ECON/FIN 250: Forecasting in Finance and Economics: Section 6: Standard Univariate Models

Size: px
Start display at page:

Download "ECON/FIN 250: Forecasting in Finance and Economics: Section 6: Standard Univariate Models"

Transcription

1 ECON/FIN 250: Forecasting in Finance and Economics: Section 6: Standard Univariate Models Patrick Herb Brandeis University Spring 2016 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

2 Course Overview 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

3 Key Objectives 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

4 Key Objectives Basic time series tools Moving average models Autoregressive models Mixed ARMA models Some special models Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

5 Lag Operators 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

6 Lag Operators Useful tools for time series Not too complicated Sometimes can get difficult Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

7 Lag Operators Ly t = y t 1 (1) L(Ly t ) = L 2 y t = y t 2 (2) L(L 2 y t ) = L 3 y t = y t 3 (3) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

8 Lag Operators y t = µ + ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2 (4) y t = µ + ɛ t + θ 1 Lɛ t + θ 2 L 2 ɛ t (5) y t = µ + (1 + θ 1 L + θ 2 L 2 )ɛ t (6) y t = µ + θ(l)e t (7) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

9 Other Operators Fy t = y t+1 (8) Dy t = y t y t 1 (9) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

10 Operators in Stata This notation moves easily into Stata: reg y L.y L2.y reg L (0/2). y reg F.y D.y Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

11 Moving Average Models 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

12 MA(1): Moving Average of Order 1 y t = ɛ t + θɛ t 1 (10) ɛ t WN(0, σ 2 ɛ ) (11) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

13 Comparing Two MA(1) Models t θ = 0.4 θ = 0.95 see software file compma1.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

14 Comparing Two MA(1) Models t θ = 0.3 θ = 0.9 see software file compma1same.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

15 MA(1): Unconditional Mean y t = ɛ t + θɛ t 1 E[y t ] = E[ɛ t ] + θe[ɛ t 1 ] = 0 (12) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

16 MA(1): Unconditional Variance y t = ɛ t + θɛ t 1 Why? Var[y t ] = Var[ɛ t ] + θ 2 Var[ɛ t 1 ] (13) = σ 2 ɛ (1 + θ 2 ) (14) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

17 MA(1): Unconditional Variance y t = ɛ t + θɛ t 1 Var[y t ] = Var[ɛ t ] + θ 2 Var[ɛ t 1 ] (15) = σɛ 2 (1 + θ 2 ) (16) Why? E[ɛ t ɛ t 1 ] = 0 (17) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

18 MA(1): Conditional Mean y t = ɛ t + θɛ t 1 E[y t Ω t 1 ] = E[ɛ t Ω t 1 ] + θe[ɛ t 1 Ω t 1 ] (18) = 0 + θɛ t 1 (19) = θɛ t 1 (20) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

19 MA(1): Conditional Variance y t = ɛ t + θɛ t 1 Var[y t Ω t 1 ] = E[(y t E[y t Ω t 1 ]) 2 Ω t 1 ] (21) = E[(ɛ t + θɛ t 1 θɛ t 1 ) 2 Ω t 1 ] (22) = σ 2 ɛ (23) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

20 MA(1): Autocovariance y t = ɛ t + θɛ t 1 γ j = E[y t y t j ] (24) γ 1 = E[y t y t 1 ] (25) γ 1 = E[(ɛ t + θɛ t 1 )(ɛ t 1 + θɛ t 2 )] (26) γ 1 = θe[ɛ t 1 ɛ t 1 ] = θσ 2 ɛ (27) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

21 MA(1): Autocovariance y t = ɛ t + θɛ t 1 γ 2 = E[y t y t 2 ] (28) γ 2 = E[(ɛ t + θɛ t 1 )(ɛ t 2 + θɛ t 3 )] (29) γ 2 = 0 (30) γ j = 0 j 2 (31) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

22 MA(1): Autocovariance y t = ɛ t + θɛ t 1 θσɛ 2 if j = 1 γ j = 0 if j > 1 (32) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

23 MA(1): Autocorrelation y t = ɛ t + θɛ t 1 ρ j = γ j Var[y t ] = γ j γ j = (33) γ 0 (1 + θ 2 )σ 2 θ ρ j =, if j = 1 (1+θ 2 ) (34) 0, if j > 1 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

24 MA(q): Moving Average of Order q y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q = Θ(L)ɛ t (35) ɛ t WN(0, σ 2 ɛ ) (36) Θ(L) = 1 + θ 1 L + θ 2 L θ q L q (37) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

25 MA(q): Unconditional Mean y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q = Θ(L)ɛ t E[y t ] = E[ɛ t ] + θ 1 E[ɛ t 1 ] + θ 2 E[ɛ t 2 ] + + θ q E[ɛ t q ] = 0 (38) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

26 MA(q): Unconditional Variance y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q = Θ(L)ɛ t Var[y t ] = Var[ɛ t ] + θ 2 1Var[ɛ t 1 ] + + θ 2 qvar[ɛ t q ] (39) Var[y t ] = σ 2 ɛ (1 + θ θ θ 2 q) (40) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

27 MA(q): Conditional Mean y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q = Θ(L)ɛ t E[y t Ω t 1 ] = E[ɛ t Ω t 1 ]+θ 1 E[ɛ t 1 Ω t 1 ]+ +θ q E[ɛ t q Ω t 1 ] (41) E[y t Ω t 1 ] = 0 + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q (42) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

28 MA(q): Conditional Variance y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q = Θ(L)ɛ t Var[y t Ω t 1 ] = E[(y t E[y t Ω t 1 ]) 2 Ω t 1 ] (43) Var[y t Ω t 1 ] = E[ɛ 2 t Ω t 1 ] (44) Var[y t Ω t 1 ] = σ 2 ɛ (45) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

29 MA(2): Autocovariance y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2 γ j = E[y t y t j ] (46) γ 2 = E[y t y t 2 ] (47) γ 2 = E[(ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2 )(ɛ t 2 + θ 1 ɛ t 3 + θ 2 ɛ t 4 )] (48) γ 2 = θ 2 σ 2 ɛ (49) γ j = 0 j > 2 (50) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

30 MA(q): Autocovariance & Autocorrelation y t = ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t θ q ɛ t q 0 if j q γ j = 0 if j > q 0 if j q ρ j = 0 if j > q (51) (52) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

31 Autocorrelation Function: MA(1) & MA(3) Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands see software file compma1ma3.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

32 MA( ): Moving Average of Order Complicated Important tool y t = ɛ t + θ j ɛ t j (53) j=1 ɛ t WN(0, σ 2 ɛ ) (54) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

33 Autoregressive Models 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

34 AR(1): Autoregressive Model of Order 1 y t = φy t 1 + ɛ t (55) ɛ t WN(0, σ 2 ɛ ) (56) (1 φl)y t = ɛ t (57) φ < 1 (58) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

35 Comparing Two AR(1) Models t φ = 0.2 φ = 0.95 see software file compar1.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

36 AR(1): Unconditional Mean y t = φy t 1 + ɛ t E[y t ] = φe[y t 1 ] + E[ɛ t ] (59) E[y t ] = φe[y t ] + E[ɛ t ] (60) (1 φ)e[y t ] = E[ɛ t ] = 0 (61) E[y t ] = 0 (62) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

37 AR(1): Unconditional Variance y t = φy t 1 + ɛ t Var[y t ] = φ 2 Var[y t 1 ] + Var[ɛ t ] (63) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

