instructions: This exam is closed books and notes. The time allowed is three hours and you need to work on any three out of questions 1-4 and any

More information

CS 323: Numerical Analysis and Computing

CS 323: Numerical Analysis and Computing CS 323: Numerical Analysis and Computing MIDTERM #2 Instructions: This is an open notes exam, i.e., you are allowed to consult any textbook, your class notes, homeworks, or any of the handouts from us.

More information

Monte Carlo Methods for Statistical Inference: Variance Reduction Techniques

Monte Carlo Methods for Statistical Inference: Variance Reduction Techniques Monte Carlo Methods for Statistical Inference: Variance Reduction Techniques Hung Chen hchen@math.ntu.edu.tw Department of Mathematics National Taiwan University 3rd March 2004 Meet at NS 104 On Wednesday

More information

Chapter 3: Root Finding. September 26, 2005

Chapter 3: Root Finding. September 26, 2005 Chapter 3: Root Finding September 26, 2005 Outline 1 Root Finding 2 3.1 The Bisection Method 3 3.2 Newton s Method: Derivation and Examples 4 3.3 How To Stop Newton s Method 5 3.4 Application: Division

More information

ECOM 009 Macroeconomics B. Lecture 2

ECOM 009 Macroeconomics B. Lecture 2 ECOM 009 Macroeconomics B Lecture 2 Giulio Fella c Giulio Fella, 2014 ECOM 009 Macroeconomics B - Lecture 2 40/197 Aim of consumption theory Consumption theory aims at explaining consumption/saving decisions

More information

1 Solution to Problem 2.1

1 Solution to Problem 2.1 Solution to Problem 2. I incorrectly worked this exercise instead of 2.2, so I decided to include the solution anyway. a) We have X Y /3, which is a - function. It maps the interval, ) where X lives) onto

More information

Linear Models in Machine Learning

Linear Models in Machine Learning CS540 Intro to AI Linear Models in Machine Learning Lecturer: Xiaojin Zhu jerryzhu@cs.wisc.edu We briefly go over two linear models frequently used in machine learning: linear regression for, well, regression,

More information

A simple macro dynamic model with endogenous saving rate: the representative agent model

A simple macro dynamic model with endogenous saving rate: the representative agent model A simple macro dynamic model with endogenous saving rate: the representative agent model Virginia Sánchez-Marcos Macroeconomics, MIE-UNICAN Macroeconomics (MIE-UNICAN) A simple macro dynamic model with

More information

Ambiguous Business Cycles: Online Appendix

Ambiguous Business Cycles: Online Appendix Ambiguous Business Cycles: Online Appendix By Cosmin Ilut and Martin Schneider This paper studies a New Keynesian business cycle model with agents who are averse to ambiguity (Knightian uncertainty). Shocks

More information

Interpolation. 1. Judd, K. Numerical Methods in Economics, Cambridge: MIT Press. Chapter

Interpolation. 1. Judd, K. Numerical Methods in Economics, Cambridge: MIT Press. Chapter Key References: Interpolation 1. Judd, K. Numerical Methods in Economics, Cambridge: MIT Press. Chapter 6. 2. Press, W. et. al. Numerical Recipes in C, Cambridge: Cambridge University Press. Chapter 3

More information

Basic Deterministic Dynamic Programming

Basic Deterministic Dynamic Programming Basic Deterministic Dynamic Programming Timothy Kam School of Economics & CAMA Australian National University ECON8022, This version March 17, 2008 Motivation What do we do? Outline Deterministic IHDP

More information

Numerical Methods in Physics and Astrophysics

Numerical Methods in Physics and Astrophysics Kostas Kokkotas 2 October 20, 2014 2 http://www.tat.physik.uni-tuebingen.de/ kokkotas Kostas Kokkotas 3 TOPICS 1. Solving nonlinear equations 2. Solving linear systems of equations 3. Interpolation, approximation

More information

Machine Learning Lecture 7

Machine Learning Lecture 7 Course Outline Machine Learning Lecture 7 Fundamentals (2 weeks) Bayes Decision Theory Probability Density Estimation Statistical Learning Theory 23.05.2016 Discriminative Approaches (5 weeks) Linear Discriminant

More information

CS 323: Numerical Analysis and Computing

CS 323: Numerical Analysis and Computing CS 323: Numerical Analysis and Computing MIDTERM #2 Instructions: This is an open notes exam, i.e., you are allowed to consult any textbook, your class notes, homeworks, or any of the handouts from us.

