AR(p) + MA(q) = ARMA(p, q)



Similar documents
Introduction to Time Series Analysis. Lecture 6.

Time Series Analysis

Lecture 4: Seasonal Time Series, Trend Analysis & Component Model Bus 41910, Time Series Analysis, Mr. R. Tsay

ITSM-R Reference Manual

Lecture 2: ARMA(p,q) models (part 3)

Discrete Time Series Analysis with ARMA Models

Some useful concepts in univariate time series analysis

Sales forecasting # 2

Financial TIme Series Analysis: Part II

Time Series Analysis 1. Lecture 8: Time Series Analysis. Time Series Analysis MIT 18.S096. Dr. Kempthorne. Fall 2013 MIT 18.S096

Analysis and Computation for Finance Time Series - An Introduction

Luciano Rispoli Department of Economics, Mathematics and Statistics Birkbeck College (University of London)

Univariate and Multivariate Methods PEARSON. Addison Wesley

Analysis of algorithms of time series analysis for forecasting sales

Univariate Time Series Analysis; ARIMA Models

Graphical Tools for Exploring and Analyzing Data From ARIMA Time Series Models

Time Series - ARIMA Models. Instructor: G. William Schwert

THE SVM APPROACH FOR BOX JENKINS MODELS

Promotional Analysis and Forecasting for Demand Planning: A Practical Time Series Approach Michael Leonard, SAS Institute Inc.

Univariate Time Series Analysis; ARIMA Models

Time Series Analysis

1 Short Introduction to Time Series

Time Series Analysis

Time Series Analysis

Algebraic Concepts Algebraic Concepts Writing

Time Series Analysis

(Refer Slide Time: 1:42)

Solutions to Homework 6 Mathematics 503 Foundations of Mathematics Spring 2014

A FULLY INTEGRATED ENVIRONMENT FOR TIME-DEPENDENT DATA ANALYSIS

Determine If An Equation Represents a Function

Do Electricity Prices Reflect Economic Fundamentals?: Evidence from the California ISO

Factoring A Quadratic Polynomial

Traffic Safety Facts. Research Note. Time Series Analysis and Forecast of Crash Fatalities during Six Holiday Periods Cejun Liu* and Chou-Lin Chen

Lecture 20: Transmission (ABCD) Matrix.

Time Series Analysis in Economics. Klaus Neusser

Forecasting model of electricity demand in the Nordic countries. Tone Pedersen

FURTHER VECTORS (MEI)

Software Review: ITSM 2000 Professional Version 6.0.

Time Series Analysis of Aviation Data

TIME SERIES. Syllabus... Keywords...

Time Series Analysis III

Statistics and Probability Letters

Lines and Planes in R 3

DIRECT ORBITAL DYNAMICS: USING INDEPENDENT ORBITAL TERMS TO TREAT BODIES AS ORBITING EACH OTHER DIRECTLY WHILE IN MOTION


the points are called control points approximating curve

Annual Report H I G H E R E D U C AT I O N C O M M I S S I O N - PA K I S TA N

Time Series Analysis and Forecasting

Section 1.1. Introduction to R n

Preliminary Mathematics

Visualizing Differential Equations Slope Fields. by Lin McMullin

In this paper we study how the time-series structure of the demand process affects the value of information

3.1 Stationary Processes and Mean Reversion

JUST THE MATHS UNIT NUMBER 8.5. VECTORS 5 (Vector equations of straight lines) A.J.Hobson

15.1 Factoring Polynomials

Solving Compound Interest Problems

Chapter 9: Univariate Time Series Analysis

Correlational Research

Time Series Analysis

PYTHAGOREAN TRIPLES KEITH CONRAD

Section 9.5: Equations of Lines and Planes

4.4 Transforming Circles

TIME SERIES ANALYSIS: AN OVERVIEW

Welcome! 8 th Grade Honors Algebra Informational Meeting

SOLVING EQUATIONS WITH EXCEL

Agenda. Managing Uncertainty in the Supply Chain. The Economic Order Quantity. Classic inventory theory

Time Series in Mathematical Finance

Time Series HILARY TERM 2010 PROF. GESINE REINERT

Copyrighted Material. Chapter 1 DEGREE OF A CURVE

Newton s Law of Motion

Predicting Indian GDP. And its relation with FMCG Sales

C Programming Language

TIME SERIES ANALYSIS

Background Knowledge

Welcome to Math Video Lessons. Stanley Ocken. Department of Mathematics The City College of New York Fall 2013

COURSE SYLLABUS Pre-Calculus A/B Last Modified: April 2015

1.5. Factorisation. Introduction. Prerequisites. Learning Outcomes. Learning Style

Forecasting the US Dollar / Euro Exchange rate Using ARMA Models

Year 9 set 1 Mathematics notes, to accompany the 9H book.