38 AR(1): Unconditional Variance y t = φy t 1 + ɛ t Why? Var[y t ] = φ 2 Var[y t 1 ] + Var[ɛ t ] (63) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

39 AR(1): Unconditional Variance y t = φy t 1 + ɛ t Why? Var[y t ] = φ 2 Var[y t 1 ] + Var[ɛ t ] (63) E[ɛ t y t 1 ] = 0 (64) Var[y t ] = φ 2 Var[y t ] + Var[ɛ t ] (65) Var[y t ] = E[y 2 t ] = σ2 ɛ 1 φ 2 (66) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

40 AR(1): Conditional Mean y t = φy t 1 + ɛ t E[y t Ω t 1 ] = E[φy t 1 Ω t 1 ] + E[ɛ t Ω t 1 ] (67) E[y t Ω t 1 ] = φy t 1 (68) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

41 AR(1): Conditional Variance y t = φy t 1 + ɛ t Var[y t Ω t 1 ] = E[(y t E[y t Ω t 1 ]) 2 Ω t 1 ] (69) Var[y t Ω t 1 ] = E[ɛ 2 t Ω t 1 ] (70) Var[y t Ω t 1 ] = σ 2 ɛ (71) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

42 AR(1): Autocovariance & Yule-Walker Equation y t = φy t 1 + ɛ t y t y t j = φy t 1 y t j + ɛ t y t j (72) E[y t y t j ] = φe[y t 1 y t j ] + E[ɛ t y t j ] (73) γ j = φγ j 1 (74) Since we know γ 0 = Var[y t ], we can use the Yule-Walker equation to get all the autocovariances. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

43 AR(1): Autocovariance γ j = φγ j 1 γ 0 = σ2 ɛ 1 φ 2 (75) σ2 ɛ γ 1 = φ (76) 1 φ 2 σ 2 ɛ γ 2 = φ 2 (77) 1 φ 2 σ 2 ɛ γ j = φ j (78) 1 φ 2 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

44 AR(1): Autocorrelation ρ j = γ j γ 0 ρ 1 = φ (79) ρ 2 = φ 2 (80) ρ j = φ j (81) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

45 AR(1): Comparing Autocorrelation Functions: φ = 0.4, φ = 0.95 Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands see software file compar1acf.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

46 AR(1): Comparing Autocorrelation Functions: φ = 0.9, φ = 0.9 Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands see software file compar1acfsign.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

47 Partial Autocorrelation Function (PACF) Regress y t on lags: y t = φ 1 y t 1 + φ 2 y t 2 + φ 3 y t Estimates of ˆφ j are the partial autocorrelations for lag j. For an AR(1) we would get: = φ if j = 1 ˆφ j = 0 if j > 1 (82) For an AR(p): 0 if j p ˆφ j = 0 if j > p (83) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

48 AR(1): Comparing PACF with φ = 0.4, φ = 0.95 Partial autocorrelations of y Lag 95% Confidence bands [se = 1/sqrt(n)] Partial autocorrelations of y Lag 95% Confidence bands [se = 1/sqrt(n)] see software file compar1pacf.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

49 AR(p): Autoregressive Model of Order p p y t = φ j y t j + ɛ t (84) j=1 ɛ t WN(0, σɛ 2 ) (85) Φ(L) = (1 φ 1 L φ 2 L 2... φ p L p ) (86) Φ(L)y t = ɛ t (87) y t = ɛ t Φ(L) (88) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

50 AR(p): Autoregressive Model of Order p AR(1) only has two kinds of persistence: φ is positive or negative. If y t = 0.9y t 1 + ɛ t, then the ACF is positive with decay If y t = 0.9y t 1 + ɛ t, then the ACF oscillates between positive and negative with decay AR(p) has more complicated patterns and can have large positive and negative swings Also, stationarity is trickier Example: y t = 1.5y t 1 0.9y t 2 + ɛ t (89) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

51 AR(2): Autocorrelation Function Autocorrelations of y Lag Bartlett s formula for MA(q) 95% confidence bands see software file acfar2.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

52 AR(2): Autocovariance & Yule-Walker Equation y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

53 AR(2): Autocovariance & Yule-Walker Equation y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t y t j = φ 1 y t 1 y t j + φ 2 y t 2 y t j + ɛ t y t j (90) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

54 AR(2): Autocovariance & Yule-Walker Equation y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t y t j = φ 1 y t 1 y t j + φ 2 y t 2 y t j + ɛ t y t j (90) E[y t y t j ] = φ 1 E[y t 1 y t j ] + φ 2 E[y t 2 y t j ] + E[ɛ t y t j ] (91) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

55 AR(2): Autocovariance & Yule-Walker Equation y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t y t j = φ 1 y t 1 y t j + φ 2 y t 2 y t j + ɛ t y t j (90) E[y t y t j ] = φ 1 E[y t 1 y t j ] + φ 2 E[y t 2 y t j ] + E[ɛ t y t j ] (91) γ j = φ 1 γ j 1 + φ 2 γ j 2 (92) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

56 AR(2): Autocovariance & Yule-Walker Equation y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t y t j = φ 1 y t 1 y t j + φ 2 y t 2 y t j + ɛ t y t j (90) E[y t y t j ] = φ 1 E[y t 1 y t j ] + φ 2 E[y t 2 y t j ] + E[ɛ t y t j ] (91) Divide by γ 0 = Var[y t ] γ j = φ 1 γ j 1 + φ 2 γ j 2 (92) ρ j = φ 1 ρ j 1 + φ 2 ρ j 2 (93) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

57 AR(2): Autocorrelation & Yule-Walker Equation ρ j = φ 1 ρ j 1 + φ 2 ρ j 2 ρ 0 = 1 (94) ρ 1 = φ 1 ρ 0 + φ 2 ρ 1 = φ 1 ρ 0 + φ 2 ρ 1 (95) ρ 1 = φ 1 1 φ 2 (96) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

58 Stationarity Question: Is the process blowing up (nonstationary)? What are the features of a stationary process? Constant unconditional mean Constant and finite unconditional variance Autocorrelations depends only on displacement (time) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

59 Stationary vs. Nonstationary AR(1) t φ = 1.1 φ = 0.9 see software file unstablear1.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

60 Quick & Dirty AR(p) Stationarity Check A necessary condition for covariance stationarity is p φ i < 1 (97) i=1 If this condition fails, you know the process is not stationary. However, if the condition holds, it may or may not be stationary. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

61 AR(2): Covariance Stationarity Covariance stationarity depends upon the roots of the lag operator polynomial. y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t (1 φ 1 L φ 2 L 2 ) = ɛ t (98) y t (1 φ 1 z φ 2 z 2 ) = ɛ t (99) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

62 AR(2): Covariance Stationarity Can find roots using quadratic formula: (1 φ 1 z φ 2 z 2 ) = 0 (100) z 1 = φ 1 φ φ 2 (101) 2φ 2 z 2 = φ 1 + φ φ 2 (102) 2φ 2 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

63 AR(2): Finding Roots y t = 1.5y t 1 0.9y t 2 + ɛ t y t (1 1.5L + 0.9L 2 ) = ɛ t (103) y t (1 1.5z + 0.9z 2 ) = ɛ t (104) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

64 AR(2): Finding Roots (1 1.5z + 0.9z 2 ) = 0 (105) z 1 = 1.5 ( 1.5) 2 4(0.9)(1) = i 2(0.9) (106) z 1 = 1.5 ( 1.5) 2 4(0.9)(1) = i 2(0.9) (107) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