More information

Mathematical Economics: Lecture 2

Mathematical Economics: Lecture 2 Mathematical Economics: Lecture 2 Yu Ren WISE, Xiamen University September 25, 2012 Outline 1 Number Line The number line, origin (Figure 2.1 Page 11) Number Line Interval (a, b) = {x R 1 : a < x < b}

More information

Calculus of Variations Summer Term 2014

Calculus of Variations Summer Term 2014 Calculus of Variations Summer Term 2014 Lecture 12 26. Juni 2014 c Daria Apushkinskaya 2014 () Calculus of variations lecture 12 26. Juni 2014 1 / 25 Purpose of Lesson Purpose of Lesson: To discuss numerical

More information

Numerical integration and differentiation. Unit IV. Numerical Integration and Differentiation. Plan of attack. Numerical integration.

Numerical integration and differentiation. Unit IV. Numerical Integration and Differentiation. Plan of attack. Numerical integration. Unit IV Numerical Integration and Differentiation Numerical integration and differentiation quadrature classical formulas for equally spaced nodes improper integrals Gaussian quadrature and orthogonal

More information

MATH 56A SPRING 2008 STOCHASTIC PROCESSES

MATH 56A SPRING 2008 STOCHASTIC PROCESSES MATH 56A SPRING 008 STOCHASTIC PROCESSES KIYOSHI IGUSA Contents 4. Optimal Stopping Time 95 4.1. Definitions 95 4.. The basic problem 95 4.3. Solutions to basic problem 97 4.4. Cost functions 101 4.5.

More information

Announcements. Topics: Homework:

Announcements. Topics: Homework: Announcements Topics: - sections 7.5 (additional techniques of integration), 7.6 (applications of integration), * Read these sections and study solved examples in your textbook! Homework: - review lecture

More information

AM 205: lecture 19. Last time: Conditions for optimality Today: Newton s method for optimization, survey of optimization methods

AM 205: lecture 19. Last time: Conditions for optimality Today: Newton s method for optimization, survey of optimization methods AM 205: lecture 19 Last time: Conditions for optimality Today: Newton s method for optimization, survey of optimization methods Optimality Conditions: Equality Constrained Case As another example of equality

More information

Lecture 4 The Centralized Economy: Extensions

Lecture 4 The Centralized Economy: Extensions Lecture 4 The Centralized Economy: Extensions Leopold von Thadden University of Mainz and ECB (on leave) Advanced Macroeconomics, Winter Term 2013 1 / 36 I Motivation This Lecture considers some applications

More information

Examination paper for TMA4215 Numerical Mathematics

Examination paper for TMA4215 Numerical Mathematics Department of Mathematical Sciences Examination paper for TMA425 Numerical Mathematics Academic contact during examination: Trond Kvamsdal Phone: 93058702 Examination date: 6th of December 207 Examination

More information

Lecture Note 3: Polynomial Interpolation. Xiaoqun Zhang Shanghai Jiao Tong University

Lecture Note 3: Polynomial Interpolation. Xiaoqun Zhang Shanghai Jiao Tong University Lecture Note 3: Polynomial Interpolation Xiaoqun Zhang Shanghai Jiao Tong University Last updated: October 24, 2013 1.1 Introduction We first look at some examples. Lookup table for f(x) = 2 π x 0 e x2

More information

Final Exam May 4, 2016

Final Exam May 4, 2016 1 Math 425 / AMCS 525 Dr. DeTurck Final Exam May 4, 2016 You may use your book and notes on this exam. Show your work in the exam book. Work only the problems that correspond to the section that you prepared.

More information

LECTURE NOTES ELEMENTARY NUMERICAL METHODS. Eusebius Doedel

LECTURE NOTES ELEMENTARY NUMERICAL METHODS. Eusebius Doedel LECTURE NOTES on ELEMENTARY NUMERICAL METHODS Eusebius Doedel TABLE OF CONTENTS Vector and Matrix Norms 1 Banach Lemma 20 The Numerical Solution of Linear Systems 25 Gauss Elimination 25 Operation Count

More information

Uncertainty Quantification in Computational Science

Uncertainty Quantification in Computational Science DTU 2010 - Lecture I Uncertainty Quantification in Computational Science Jan S Hesthaven Brown University Jan.Hesthaven@Brown.edu Objective of lectures The main objective of these lectures are To offer

More information

Interpolation Theory

Interpolation Theory Numerical Analysis Massoud Malek Interpolation Theory The concept of interpolation is to select a function P (x) from a given class of functions in such a way that the graph of y P (x) passes through the