State Space Time Series Analysis

Estimating an ARMA Process

Simple Linear Regression

Advanced Forecasting Techniques and Models: ARIMA

Readers will be provided a link to download the software and Excel files that are used in the book after payment. Please visit

north seattle community college

Survey of the Mathematics of Big Data

Moon Phases & Eclipses Notes

Matrices and Polynomials

Prentice Hall Mathematics: Algebra Correlated to: Utah Core Curriculum for Math, Intermediate Algebra (Secondary)

= y y 0. = z z 0. (a) Find a parametric vector equation for L. (b) Find parametric (scalar) equations for L.

ARMA, GARCH and Related Option Pricing Method

4.1. Title: data analysis (systems analysis) Annotation of educational discipline: educational discipline includes in itself the mastery of the

Using the ac Method to Factor

Inversion. Chapter Constructing The Inverse of a Point: If P is inside the circle of inversion: (See Figure 7.1)

Core Maths C2. Revision Notes

0.8 Rational Expressions and Equations

CHAPTER FIVE. 5. Equations of Lines in R 3

Zeros of Polynomial Functions

Transcription:

AR(p) + MA(q) = ARMA(p, q)

Outline 1 3.4: ARMA(p, q) Model 2 Homework 3a Arthur Berg AR(p) + MA(q) = ARMA(p, q) 2/ 12

ARMA(p, q) Model Definition (ARMA(p, q) Model) A time series is ARMA(p, q) if it is stationary and satisfies Z t = α + φ 1 Z t 1 + + φ p Z t p + θ 1 a t 1 + + θ q a t q +a t ( ) } {{ } } {{ } AR part MA part Using the AR and MA operators, we can rewrite ( ) as φ(b)ż t = θ(b)a t The textbook defines the ARMA process with a slightly different parametrization given by Z t = α + φ 1 Z t 1 + + φ p Z t p θ 1 a t 1 θ q a t q + a t ( ) Arthur Berg AR(p) + MA(q) = ARMA(p, q) 3/ 12

ARMA(p, q) Model Definition (ARMA(p, q) Model) A time series is ARMA(p, q) if it is stationary and satisfies Z t = α + φ 1 Z t 1 + + φ p Z t p + θ 1 a t 1 + + θ q a t q +a t ( ) } {{ } } {{ } AR part MA part Using the AR and MA operators, we can rewrite ( ) as φ(b)ż t = θ(b)a t The textbook defines the ARMA process with a slightly different parametrization given by Z t = α + φ 1 Z t 1 + + φ p Z t p θ 1 a t 1 θ q a t q + a t ( ) Arthur Berg AR(p) + MA(q) = ARMA(p, q) 3/ 12

ARMA(p, q) Model Definition (ARMA(p, q) Model) A time series is ARMA(p, q) if it is stationary and satisfies Z t = α + φ 1 Z t 1 + + φ p Z t p + θ 1 a t 1 + + θ q a t q +a t ( ) } {{ } } {{ } AR part MA part Using the AR and MA operators, we can rewrite ( ) as φ(b)ż t = θ(b)a t The textbook defines the ARMA process with a slightly different parametrization given by Z t = α + φ 1 Z t 1 + + φ p Z t p θ 1 a t 1 θ q a t q + a t ( ) Arthur Berg AR(p) + MA(q) = ARMA(p, q) 3/ 12

Causality and Invertibility of ARMA(p, q) Facts: The ARMA(p, q) model given by φ(b)z t = θ(b)a t is causal if and only if φ(z) 0 when z 1 i.e. all roots including complex roots of φ(z) lie outside the unit disk. The ARMA(p, q) model given by φ(b)z t = θ(b)a t is invertible if and only if θ(z) 0 when θ 1 i.e. all roots including complex roots of θ(z) lie outside the unit disk. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 4/ 12

Causality and Invertibility of ARMA(p, q) Facts: The ARMA(p, q) model given by φ(b)z t = θ(b)a t is causal if and only if φ(z) 0 when z 1 i.e. all roots including complex roots of φ(z) lie outside the unit disk. The ARMA(p, q) model given by φ(b)z t = θ(b)a t is invertible if and only if θ(z) 0 when θ 1 i.e. all roots including complex roots of θ(z) lie outside the unit disk. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 4/ 12

Causality and Invertibility of ARMA(p, q) Facts: The ARMA(p, q) model given by φ(b)z t = θ(b)a t is causal if and only if φ(z) 0 when z 1 i.e. all roots including complex roots of φ(z) lie outside the unit disk. The ARMA(p, q) model given by φ(b)z t = θ(b)a t is invertible if and only if θ(z) 0 when θ 1 i.e. all roots including complex roots of θ(z) lie outside the unit disk. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 4/ 12