65 AR(2): Covariance Stationarity: Modulus Stationarity: Roots must be greater than 1 in length, and can be calculated from the modulus: a 2 + b 2 = c 2 Modulus = R 2 + C 2 (108) R = Real Part and C = Complex Part. If root length > 1, then root lies outside of unit circle and the process is stationary. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

66 AR(2): Covariance Stationarity: Modulus y t = 1.5y t 1 0.9y t 2 + ɛ t roots = 0.83 ± 0.65i (109) Modulus = = 1.11 > 1 (110) The roots are outside the unit circle, so this process is stationary If root contains only real numbers, then the correlations decay over time If roots are complex, then correlations have cycles Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

67 Inverse Roots Some books and software alternatively test and report the inverse roots. y t = φ 1 y t 1 + φ 2 y t 2 + ɛ t y t (1 φ 1 L φ 2 L 2 ) = ɛ t (111) (z 2 φ 1 z 1 φ 2 ) = 0 (112) (λ 2 φ 1 λ φ 2 ) = 0, λ z 1 (113) Stationarity requires that the roots of λ must lie inside the unit circle such that: Modulus = R 2 + C 2 < 1 (114) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

68 AR(2): Inverse Roots y t = 1.5y t 1 0.9y t 2 + ɛ t y t (1 1.5L + 0.9L 2 ) = ɛ t (115) (λ 2 1.5λ + 0.9) = 0 (116) From the quadratic formula: Modulus = = < 1 (117) Stata tests the inverse roots using the estat aroots command Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

69 Technical Aside Alternatively we could write the second order difference equation in matrix notation [ yt y t 1 ] = [ φ1 φ ] [ ] yt 1 + y t 2 [ ] et 0 (118) In this case, the inverse roots λ 1, λ 2 are the eigenvalues of the matrix Which is the solution to F λi = F = [ ] φ1 φ [ ] φ1 φ (119) [ ] λ 0 0 λ = 0 (120) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

70 AR(p): Autoregressive of Order p p y t = φ j y t j + ɛ t j=1 Stationarity conditions more complicated Involve pth order polynomials Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

71 Converting Between AR & MA Models 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

72 Finite Order Moving Averages All finite order moving averages are covariance stationary Not all finite order moving averages are invertible Finite order moving averages are invertible if the inverse roots of the lag operator polynomial are inside the unit circle If the finite order moving average is invertible, then it can be written as an AR( ) process Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

73 MA(1) to AR( ) MA(1) y t = ɛ t + θɛ t 1, ɛ t WN(0, σ 2 ), θ < 1 We can solve for the innovation ɛ t = y t θɛ t 1 (121) The lagged innovations are ɛ t 1 = y t 1 θɛ t 2 (122) ɛ t 2 = y t 2 θɛ t 3 (123) ɛ t 3 = y t 3 θɛ t 4 (124) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

74 MA(1) to AR( ) Substitute the innovations into the MA(1) process y t = ɛ t + θɛ t 1, ɛ t WN(0, σ 2 ), θ < 1 to yield an AR( ) y t = ɛ t + θy t 1 θ 2 y t 2 + θ 3 y t 3... (125) Or in lag notation θl y t = ɛ t (126) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

75 MA(1): Comparing PACF with θ = 0.9, θ = 0.9 Partial autocorrelations of y Lag 95% Confidence bands [se = 1/sqrt(n)] Partial autocorrelations of y Lag 95% Confidence bands [se = 1/sqrt(n)] see software file compma1pacf.do Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

76 MA(q) & MA( ) The same approach can be used to determine invertibility of finite MA(q) processes q < When q, stability is no longer guaranteed Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

77 MA( ) For an MA( ) process y t = ψ j ɛ t j = ψ 0 ɛ t + ψ 1 ɛ t 1 + ψ 1 ɛ t (127) j=0 The infinite sequence is covariance stationary provided that ψj 2 < (128) j=0 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

78 AR(1) to MA( ) by Substitution y t = φy t 1 + ɛ t, y t 1 = φy t 2 + ɛ t 1 y t = φ(φy t 2 + ɛ t 1 ) + ɛ t (129) y t = φ 2 y t 2 + φɛ t 1 + ɛ t (130) y t = φ m y t m + m 1 j=1 φ j ɛ t j + ɛ t (131) y t = φ j ɛ t j + ɛ t = ɛ t j (132) j=1 j=0 Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

79 Stationary AR Processes AR processes are always invertible AR processes are not always stationary Stationarity requires the inverse roots of the lag operator polynomial to be inside the unit circle Any stationary AR(p) process can be converted to an MA( ) Conversion can make some calculations easier Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

80 AR(1) to MA( ) with Lag Operator y t = φy t 1 + ɛ t, φ < 1 y t = (1 φl)y t = ɛ t (133) 1 1 φl ɛ t = φ j ɛ t j (134) j=0 1 1 φl = 1 + φl + φ2 L 2 + φ 3 L (135) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

81 Autoregressive Moving Average (ARMA) Models 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

82 ARMA(p,q) Models p q y t = φ j y t j + θ j ɛ t j + ɛ t (136) j=1 j=1 Combines AR and MA components Richer Dynamics Can often reduce parameter estimates (increase parsimony) Can result from aggregation Sums of AR processes, or sums of AR and MA process can be ARMA processes AR processes observed subject to measurement error also turn out to be ARMA processes Connection to state space models Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

83 ARMA(1,1) In lag notation y t = φy t 1 + θɛ t 1 + ɛ t, ɛ t WN(0, σ 2 ) (137) Stationarity requires φ < 1 Invertibility requires θ < 1 (1 φl)y t = (1 + θl)ɛ t (138) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

84 ARMA(1,1) If the covariance stationarity condition is satisfied, then we can convert this to an infinite MA process y t = (1 + θl) (1 φl) ɛ t (139) If the invertibility condition is satisfied, then we can convert this to the infinite order AR process (1 φl) (1 + θl) y t = ɛ t (140) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

85 ARMA(1,1): Extreme Example y t = 0.99y t ɛ t 1 + ɛ t (141) Run arma11acf.do Note the ACF pattern Extreme persistence and low level Sometimes difficult to see Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

86 Possible Model Origins y t = x t + η t (142) x t = ρx t 1 + µ t (143) Observe only y t x t is hidden With ρ large (close to one), x t moves slowly With Var[η t ] large this is difficult to see/measure Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

87 Put These Together y t = x t + η t x t = ρx t 1 + µ t y t = ρx t 1 + µ t + η t = ρ(y t 1 η t 1 ) + µ t + η t = ρy t 1 ρη t 1 + (µ t + η t ) = ρy t 1 θe t 1 + e t Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

88 Solving for ARMA Parameters (Optional Note) Given that you know the parameters for the noisy random walk, how do you map to an ARMA(1,1)? y t = x t + η t, x t = ρx t 1 + µ t y t = ρy t 1 ρη t 1 + (µ t + η t ) = ρy t 1 θe t 1 + e t = ρy t 1 + z t Var[z t ] = (1 + θ 2 )σ 2 e = (1 + ρ 2 )σ 2 η + σ 2 µ Cov[z t, z t 1 ] = θσ 2 e = ρσ 2 η σe 2 = (ρ/θ)ση 2 (1 + θ 2 )(ρ/θ)ση 2 = (1 + ρ 2 )ση 2 + σµ 2 Now solve quadratic for θ. You know all the other values. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