More information

CSCI-6971 Lecture Notes: Probability theory

CSCI-6971 Lecture Notes: Probability theory CSCI-6971 Lecture Notes: Probability theory Kristopher R. Beevers Department of Computer Science Rensselaer Polytechnic Institute beevek@cs.rpi.edu January 31, 2006 1 Properties of probabilities Let, A,

More information

5 Operations on Multiple Random Variables

5 Operations on Multiple Random Variables EE360 Random Signal analysis Chapter 5: Operations on Multiple Random Variables 5 Operations on Multiple Random Variables Expected value of a function of r.v. s Two r.v. s: ḡ = E[g(X, Y )] = g(x, y)f X,Y

More information

Lecture 11. Probability Theory: an Overveiw

Lecture 11. Probability Theory: an Overveiw Math 408 - Mathematical Statistics Lecture 11. Probability Theory: an Overveiw February 11, 2013 Konstantin Zuev (USC) Math 408, Lecture 11 February 11, 2013 1 / 24 The starting point in developing the

More information

4 Stability analysis of finite-difference methods for ODEs

4 Stability analysis of finite-difference methods for ODEs MATH 337, by T. Lakoba, University of Vermont 36 4 Stability analysis of finite-difference methods for ODEs 4.1 Consistency, stability, and convergence of a numerical method; Main Theorem In this Lecture

More information

OR MSc Maths Revision Course

OR MSc Maths Revision Course OR MSc Maths Revision Course Tom Byrne School of Mathematics University of Edinburgh t.m.byrne@sms.ed.ac.uk 15 September 2017 General Information Today JCMB Lecture Theatre A, 09:30-12:30 Mathematics revision

More information

Kasetsart University Workshop. Multigrid methods: An introduction

Kasetsart University Workshop. Multigrid methods: An introduction Kasetsart University Workshop Multigrid methods: An introduction Dr. Anand Pardhanani Mathematics Department Earlham College Richmond, Indiana USA pardhan@earlham.edu A copy of these slides is available

More information

Jim Lambers MAT 460/560 Fall Semester Practice Final Exam

Jim Lambers MAT 460/560 Fall Semester Practice Final Exam Jim Lambers MAT 460/560 Fall Semester 2009-10 Practice Final Exam 1. Let f(x) = sin 2x + cos 2x. (a) Write down the 2nd Taylor polynomial P 2 (x) of f(x) centered around x 0 = 0. (b) Write down the corresponding

More information

March Algebra 2 Question 1. March Algebra 2 Question 1

March Algebra 2 Question 1. March Algebra 2 Question 1 March Algebra 2 Question 1 If the statement is always true for the domain, assign that part a 3. If it is sometimes true, assign it a 2. If it is never true, assign it a 1. Your answer for this question

More information

x log x, which is strictly convex, and use Jensen s Inequality:

x log x, which is strictly convex, and use Jensen s Inequality: 2. Information measures: mutual information 2.1 Divergence: main inequality Theorem 2.1 (Information Inequality). D(P Q) 0 ; D(P Q) = 0 iff P = Q Proof. Let ϕ(x) x log x, which is strictly convex, and

More information

Method of Finite Elements I

Method of Finite Elements I Method of Finite Elements I PhD Candidate - Charilaos Mylonas HIL H33.1 and Boundary Conditions, 26 March, 2018 Institute of Structural Engineering Method of Finite Elements I 1 Outline 1 2 Penalty method

More information

Dynamic Macroeconomics

Dynamic Macroeconomics Dynamic Macroeconomics Summer Term 2011 Christian Bayer Universität Bonn May 15, 2011 Overview of the Course: My Part 1. Introduction, Basic Theory 2. Stochastic Fluctuations 3. Numerical Tools 4. Estimation

More information

Week 9 Generators, duality, change of measure

Week 9 Generators, duality, change of measure Week 9 Generators, duality, change of measure Jonathan Goodman November 18, 013 1 Generators This section describes a common abstract way to describe many of the differential equations related to Markov

More information

PROJECTION METHODS FOR DYNAMIC MODELS

PROJECTION METHODS FOR DYNAMIC MODELS PROJECTION METHODS FOR DYNAMIC MODELS Kenneth L. Judd Hoover Institution and NBER June 28, 2006 Functional Problems Many problems involve solving for some unknown function Dynamic programming Consumption

More information

(f(x) P 3 (x)) dx. (a) The Lagrange formula for the error is given by

(f(x) P 3 (x)) dx. (a) The Lagrange formula for the error is given by 1. QUESTION (a) Given a nth degree Taylor polynomial P n (x) of a function f(x), expanded about x = x 0, write down the Lagrange formula for the truncation error, carefully defining all its elements. How