Parameter Redundancy The process Z t = a t ( ) is equivalent to.5z t 1 =.5a t 1 ( ) which is also equivalent to ( ) ( ), i.e. Z t.5z t 1 = a t.5a t 1 Z t in the last representation is still white noise, but has an ARMA(1,1) representation. We remove redundancies in an ARMA model φ(b)z t = θ(b)a t by canceling the common factors in θ(b) and φ(b). Arthur Berg AR(p) + MA(q) = ARMA(p, q) 5/ 12

Parameter Redundancy The process Z t = a t ( ) is equivalent to.5z t 1 =.5a t 1 ( ) which is also equivalent to ( ) ( ), i.e. Z t.5z t 1 = a t.5a t 1 Z t in the last representation is still white noise, but has an ARMA(1,1) representation. We remove redundancies in an ARMA model φ(b)z t = θ(b)a t by canceling the common factors in θ(b) and φ(b). Arthur Berg AR(p) + MA(q) = ARMA(p, q) 5/ 12

Parameter Redundancy The process Z t = a t ( ) is equivalent to.5z t 1 =.5a t 1 ( ) which is also equivalent to ( ) ( ), i.e. Z t.5z t 1 = a t.5a t 1 Z t in the last representation is still white noise, but has an ARMA(1,1) representation. We remove redundancies in an ARMA model φ(b)z t = θ(b)a t by canceling the common factors in θ(b) and φ(b). Arthur Berg AR(p) + MA(q) = ARMA(p, q) 5/ 12

Parameter Redundancy The process Z t = a t ( ) is equivalent to.5z t 1 =.5a t 1 ( ) which is also equivalent to ( ) ( ), i.e. Z t.5z t 1 = a t.5a t 1 Z t in the last representation is still white noise, but has an ARMA(1,1) representation. We remove redundancies in an ARMA model φ(b)z t = θ(b)a t by canceling the common factors in θ(b) and φ(b). Arthur Berg AR(p) + MA(q) = ARMA(p, q) 5/ 12

Parameter Redundancy The process Z t = a t ( ) is equivalent to.5z t 1 =.5a t 1 ( ) which is also equivalent to ( ) ( ), i.e. Z t.5z t 1 = a t.5a t 1 Z t in the last representation is still white noise, but has an ARMA(1,1) representation. We remove redundancies in an ARMA model φ(b)z t = θ(b)a t by canceling the common factors in θ(b) and φ(b). Arthur Berg AR(p) + MA(q) = ARMA(p, q) 5/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) a t Is this process really ARMA(2,2)? Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) a t Is this process really ARMA(2,2)? Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) a t Is this process really ARMA(2,2)? Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) Is this process really ARMA(2,2)? No! Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) Is this process really ARMA(2,2)? No! Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Removing Redundancy Example Consider the process which is equivalent to Z t =.4Z t 1 +.45Z t 2 + a t + a t 1 +.25a t 2 (1.4B.45B 2 ) Z } {{ } t = (1 + B +.25B 2 ) } {{ } φ(b) θ(b) Is this process really ARMA(2,2)? No! Factoring φ(z) and θ(z) gives φ(z) = 1.4z.45z 2 = (1 +.5z)(1.9z) θ(z) = (1 + z +.25z 2 ) = (1 +.5z) 2 Removing the common term (1 +.5z) gives the reduced model Z t =.9Z t 1 +.5a t 1 + a t a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 6/ 12

Example Causality and Invertibility Example (cont.) The reduced model is φ(b)z t = θ(b)a t where φ(z) = 1.9z θ(z) = 1 +.5z There is only one root of φ(z) which is z = 10/9, and 10/9 lies outside the unit circle so this ARMA model is causal. There is only one root of θ(z) which is z = 2, and 2 lies outside the unit circle so this ARMA model is invertible. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 7/ 12

Example Causality and Invertibility Example (cont.) The reduced model is φ(b)z t = θ(b)a t where φ(z) = 1.9z θ(z) = 1 +.5z There is only one root of φ(z) which is z = 10/9, and 10/9 lies outside the unit circle so this ARMA model is causal. There is only one root of θ(z) which is z = 2, and 2 lies outside the unit circle so this ARMA model is invertible. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 7/ 12

Example Causality and Invertibility Example (cont.) The reduced model is φ(b)z t = θ(b)a t where φ(z) = 1.9z θ(z) = 1 +.5z There is only one root of φ(z) which is z = 10/9, and 10/9 lies outside the unit circle so this ARMA model is causal. There is only one root of θ(z) which is z = 2, and 2 lies outside the unit circle so this ARMA model is invertible. Arthur Berg AR(p) + MA(q) = ARMA(p, q) 7/ 12