89 Nonzero Expected Values 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

90 Adding Constants q y t = θ 0 + θ j ɛ t j + ɛ t (144) j=1 p y t = φ 0 + φ j y t j + ɛ t (145) j=1 For the MA model this is a trivial change. For the AR, it can be trickier. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

91 AR(1): Adding a Constant y t = φ 0 + φ 1 y t 1 + ɛ t (146) E(y t ) = φ 0 + φ 1 E(y t 1 ) + E(ɛ t ) (147) E(y t ) = φ 0 + φ 1 E(y t ) + 0 (148) E(y t ) = φ 0 1 φ 1 (149) Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

92 Linear Time Series: Big Picture 1 Key Objectives 2 Lag Operators 3 Moving Average Models 4 Autoregressive Models 5 Converting Between AR & MA Models 6 Autoregressive Moving Average (ARMA) Models 7 Nonzero Expected Values 8 Linear Time Series: Big Picture Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

93 Frequently Used Models MA(q) AR(1) AR(2) ARMA(1,1) Random Walks Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

94 Thoughts on Linear Models Many flavors of models (MA, AR, ARMA) AR(2) very different from AR(1) ARMA(1,1) is more useful than you might think Most of these do map to a MA( ) Certain approximate models may have fewer parameters (parsimonious), and be more useful for forecasting. Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

95 Thoughts on Linear Models Many flavors of models (MA, AR, ARMA) AR(2) very different from AR(1) ARMA(1,1) is more useful than you might think Most of these do map to a MA( ) Certain approximate models may have fewer parameters (parsimonious), and be more useful for forecasting. In practice, model identification can be tricky Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

96 Thoughts on Linear Models Many flavors of models (MA, AR, ARMA) AR(2) very different from AR(1) ARMA(1,1) is more useful than you might think Most of these do map to a MA( ) Certain approximate models may have fewer parameters (parsimonious), and be more useful for forecasting. In practice, model identification can be tricky Other families of models Nonlinear Long memory Regime shifting Patrick Herb (Brandeis University) Standard Univariate Models ECON/FIN 250: Spring / 88

ECON/FIN 250: Forecasting in Finance and Economics: Section 7: Unit Roots & Dickey-Fuller Tests

ECON/FIN 250: Forecasting in Finance and Economics: Section 7: Unit Roots & Dickey-Fuller Tests ECON/FIN 250: Forecasting in Finance and Economics: Section 7: Unit Roots & Dickey-Fuller Tests Patrick Herb Brandeis University Spring 2016 Patrick Herb (Brandeis University) Unit Root Tests ECON/FIN

More information

Chapter 4: Models for Stationary Time Series

Chapter 4: Models for Stationary Time Series Chapter 4: Models for Stationary Time Series Now we will introduce some useful parametric models for time series that are stationary processes. We begin by defining the General Linear Process. Let {Y t

More information

Lecture 1: Fundamental concepts in Time Series Analysis (part 2)

Lecture 1: Fundamental concepts in Time Series Analysis (part 2) Lecture 1: Fundamental concepts in Time Series Analysis (part 2) Florian Pelgrin University of Lausanne, École des HEC Department of mathematics (IMEA-Nice) Sept. 2011 - Jan. 2012 Florian Pelgrin (HEC)

More information

Covariance Stationary Time Series. Example: Independent White Noise (IWN(0,σ 2 )) Y t = ε t, ε t iid N(0,σ 2 )

Covariance Stationary Time Series. Example: Independent White Noise (IWN(0,σ 2 )) Y t = ε t, ε t iid N(0,σ 2 ) Covariance Stationary Time Series Stochastic Process: sequence of rv s ordered by time {Y t } {...,Y 1,Y 0,Y 1,...} Defn: {Y t } is covariance stationary if E[Y t ]μ for all t cov(y t,y t j )E[(Y t μ)(y

More information

Ch 4. Models For Stationary Time Series. Time Series Analysis

Ch 4. Models For Stationary Time Series. Time Series Analysis This chapter discusses the basic concept of a broad class of stationary parametric time series models the autoregressive moving average (ARMA) models. Let {Y t } denote the observed time series, and {e

More information

Forecasting with ARMA

Forecasting with ARMA Forecasting with ARMA Eduardo Rossi University of Pavia October 2013 Rossi Forecasting Financial Econometrics - 2013 1 / 32 Mean Squared Error Linear Projection Forecast of Y t+1 based on a set of variables

More information

Midterm Suggested Solutions

Midterm Suggested Solutions CUHK Dept. of Economics Spring 2011 ECON 4120 Sung Y. Park Midterm Suggested Solutions Q1 (a) In time series, autocorrelation measures the correlation between y t and its lag y t τ. It is defined as. ρ(τ)

More information

ECON/FIN 250: Forecasting in Finance and Economics: Section 8: Forecast Examples: Part 1

ECON/FIN 250: Forecasting in Finance and Economics: Section 8: Forecast Examples: Part 1 ECON/FIN 250: Forecasting in Finance and Economics: Section 8: Forecast Examples: Part 1 Patrick Herb Brandeis University Spring 2016 Patrick Herb (Brandeis University) Forecast Examples: Part 1 ECON/FIN

More information

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

Lecture 2: ARMA(p,q) models (part 2) Lecture 2: ARMA(p,q) models (part 2) Florian Pelgrin University of Lausanne, École des HEC Department of mathematics (IMEA-Nice) Sept. 2011 - Jan. 2012 Florian Pelgrin (HEC) Univariate time series Sept.

More information

Ch. 14 Stationary ARMA Process

Ch. 14 Stationary ARMA Process Ch. 14 Stationary ARMA Process A general linear stochastic model is described that suppose a time series to be generated by a linear aggregation of random shock. For practical representation it is desirable

More information

3. ARMA Modeling. Now: Important class of stationary processes

3. ARMA Modeling. Now: Important class of stationary processes 3. ARMA Modeling Now: Important class of stationary processes Definition 3.1: (ARMA(p, q) process) Let {ɛ t } t Z WN(0, σ 2 ) be a white noise process. The process {X t } t Z is called AutoRegressive-Moving-Average

More information

Econometrics II Heij et al. Chapter 7.1

Econometrics II Heij et al. Chapter 7.1 Chapter 7.1 p. 1/2 Econometrics II Heij et al. Chapter 7.1 Linear Time Series Models for Stationary data Marius Ooms Tinbergen Institute Amsterdam Chapter 7.1 p. 2/2 Program Introduction Modelling philosophy

More information

Autoregressive and Moving-Average Models

Autoregressive and Moving-Average Models Chapter 3 Autoregressive and Moving-Average Models 3.1 Introduction Let y be a random variable. We consider the elements of an observed time series {y 0,y 1,y2,...,y t } as being realizations of this randoms

More information

Univariate Time Series Analysis; ARIMA Models

Univariate Time Series Analysis; ARIMA Models Econometrics 2 Fall 24 Univariate Time Series Analysis; ARIMA Models Heino Bohn Nielsen of4 Outline of the Lecture () Introduction to univariate time series analysis. (2) Stationarity. (3) Characterizing

More information

at least 50 and preferably 100 observations should be available to build a proper model

at least 50 and preferably 100 observations should be available to build a proper model III Box-Jenkins Methods 1. Pros and Cons of ARIMA Forecasting a) need for data at least 50 and preferably 100 observations should be available to build a proper model used most frequently for hourly or

