TS_Fall_2011_Lecture_4

# TS_Fall_2011_Lecture_4 - FIN 271 FINANCIAL MODELING AND...

This preview shows pages 1–6. Sign up to view the full content.

FIN 271 FINANCIAL MODELING AND ECONOMETRICS LECTURE SET 4 TIME SERIES MODELING REFIK SOYER THE GEORGE WASHINGTON UNIVERSITY

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
2 MOVING AVERAGE (MA) PROCESSES ]  œ ] œ   â  t t t 1 t 1 2 t 2 q t q . % ) % ) % ) % . is a MA(q). MA process of order q Ê Note: E( ) and E( ) 0 . ] œ ] œ t t . The MA process is a linear combination of the current and the past values of the white noise process; thus, the MA process is always stationary. The back shift operator representation : ] œ  â  . ( 1 B B B ) t 1 2 q t 2 q ðóóóóóóóóóóóóóóóóñóóóóóóóóóóóóóóóóò ) ) ) % (B) ) Ê ] œ (B) . t t ) % ) (B) transfer function of the MA process. ´ The MA process is a linear combination of the current and the past values of the white noise process; recall the MA representation of AR(1) process. 1
3 • Special Case: First order MA process: MA(1) ] œ   t t 1 t 1 . % ) % (Conventional representation) Variable of interest is written as a "moving average" of random disturbance terms. ] t For AR and MA models (also for ARMA models), it can be shown that the random disturbance terms are one step ahead forecast errors (1) ^ % t t t 1 œ ]  ] where (1) is the forecast for at time (t 1). ^ ] ] t 1 t Ê ] % t t is the one step ahead forecast error for forecasting based on the information at time (t 1). We can write the MA(1) process ] œ œ . (1 B) t t 1 t 1 1 t % ) % ) % where ( ) 0 for all t, since ( ) 0 for all t. . . ] œ ]  Ê I ] œ I œ t t t t . % 2

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
4 Z +< ] œ Z +< œ Z +< Z +< ( ) ( ) ( ) ( ) . t t 1 t 1 t t 1 1 2 % ) % % ) % (1 ). Ê œ # 5 ) 0 2 2 1 % The ACF of MA(1) Process: # . . k t t k t t k œ IÒ ]  ]  Ó œ I ] ] ( ) ( ) ( ) . . First, consider . We can multiply both sides by and take expectations: . # 1 t 1 ] ] ] œ ] œ . . . ) ) ( ) t t 1 t 1 t 1 t 1 t 1 t 1 t 1 1 t 2 % ) % % ) % % ) % I ] ] œ I ( ) ( ) ( ) . . t t 1 t 1 t 1 t 1 1 t 2 % ) % % ) % I ] ] œ I  I  I  I ( ) ( ) ( ) ( ) ( ) . . t t 1 t t 1 1 t t 2 1 t 1 t 2 2 2 t 1 1 % % ) % % ) % ) % % I œ Ê œ ] ] œ  œ  ( ) for all t E( ) E( ) . . % 5 # ) % ) 5 t 2 2 2 2 1 t t 1 1 1 t 1 % % Thus, we have (1 ) (1 ) 3 # ) 5 ) # 5 ) ) 1 1 1 1 0 2 2 1 1 2 2 œ œ œ % % What about higher order autocorrelations ? # % ) % % ) % 2 t t 2 t 1 t 1 t 2 1 t 3 œ ] ] œ œ E( ) E( ) ( ) 0 . . 3
5 In general for the MA(1) process # 3 k k œ Ê œ 0 for k 1 0 for k 1. Recall that a stationary AR process has an infinite MA process representation. Similarly an MA process can be represented as infinite AR process under certain conditions.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 02/29/2012 for the course FINA 6271 taught by Professor Phillipwirtz,refiksoyer during the Fall '11 term at GWU.

### Page1 / 23

TS_Fall_2011_Lecture_4 - FIN 271 FINANCIAL MODELING AND...

This preview shows document pages 1 - 6. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online