ARMA Mathematics Start with the ARMA equation φ(b)z t = θ(b)a t To solve for Z t, we simply divide! That is Z t = θ(b) φ(b) } {{ } ψ(b) Similarly, solving for a t gives a t = φ(b) θ(b) } {{ } π(b) The key is to figuring out the ψ j and π j in ψ(b) = 1 + ψ j B j ψ(b) = 1 + ψ j B j a t Z t MA( ) representation AR( ) representation Arthur Berg AR(p) + MA(q) = ARMA(p, q) 8/ 12

ARMA Mathematics Start with the ARMA equation φ(b)z t = θ(b)a t To solve for Z t, we simply divide! That is Z t = θ(b) φ(b) } {{ } ψ(b) Similarly, solving for a t gives a t = φ(b) θ(b) } {{ } π(b) The key is to figuring out the ψ j and π j in ψ(b) = 1 + ψ j B j ψ(b) = 1 + ψ j B j a t Z t MA( ) representation AR( ) representation Arthur Berg AR(p) + MA(q) = ARMA(p, q) 8/ 12

ARMA Mathematics Start with the ARMA equation φ(b)z t = θ(b)a t To solve for Z t, we simply divide! That is Z t = θ(b) φ(b) } {{ } ψ(b) Similarly, solving for a t gives a t = φ(b) θ(b) } {{ } π(b) The key is to figuring out the ψ j and π j in ψ(b) = 1 + ψ j B j ψ(b) = 1 + ψ j B j a t Z t MA( ) representation AR( ) representation Arthur Berg AR(p) + MA(q) = ARMA(p, q) 8/ 12

ARMA Mathematics Start with the ARMA equation φ(b)z t = θ(b)a t To solve for Z t, we simply divide! That is Z t = θ(b) φ(b) } {{ } ψ(b) Similarly, solving for a t gives a t = φ(b) θ(b) } {{ } π(b) The key is to figuring out the ψ j and π j in ψ(b) = 1 + ψ j B j ψ(b) = 1 + ψ j B j a t Z t MA( ) representation AR( ) representation Arthur Berg AR(p) + MA(q) = ARMA(p, q) 8/ 12

Example MA( ) Representation Example (cont.) Note that ψ(z) = θ(z) φ(z) = 1 +.5z 1.9z = (1 +.5z)(1 +.9z +.9 2 z 2 +.9 3 z 3 + ) for z 1 = 1 + (.5 +.9)z + (.5(.9) +.9 2 )z 2 + (.5(.9 2 ) +.9 3 )z 3 + for z 1 = 1 + (.5 +.9)z + (.5 +.9)(.9)z 2 + (.5 +.9)(.9 2 )z 3 + for z 1 = 1 + (.5 +.9).9 j 1 z j for z 1 } {{ } 1.4 This gives the following MA( ) representation: Z t = θ(b) φ(b) a t = 1 + 1.4.9 j 1 B j a t = a t + 1.4.9 j 1 a t j Arthur Berg AR(p) + MA(q) = ARMA(p, q) 9/ 12

Example MA( ) Representation Example (cont.) Note that ψ(z) = θ(z) φ(z) = 1 +.5z 1.9z = (1 +.5z)(1 +.9z +.9 2 z 2 +.9 3 z 3 + ) for z 1 = 1 + (.5 +.9)z + (.5(.9) +.9 2 )z 2 + (.5(.9 2 ) +.9 3 )z 3 + for z 1 = 1 + (.5 +.9)z + (.5 +.9)(.9)z 2 + (.5 +.9)(.9 2 )z 3 + for z 1 = 1 + (.5 +.9).9 j 1 z j for z 1 } {{ } 1.4 This gives the following MA( ) representation: Z t = θ(b) φ(b) a t = 1 + 1.4.9 j 1 B j a t = a t + 1.4.9 j 1 a t j Arthur Berg AR(p) + MA(q) = ARMA(p, q) 9/ 12

Example AR( ) Representation Example (cont.) Similarly, one can show π(z) = φ(z) θ(z). = 1.9z 1 +.5z = 1 1.4 (.5) j 1 z j for z 1 This gives the following AR( ) representation: Z t = 1.4 (.5) j 1 a t j + a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 10/ 12

Example AR( ) Representation Example (cont.) Similarly, one can show π(z) = φ(z) θ(z). = 1.9z 1 +.5z = 1 1.4 (.5) j 1 z j for z 1 This gives the following AR( ) representation: Z t = 1.4 (.5) j 1 a t j + a t Arthur Berg AR(p) + MA(q) = ARMA(p, q) 10/ 12

Outline 1 3.4: ARMA(p, q) Model 2 Homework 3a Arthur Berg AR(p) + MA(q) = ARMA(p, q) 11/ 12

Homework 3a Read 4.1 and 4.2. Derive the equation for π(z) on slide 10. Do exercise #3.14(a) only for models (i), (ii), and (iv) Do exercise #3.14(b,c) only for model (ii) Arthur Berg AR(p) + MA(q) = ARMA(p, q) 12/ 12