More information

Autoregressive Moving Average (ARMA) Models and their Practical Applications

Autoregressive Moving Average (ARMA) Models and their Practical Applications Autoregressive Moving Average (ARMA) Models and their Practical Applications Massimo Guidolin February 2018 1 Essential Concepts in Time Series Analysis 1.1 Time Series and Their Properties Time series:

More information

Econometría 2: Análisis de series de Tiempo

Econometría 2: Análisis de series de Tiempo Econometría 2: Análisis de series de Tiempo Karoll GOMEZ kgomezp@unal.edu.co http://karollgomez.wordpress.com Segundo semestre 2016 III. Stationary models 1 Purely random process 2 Random walk (non-stationary)

More information

Lecture 3: Autoregressive Moving Average (ARMA) Models and their Practical Applications

Lecture 3: Autoregressive Moving Average (ARMA) Models and their Practical Applications Lecture 3: Autoregressive Moving Average (ARMA) Models and their Practical Applications Prof. Massimo Guidolin 20192 Financial Econometrics Winter/Spring 2018 Overview Moving average processes Autoregressive

More information

Chapter 9: Forecasting

Chapter 9: Forecasting Chapter 9: Forecasting One of the critical goals of time series analysis is to forecast (predict) the values of the time series at times in the future. When forecasting, we ideally should evaluate the

More information

Discrete time processes

Discrete time processes Discrete time processes Predictions are difficult. Especially about the future Mark Twain. Florian Herzog 2013 Modeling observed data When we model observed (realized) data, we encounter usually the following

More information

University of Oxford. Statistical Methods Autocorrelation. Identification and Estimation

University of Oxford. Statistical Methods Autocorrelation. Identification and Estimation University of Oxford Statistical Methods Autocorrelation Identification and Estimation Dr. Órlaith Burke Michaelmas Term, 2011 Department of Statistics, 1 South Parks Road, Oxford OX1 3TG Contents 1 Model

More information

ECON 616: Lecture 1: Time Series Basics

ECON 616: Lecture 1: Time Series Basics ECON 616: Lecture 1: Time Series Basics ED HERBST August 30, 2017 References Overview: Chapters 1-3 from Hamilton (1994). Technical Details: Chapters 2-3 from Brockwell and Davis (1987). Intuition: Chapters

More information

Introduction to ARMA and GARCH processes

Introduction to ARMA and GARCH processes Introduction to ARMA and GARCH processes Fulvio Corsi SNS Pisa 3 March 2010 Fulvio Corsi Introduction to ARMA () and GARCH processes SNS Pisa 3 March 2010 1 / 24 Stationarity Strict stationarity: (X 1,

More information

Empirical Market Microstructure Analysis (EMMA)

Empirical Market Microstructure Analysis (EMMA) Empirical Market Microstructure Analysis (EMMA) Lecture 3: Statistical Building Blocks and Econometric Basics Prof. Dr. Michael Stein michael.stein@vwl.uni-freiburg.de Albert-Ludwigs-University of Freiburg

More information

Class 1: Stationary Time Series Analysis

Class 1: Stationary Time Series Analysis Class 1: Stationary Time Series Analysis Macroeconometrics - Fall 2009 Jacek Suda, BdF and PSE February 28, 2011 Outline Outline: 1 Covariance-Stationary Processes 2 Wold Decomposition Theorem 3 ARMA Models

More information

Problem Set 2 Solution Sketches Time Series Analysis Spring 2010

Problem Set 2 Solution Sketches Time Series Analysis Spring 2010 Problem Set 2 Solution Sketches Time Series Analysis Spring 2010 Forecasting 1. Let X and Y be two random variables such that E(X 2 ) < and E(Y 2 )

More information

3 Theory of stationary random processes

3 Theory of stationary random processes 3 Theory of stationary random processes 3.1 Linear filters and the General linear process A filter is a transformation of one random sequence {U t } into another, {Y t }. A linear filter is a transformation

More information

Review Session: Econometrics - CLEFIN (20192)

Review Session: Econometrics - CLEFIN (20192) Review Session: Econometrics - CLEFIN (20192) Part II: Univariate time series analysis Daniele Bianchi March 20, 2013 Fundamentals Stationarity A time series is a sequence of random variables x t, t =

More information

Some Time-Series Models

Some Time-Series Models Some Time-Series Models Outline 1. Stochastic processes and their properties 2. Stationary processes 3. Some properties of the autocorrelation function 4. Some useful models Purely random processes, random

More information

EASTERN MEDITERRANEAN UNIVERSITY ECON 604, FALL 2007 DEPARTMENT OF ECONOMICS MEHMET BALCILAR ARIMA MODELS: IDENTIFICATION

EASTERN MEDITERRANEAN UNIVERSITY ECON 604, FALL 2007 DEPARTMENT OF ECONOMICS MEHMET BALCILAR ARIMA MODELS: IDENTIFICATION ARIMA MODELS: IDENTIFICATION A. Autocorrelations and Partial Autocorrelations 1. Summary of What We Know So Far: a) Series y t is to be modeled by Box-Jenkins methods. The first step was to convert y t

More information

2. An Introduction to Moving Average Models and ARMA Models

2. An Introduction to Moving Average Models and ARMA Models . An Introduction to Moving Average Models and ARMA Models.1 White Noise. The MA(1) model.3 The MA(q) model..4 Estimation and forecasting of MA models..5 ARMA(p,q) models. The Moving Average (MA) models

More information

Lecture on ARMA model

Lecture on ARMA model Lecture on ARMA model Robert M. de Jong Ohio State University Columbus, OH 43210 USA Chien-Ho Wang National Taipei University Taipei City, 104 Taiwan ROC October 19, 2006 (Very Preliminary edition, Comment

More information

Time Series Solutions HT 2009

Time Series Solutions HT 2009 Time Series Solutions HT 2009 1. Let {X t } be the ARMA(1, 1) process, X t φx t 1 = ɛ t + θɛ t 1, {ɛ t } WN(0, σ 2 ), where φ < 1 and θ < 1. Show that the autocorrelation function of {X t } is given by

More information

4. MA(2) +drift: y t = µ + ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2. Mean: where θ(l) = 1 + θ 1 L + θ 2 L 2. Therefore,

4. MA(2) +drift: y t = µ + ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2. Mean: where θ(l) = 1 + θ 1 L + θ 2 L 2. Therefore, 61 4. MA(2) +drift: y t = µ + ɛ t + θ 1 ɛ t 1 + θ 2 ɛ t 2 Mean: y t = µ + θ(l)ɛ t, where θ(l) = 1 + θ 1 L + θ 2 L 2. Therefore, E(y t ) = µ + θ(l)e(ɛ t ) = µ 62 Example: MA(q) Model: y t = ɛ t + θ 1 ɛ

More information

Permanent Income Hypothesis (PIH) Instructor: Dmytro Hryshko

Permanent Income Hypothesis (PIH) Instructor: Dmytro Hryshko Permanent Income Hypothesis (PIH) Instructor: Dmytro Hryshko 1 / 36 The PIH Utility function is quadratic, u(c t ) = 1 2 (c t c) 2 ; borrowing/saving is allowed using only the risk-free bond; β(1 + r)

More information

Lecture 1: Stationary Time Series Analysis

Lecture 1: Stationary Time Series Analysis Syllabus Stationarity ARMA AR MA Model Selection Estimation Lecture 1: Stationary Time Series Analysis 222061-1617: Time Series Econometrics Spring 2018 Jacek Suda Syllabus Stationarity ARMA AR MA Model