More information

CHAPTER 5: Linear Multistep Methods

CHAPTER 5: Linear Multistep Methods CHAPTER 5: Linear Multistep Methods Multistep: use information from many steps Higher order possible with fewer function evaluations than with RK. Convenient error estimates. Changing stepsize or order

More information

Differentiation and Integration

Differentiation and Integration Differentiation and Integration (Lectures on Numerical Analysis for Economists II) Jesús Fernández-Villaverde 1 and Pablo Guerrón 2 February 12, 2018 1 University of Pennsylvania 2 Boston College Motivation

More information

Stochastic Dynamic Programming: The One Sector Growth Model

Stochastic Dynamic Programming: The One Sector Growth Model Stochastic Dynamic Programming: The One Sector Growth Model Esteban Rossi-Hansberg Princeton University March 26, 2012 Esteban Rossi-Hansberg () Stochastic Dynamic Programming March 26, 2012 1 / 31 References

More information

1 Markov decision processes

1 Markov decision processes 2.997 Decision-Making in Large-Scale Systems February 4 MI, Spring 2004 Handout #1 Lecture Note 1 1 Markov decision processes In this class we will study discrete-time stochastic systems. We can describe

More information

Numerical Dynamic Programming

Numerical Dynamic Programming Numerical Dynamic Programming Kenneth L. Judd Hoover Institution Prepared for ICE05 July 20, 2005 - lectures 1 Discrete-Time Dynamic Programming Objective: X: set of states E ( TX D: the set of controls

More information

Asymptotic distribution of the sample average value-at-risk

Asymptotic distribution of the sample average value-at-risk Asymptotic distribution of the sample average value-at-risk Stoyan V. Stoyanov Svetlozar T. Rachev September 3, 7 Abstract In this paper, we prove a result for the asymptotic distribution of the sample

More information

Q 0 x if x 0 x x 1. S 1 x if x 1 x x 2. i 0,1,...,n 1, and L x L n 1 x if x n 1 x x n

Q 0 x if x 0 x x 1. S 1 x if x 1 x x 2. i 0,1,...,n 1, and L x L n 1 x if x n 1 x x n . - Piecewise Linear-Quadratic Interpolation Piecewise-polynomial Approximation: Problem: Givenn pairs of data points x i, y i, i,,...,n, find a piecewise-polynomial Sx S x if x x x Sx S x if x x x 2 :

More information

Modeling Data with Linear Combinations of Basis Functions. Read Chapter 3 in the text by Bishop

Modeling Data with Linear Combinations of Basis Functions. Read Chapter 3 in the text by Bishop Modeling Data with Linear Combinations of Basis Functions Read Chapter 3 in the text by Bishop A Type of Supervised Learning Problem We want to model data (x 1, t 1 ),..., (x N, t N ), where x i is a vector

More information

Functions of Random Variables

Functions of Random Variables Functions of Random Variables Statistics 110 Summer 2006 Copyright c 2006 by Mark E. Irwin Functions of Random Variables As we ve seen before, if X N(µ, σ 2 ), then Y = ax + b is also normally distributed.

More information

Some definitions. Math 1080: Numerical Linear Algebra Chapter 5, Solving Ax = b by Optimization. A-inner product. Important facts

Some definitions. Math 1080: Numerical Linear Algebra Chapter 5, Solving Ax = b by Optimization. A-inner product. Important facts Some definitions Math 1080: Numerical Linear Algebra Chapter 5, Solving Ax = b by Optimization M. M. Sussman sussmanm@math.pitt.edu Office Hours: MW 1:45PM-2:45PM, Thack 622 A matrix A is SPD (Symmetric

More information

First-Order Differential Equations

First-Order Differential Equations CHAPTER 1 First-Order Differential Equations 1. Diff Eqns and Math Models Know what it means for a function to be a solution to a differential equation. In order to figure out if y = y(x) is a solution

More information

Numerical Analysis Preliminary Exam 10 am to 1 pm, August 20, 2018

Numerical Analysis Preliminary Exam 10 am to 1 pm, August 20, 2018 Numerical Analysis Preliminary Exam 1 am to 1 pm, August 2, 218 Instructions. You have three hours to complete this exam. Submit solutions to four (and no more) of the following six problems. Please start

More information

Lecture 2: Convex Sets and Functions

Lecture 2: Convex Sets and Functions Lecture 2: Convex Sets and Functions Hyang-Won Lee Dept. of Internet & Multimedia Eng. Konkuk University Lecture 2 Network Optimization, Fall 2015 1 / 22 Optimization Problems Optimization problems are

More information