More information

1 Linear Difference Equations

1 Linear Difference Equations ARMA Handout Jialin Yu 1 Linear Difference Equations First order systems Let {ε t } t=1 denote an input sequence and {y t} t=1 sequence generated by denote an output y t = φy t 1 + ε t t = 1, 2,... with

More information

Week 5 Quantitative Analysis of Financial Markets Characterizing Cycles

Week 5 Quantitative Analysis of Financial Markets Characterizing Cycles Week 5 Quantitative Analysis of Financial Markets Characterizing Cycles Christopher Ting http://www.mysmu.edu/faculty/christophert/ Christopher Ting : christopherting@smu.edu.sg : 6828 0364 : LKCSB 5036

More information

Lecture 4a: ARMA Model

Lecture 4a: ARMA Model Lecture 4a: ARMA Model 1 2 Big Picture Most often our goal is to find a statistical model to describe real time series (estimation), and then predict the future (forecasting) One particularly popular model

More information

We will only present the general ideas on how to obtain. follow closely the AR(1) and AR(2) cases presented before.

We will only present the general ideas on how to obtain. follow closely the AR(1) and AR(2) cases presented before. ACF and PACF of an AR(p) We will only present the general ideas on how to obtain the ACF and PACF of an AR(p) model since the details follow closely the AR(1) and AR(2) cases presented before. Recall that

More information

Problem Set 1 Solution Sketches Time Series Analysis Spring 2010

Problem Set 1 Solution Sketches Time Series Analysis Spring 2010 Problem Set 1 Solution Sketches Time Series Analysis Spring 2010 1. Construct a martingale difference process that is not weakly stationary. Simplest e.g.: Let Y t be a sequence of independent, non-identically

More information

Chapter 6: Model Specification for Time Series

Chapter 6: Model Specification for Time Series Chapter 6: Model Specification for Time Series The ARIMA(p, d, q) class of models as a broad class can describe many real time series. Model specification for ARIMA(p, d, q) models involves 1. Choosing

More information

Time Series Analysis

Time Series Analysis Time Series Analysis Christopher Ting http://mysmu.edu.sg/faculty/christophert/ christopherting@smu.edu.sg Quantitative Finance Singapore Management University March 3, 2017 Christopher Ting Week 9 March

More information

Università di Pavia. Forecasting. Eduardo Rossi

Università di Pavia. Forecasting. Eduardo Rossi Università di Pavia Forecasting Eduardo Rossi Mean Squared Error Forecast of Y t+1 based on a set of variables observed at date t, X t : Yt+1 t. The loss function MSE(Y t+1 t ) = E[Y t+1 Y t+1 t ]2 The

More information

Module 4. Stationary Time Series Models Part 1 MA Models and Their Properties

Module 4. Stationary Time Series Models Part 1 MA Models and Their Properties Module 4 Stationary Time Series Models Part 1 MA Models and Their Properties Class notes for Statistics 451: Applied Time Series Iowa State University Copyright 2015 W. Q. Meeker. February 14, 2016 20h

More information

Lesson 9: Autoregressive-Moving Average (ARMA) models

Lesson 9: Autoregressive-Moving Average (ARMA) models Lesson 9: Autoregressive-Moving Average (ARMA) models Dipartimento di Ingegneria e Scienze dell Informazione e Matematica Università dell Aquila, umberto.triacca@ec.univaq.it Introduction We have seen

More information

18.S096 Problem Set 4 Fall 2013 Time Series Due Date: 10/15/2013

18.S096 Problem Set 4 Fall 2013 Time Series Due Date: 10/15/2013 18.S096 Problem Set 4 Fall 2013 Time Series Due Date: 10/15/2013 1. Covariance Stationary AR(2) Processes Suppose the discrete-time stochastic process {X t } follows a secondorder auto-regressive process

More information

Introduction to Stochastic processes

Introduction to Stochastic processes Università di Pavia Introduction to Stochastic processes Eduardo Rossi Stochastic Process Stochastic Process: A stochastic process is an ordered sequence of random variables defined on a probability space

More information

Ch. 15 Forecasting. 1.1 Forecasts Based on Conditional Expectations

Ch. 15 Forecasting. 1.1 Forecasts Based on Conditional Expectations Ch 15 Forecasting Having considered in Chapter 14 some of the properties of ARMA models, we now show how they may be used to forecast future values of an observed time series For the present we proceed

More information

Ch 5. Models for Nonstationary Time Series. Time Series Analysis

Ch 5. Models for Nonstationary Time Series. Time Series Analysis We have studied some deterministic and some stationary trend models. However, many time series data cannot be modeled in either way. Ex. The data set oil.price displays an increasing variation from the

More information

MAT 3379 (Winter 2016) FINAL EXAM (SOLUTIONS)

MAT 3379 (Winter 2016) FINAL EXAM (SOLUTIONS) MAT 3379 (Winter 2016) FINAL EXAM (SOLUTIONS) 15 April 2016 (180 minutes) Professor: R. Kulik Student Number: Name: This is closed book exam. You are allowed to use one double-sided A4 sheet of notes.

More information

Lecture 1: Stationary Time Series Analysis

Lecture 1: Stationary Time Series Analysis Syllabus Stationarity ARMA AR MA Model Selection Estimation Forecasting Lecture 1: Stationary Time Series Analysis 222061-1617: Time Series Econometrics Spring 2018 Jacek Suda Syllabus Stationarity ARMA

More information

STAT 443 Final Exam Review. 1 Basic Definitions. 2 Statistical Tests. L A TEXer: W. Kong

STAT 443 Final Exam Review. 1 Basic Definitions. 2 Statistical Tests. L A TEXer: W. Kong STAT 443 Final Exam Review L A TEXer: W Kong 1 Basic Definitions Definition 11 The time series {X t } with E[X 2 t ] < is said to be weakly stationary if: 1 µ X (t) = E[X t ] is independent of t 2 γ X

More information

Module 3. Descriptive Time Series Statistics and Introduction to Time Series Models

Module 3. Descriptive Time Series Statistics and Introduction to Time Series Models Module 3 Descriptive Time Series Statistics and Introduction to Time Series Models Class notes for Statistics 451: Applied Time Series Iowa State University Copyright 2015 W Q Meeker November 11, 2015

More information

Time Series Analysis Fall 2008

Time Series Analysis Fall 2008 MIT OpenCourseWare http://ocw.mit.edu 14.384 Time Series Analysis Fall 008 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. Introduction 1 14.384 Time

More information

FE570 Financial Markets and Trading. Stevens Institute of Technology

FE570 Financial Markets and Trading. Stevens Institute of Technology FE570 Financial Markets and Trading Lecture 5. Linear Time Series Analysis and Its Applications (Ref. Joel Hasbrouck - Empirical Market Microstructure ) Steve Yang Stevens Institute of Technology 9/25/2012

More information

Econ 623 Econometrics II Topic 2: Stationary Time Series

Econ 623 Econometrics II Topic 2: Stationary Time Series 1 Introduction Econ 623 Econometrics II Topic 2: Stationary Time Series In the regression model we can model the error term as an autoregression AR(1) process. That is, we can use the past value of the

More information

Univariate, Nonstationary Processes

Univariate, Nonstationary Processes Univariate, Nonstationary Processes Jamie Monogan University of Georgia March 20, 2018 Jamie Monogan (UGA) Univariate, Nonstationary Processes March 20, 2018 1 / 14 Objectives By the end of this meeting,

More information

Covariances of ARMA Processes

Covariances of ARMA Processes Statistics 910, #10 1 Overview Covariances of ARMA Processes 1. Review ARMA models: causality and invertibility 2. AR covariance functions 3. MA and ARMA covariance functions 4. Partial autocorrelation

More information

Ch 6. Model Specification. Time Series Analysis

Ch 6. Model Specification. Time Series Analysis We start to build ARIMA(p,d,q) models. The subjects include: 1 how to determine p, d, q for a given series (Chapter 6); 2 how to estimate the parameters (φ s and θ s) of a specific ARIMA(p,d,q) model (Chapter

More information

ARIMA Models. Jamie Monogan. January 16, University of Georgia. Jamie Monogan (UGA) ARIMA Models January 16, / 27

ARIMA Models. Jamie Monogan. January 16, University of Georgia. Jamie Monogan (UGA) ARIMA Models January 16, / 27 ARIMA Models Jamie Monogan University of Georgia January 16, 2018 Jamie Monogan (UGA) ARIMA Models January 16, 2018 1 / 27 Objectives By the end of this meeting, participants should be able to: Argue why

More information

Econometrics of financial markets, -solutions to seminar 1. Problem 1

Econometrics of financial markets, -solutions to seminar 1. Problem 1 Econometrics of financial markets, -solutions to seminar 1. Problem 1 a) Estimate with OLS. For any regression y i α + βx i + u i for OLS to be unbiased we need cov (u i,x j )0 i, j. For the autoregressive

More information

7. MULTIVARATE STATIONARY PROCESSES

7. MULTIVARATE STATIONARY PROCESSES 7. MULTIVARATE STATIONARY PROCESSES 1 1 Some Preliminary Definitions and Concepts Random Vector: A vector X = (X 1,..., X n ) whose components are scalar-valued random variables on the same probability

More information

Lecture note 2 considered the statistical analysis of regression models for time

Lecture note 2 considered the statistical analysis of regression models for time DYNAMIC MODELS FOR STATIONARY TIME SERIES Econometrics 2 LectureNote4 Heino Bohn Nielsen March 2, 2007 Lecture note 2 considered the statistical analysis of regression models for time series data, and

More information

Lecture 2: Univariate Time Series

Lecture 2: Univariate Time Series Lecture 2: Univariate Time Series Analysis: Conditional and Unconditional Densities, Stationarity, ARMA Processes Prof. Massimo Guidolin 20192 Financial Econometrics Spring/Winter 2017 Overview Motivation:

More information

{ } Stochastic processes. Models for time series. Specification of a process. Specification of a process. , X t3. ,...X tn }

{ } Stochastic processes. Models for time series. Specification of a process. Specification of a process. , X t3. ,...X tn } Stochastic processes Time series are an example of a stochastic or random process Models for time series A stochastic process is 'a statistical phenomenon that evolves in time according to probabilistic

More information

Non-Stationary Time Series and Unit Root Testing

Non-Stationary Time Series and Unit Root Testing Econometrics II Non-Stationary Time Series and Unit Root Testing Morten Nyboe Tabor Course Outline: Non-Stationary Time Series and Unit Root Testing 1 Stationarity and Deviation from Stationarity Trend-Stationarity

More information

Univariate ARIMA Models

Univariate ARIMA Models Univariate ARIMA Models ARIMA Model Building Steps: Identification: Using graphs, statistics, ACFs and PACFs, transformations, etc. to achieve stationary and tentatively identify patterns and model components.

More information

LINEAR STOCHASTIC MODELS

LINEAR STOCHASTIC MODELS LINEAR STOCHASTIC MODELS Let {x τ+1,x τ+2,...,x τ+n } denote n consecutive elements from a stochastic process. If their joint distribution does not depend on τ, regardless of the size of n, then the process

More information

Time Series 2. Robert Almgren. Sept. 21, 2009

Time Series 2. Robert Almgren. Sept. 21, 2009 Time Series 2 Robert Almgren Sept. 21, 2009 This week we will talk about linear time series models: AR, MA, ARMA, ARIMA, etc. First we will talk about theory and after we will talk about fitting the models

More information

Ross Bettinger, Analytical Consultant, Seattle, WA

Ross Bettinger, Analytical Consultant, Seattle, WA ABSTRACT DYNAMIC REGRESSION IN ARIMA MODELING Ross Bettinger, Analytical Consultant, Seattle, WA Box-Jenkins time series models that contain exogenous predictor variables are called dynamic regression

More information

CHAPTER 8 FORECASTING PRACTICE I

CHAPTER 8 FORECASTING PRACTICE I CHAPTER 8 FORECASTING PRACTICE I Sometimes we find time series with mixed AR and MA properties (ACF and PACF) We then can use mixed models: ARMA(p,q) These slides are based on: González-Rivera: Forecasting

More information

6 NONSEASONAL BOX-JENKINS MODELS

6 NONSEASONAL BOX-JENKINS MODELS 6 NONSEASONAL BOX-JENKINS MODELS In this section, we will discuss a class of models for describing time series commonly referred to as Box-Jenkins models. There are two types of Box-Jenkins models, seasonal

More information

Stat 5100 Handout #12.e Notes: ARIMA Models (Unit 7) Key here: after stationary, identify dependence structure (and use for forecasting)

Stat 5100 Handout #12.e Notes: ARIMA Models (Unit 7) Key here: after stationary, identify dependence structure (and use for forecasting) Stat 5100 Handout #12.e Notes: ARIMA Models (Unit 7) Key here: after stationary, identify dependence structure (and use for forecasting) (overshort example) White noise H 0 : Let Z t be the stationary

More information

ARIMA Models. Jamie Monogan. January 25, University of Georgia. Jamie Monogan (UGA) ARIMA Models January 25, / 38

ARIMA Models. Jamie Monogan. January 25, University of Georgia. Jamie Monogan (UGA) ARIMA Models January 25, / 38 ARIMA Models Jamie Monogan University of Georgia January 25, 2012 Jamie Monogan (UGA) ARIMA Models January 25, 2012 1 / 38 Objectives By the end of this meeting, participants should be able to: Describe

More information

Exercises - Time series analysis

Exercises - Time series analysis Descriptive analysis of a time series (1) Estimate the trend of the series of gasoline consumption in Spain using a straight line in the period from 1945 to 1995 and generate forecasts for 24 months. Compare

More information

Forecasting and Estimation

Forecasting and Estimation February 3, 2009 Forecasting I Very frequently the goal of estimating time series is to provide forecasts of future values. This typically means you treat the data di erently than if you were simply tting

More information

STAT Financial Time Series

STAT Financial Time Series STAT 6104 - Financial Time Series Chapter 4 - Estimation in the time Domain Chun Yip Yau (CUHK) STAT 6104:Financial Time Series 1 / 46 Agenda 1 Introduction 2 Moment Estimates 3 Autoregressive Models (AR

More information

Circle the single best answer for each multiple choice question. Your choice should be made clearly.

Circle the single best answer for each multiple choice question. Your choice should be made clearly. TEST #1 STA 4853 March 6, 2017 Name: Please read the following directions. DO NOT TURN THE PAGE UNTIL INSTRUCTED TO DO SO Directions This exam is closed book and closed notes. There are 32 multiple choice

More information

ECON/FIN 250: Forecasting in Finance and Economics: Section 4.1 Forecasting Fundamentals

ECON/FIN 250: Forecasting in Finance and Economics: Section 4.1 Forecasting Fundamentals ECON/FIN 250: Forecasting in Finance and Economics: Section 4.1 Forecasting Fundamentals Patrick Herb Brandeis University Spring 2016 Patrick Herb (Brandeis University) Forecasting Fundamentals ECON/FIN

More information

Non-Stationary Time Series and Unit Root Testing

Non-Stationary Time Series and Unit Root Testing Econometrics II Non-Stationary Time Series and Unit Root Testing Morten Nyboe Tabor Course Outline: Non-Stationary Time Series and Unit Root Testing 1 Stationarity and Deviation from Stationarity Trend-Stationarity

More information

5: MULTIVARATE STATIONARY PROCESSES

5: MULTIVARATE STATIONARY PROCESSES 5: MULTIVARATE STATIONARY PROCESSES 1 1 Some Preliminary Definitions and Concepts Random Vector: A vector X = (X 1,..., X n ) whose components are scalarvalued random variables on the same probability

More information

E 4101/5101 Lecture 6: Spectral analysis

E 4101/5101 Lecture 6: Spectral analysis E 4101/5101 Lecture 6: Spectral analysis Ragnar Nymoen 3 March 2011 References to this lecture Hamilton Ch 6 Lecture note (on web page) For stationary variables/processes there is a close correspondence

More information

Quantitative Finance I

Quantitative Finance I Quantitative Finance I Linear AR and MA Models (Lecture 4) Winter Semester 01/013 by Lukas Vacha * If viewed in.pdf format - for full functionality use Mathematica 7 (or higher) notebook (.nb) version

More information

Forecasting using R. Rob J Hyndman. 2.4 Non-seasonal ARIMA models. Forecasting using R 1

Forecasting using R. Rob J Hyndman. 2.4 Non-seasonal ARIMA models. Forecasting using R 1 Forecasting using R Rob J Hyndman 2.4 Non-seasonal ARIMA models Forecasting using R 1 Outline 1 Autoregressive models 2 Moving average models 3 Non-seasonal ARIMA models 4 Partial autocorrelations 5 Estimation

More information

We use the centered realization z t z in the computation. Also used in computing sample autocovariances and autocorrelations.

We use the centered realization z t z in the computation. Also used in computing sample autocovariances and autocorrelations. Stationary Time Series Models Part 1 MA Models and Their Properties Class notes for Statistics 41: Applied Time Series Ioa State University Copyright 1 W. Q. Meeker. Segment 1 ARMA Notation, Conventions,

More information

A lecture on time series

A lecture on time series A lecture on time series Bernt Arne Ødegaard 15 November 2018 Contents 1 Survey of time series issues 1 1.1 Definition.............................. 2 1.2 Examples.............................. 2 1.3 Typical

More information

ECONOMETRICS Part II PhD LBS

ECONOMETRICS Part II PhD LBS ECONOMETRICS Part II PhD LBS Luca Gambetti UAB, Barcelona GSE February-March 2014 1 Contacts Prof.: Luca Gambetti email: luca.gambetti@uab.es webpage: http://pareto.uab.es/lgambetti/ Description This is

More information

Trend-Cycle Decompositions

Trend-Cycle Decompositions Trend-Cycle Decompositions Eric Zivot April 22, 2005 1 Introduction A convenient way of representing an economic time series y t is through the so-called trend-cycle decomposition y t = TD t + Z t (1)

More information

Circle a single answer for each multiple choice question. Your choice should be made clearly.

Circle a single answer for each multiple choice question. Your choice should be made clearly. TEST #1 STA 4853 March 4, 215 Name: Please read the following directions. DO NOT TURN THE PAGE UNTIL INSTRUCTED TO DO SO Directions This exam is closed book and closed notes. There are 31 questions. Circle

More information

5 Transfer function modelling

5 Transfer function modelling MSc Further Time Series Analysis 5 Transfer function modelling 5.1 The model Consider the construction of a model for a time series (Y t ) whose values are influenced by the earlier values of a series

More information

Time Series Models and Inference. James L. Powell Department of Economics University of California, Berkeley

Time Series Models and Inference. James L. Powell Department of Economics University of California, Berkeley Time Series Models and Inference James L. Powell Department of Economics University of California, Berkeley Overview In contrast to the classical linear regression model, in which the components of the

More information

IDENTIFICATION OF ARMA MODELS

IDENTIFICATION OF ARMA MODELS IDENTIFICATION OF ARMA MODELS A stationary stochastic process can be characterised, equivalently, by its autocovariance function or its partial autocovariance function. It can also be characterised by

More information

TIME SERIES ANALYSIS. Forecasting and Control. Wiley. Fifth Edition GWILYM M. JENKINS GEORGE E. P. BOX GREGORY C. REINSEL GRETA M.

TIME SERIES ANALYSIS. Forecasting and Control. Wiley. Fifth Edition GWILYM M. JENKINS GEORGE E. P. BOX GREGORY C. REINSEL GRETA M. TIME SERIES ANALYSIS Forecasting and Control Fifth Edition GEORGE E. P. BOX GWILYM M. JENKINS GREGORY C. REINSEL GRETA M. LJUNG Wiley CONTENTS PREFACE TO THE FIFTH EDITION PREFACE TO THE FOURTH EDITION

More information

Applied time-series analysis

Applied time-series analysis Robert M. Kunst robert.kunst@univie.ac.at University of Vienna and Institute for Advanced Studies Vienna October 18, 2011 Outline Introduction and overview Econometric Time-Series Analysis In principle,

More information

Chapter 3 - Temporal processes

Chapter 3 - Temporal processes STK4150 - Intro 1 Chapter 3 - Temporal processes Odd Kolbjørnsen and Geir Storvik January 23 2017 STK4150 - Intro 2 Temporal processes Data collected over time Past, present, future, change Temporal aspect

More information

Basics: Definitions and Notation. Stationarity. A More Formal Definition

Basics: Definitions and Notation. Stationarity. A More Formal Definition Basics: Definitions and Notation A Univariate is a sequence of measurements of the same variable collected over (usually regular intervals of) time. Usual assumption in many time series techniques is that

More information

Chapter 1. Basics. 1.1 Definition. A time series (or stochastic process) is a function Xpt, ωq such that for

Chapter 1. Basics. 1.1 Definition. A time series (or stochastic process) is a function Xpt, ωq such that for Chapter 1 Basics 1.1 Definition A time series (or stochastic process) is a function Xpt, ωq such that for each fixed t, Xpt, ωq is a random variable [denoted by X t pωq]. For a fixed ω, Xpt, ωq is simply

More information

TMA4285 December 2015 Time series models, solution.

TMA4285 December 2015 Time series models, solution. Norwegian University of Science and Technology Department of Mathematical Sciences Page of 5 TMA4285 December 205 Time series models, solution. Problem a) (i) The slow decay of the ACF of z t suggest that

More information

Basic concepts and terminology: AR, MA and ARMA processes

Basic concepts and terminology: AR, MA and ARMA processes ECON 5101 ADVANCED ECONOMETRICS TIME SERIES Lecture note no. 1 (EB) Erik Biørn, Department of Economics Version of February 1, 2011 Basic concepts and terminology: AR, MA and ARMA processes This lecture

More information