RDP 2001-09: What do Sentiment Surveys Measure? Appendix B: Additional Results

Ivan Roberts and John Simon

November 2001

Table B1: Roy Morgan Consumer Confidence Rating Component Indices
Regressor Index 1 Index 2 Index 3 Index 4 Index 5
  1.72
(0.73)**		 – – – 1.76
(0.81)**
ΔGDPt−2 2.80
(0.82)***		 – – – –
ΔGDPt−3 1.82
(0.74)***		 – – – –
ΔJVt−2 – – – – 0.46
(0.15)*
ΔHWt−2 3.56
(1.57)**		 3.44
(1.29)***		 – – –
ΔHWt−2 2.73
(1.56)*		 – – – –
ΔAOt – – 0.40
(0.18)**		 – 0.20
(0.09)**
ΔAOt−1 0.19
(0.08)**		 0.14
(0.07)*		 0.45
(0.19)**		 – –
ΔAOt−2 – 0.18
(0.08)**		 0.59
(0.19)***		 – –
ΔAOt−3 0.15
(0.08)*		 0.17
(0.08)**		 0.52
(0.18)***		 – –
ΔAOt−4 – – – – −0.27
(0.09)***
ΔERt – – – – 0.32
(0.16)**
ΔERt−3 – −0.25
(0.13)*		 – – –
CASHt −1.35
(0.37)***		 – – −1.96
(0.57)***		 −1.71
(0.51)***
CASHt−1 – −1.43
(0.26)***		 – −1.56
(0.58)***		 –
CASHt−4 – – −2.96
(0.76)***		 – –
Rho 0.66
(0.09)***		 0.53
(0.09)***		 0.62
(0.08)***		 0.55
(0.09)***		 0.73
(0.08)***
Constant 4.79
(4.35)		 12.31
(2.95)***		 26.86
(8.33)***		 34.58
(4.87)***		 15.29
(5.93)***
Observations 86 87 102 103 88
R2 0.78 0.72 0.63 0.76 0.72
Relevant R2(a) 0.54 0.56 0.46 0.61 0.51
Q (1–8) p-value 0.34 0.14 0.57 0.91 0.98

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) See Section 3.3

Table B2: Westpac-Melbourne Institute Index of Consumer Sentiment Component Indices
Regressor Index 1 Index 2 Index 3 Index 4 Index 5
ΔGDPt−1 1.24
(0.59)**		 1.48
(0.71)**		 3.02
(1.48)**		 – 1.36
(0.64)**
AGDPt−3 – – 2.65
(1.49)*		 – –
ΔJVt−2 – 0.21
(0.24)*		 – – –
ΔHWt−1 – – – −3.75
(2.06)*		 –
ΔHWt−2 3.58
(1.33)***		 – 8.30
(3.44)**		 – –
ΔHWt−3 – – – −4.66
(2.11)**		 –
ΔAOt – – 0.51
(0.20)**		 – –
ΔAOt−1 – – 0.36
(0.18)**		 – 0.18
(0.08)**
ΔAOt−4 −0.17
(0.07)**		 – −0.47
(0.18)**		 – −0.25
(0.08)***
ΔERt−3 −0.28
(0.12)**		 −0.31
(0.14)**		 – −0.44
(0.19)**		 –
ΔERt−4 – – 1.08
(0.35)***		 – –
CASHt −1.11
(0.51)**		 – −2.74
(0.92)***		 – −1.71
(0.62)***
CASH t−1 −0.96
(0.50) *		 −0.81
(0.24)***		 – – –
CASH t−2 – – – – −1.13
(0.62)*
CASH t−3 – – – −1.95
(0.50)***		 –
Rho 0.90
(0.07)***		 0.44
(0.10)***		 0.72
(0.08)***		 0.62
(0.09)***		 0.83
(0.07)***
Constant 17.13
(8.58)**		 6.35
(2.79)**		 19.33
(10.49)*		 23.52
(5.93)***		 25.45
(8.54)***
Observations 87 88 87 86 104
R2 0.78 0.53 0.67 0.57 0.74
Relevant R2(a) 0.61 0.39 0.41 0.54 0.62
Q (1−8) p-value 0.22 0.81 0.60 0.32 0.21

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) See Section 3.3

Table B3: NAB Quarterly Survey Actual Business Conditions
Regressor Trading Profitability Employment
ΔGDPt−1 3.32
(1.51)**		 3.27
(1.55)**		 –
ΔGDPt−2 5.23
(1.57)***		 3.82
(1.61)**		 –
ΔGDPt−3 5.23
(1.99)**		 7.30
(1.40)***		 3.87
(0.92)***
ΔGDPt−4 3.16
(1.50)**		 5.30
(1.43)***		 1.92
(0.83)**
ΔJVt−1 – – 0.36
(0.11)***
ΔJVt−2 – 0.67
(0.20) ***		 0.50
(0.12) ***
ΔHWt−1 6.26
(2.52)**		 – –
ΔAOt−2 0.65
(0.17)***		 0.46
(0.17)***		 –
ΔAOt−3 0.32
(0.16)*		 – 0.24
(0.10)**
ΔERt−1 – – 0.28
(0.15)*
ΔERt−2 – – 0.39
(0.16)**
ΔERt−3 0.57
(0.28)*		 – –
ΔERt−4 0.85
(0.30)***		 0.82
(0.28)***		 0.48
(0.16)***
CASHt−1 3.64
(2.08)*		 – –
CASHt−2 −4.54
(2.10)**		 – –
CASHt−3 – – 4.60
(0.91)***
CASHt−4 – – −5.85
(0.92)***
Rho 0.54
(0.16)***		 0.34
(0.18)*		 0.22
(0.18)
Constant −9.40
(6.09)		 −16.69
(3.33)***		 7.34
(2.55)***
Observations 47 47 47
R2 0.92 0.87 0.96
Relevant R2(a) 0.82 0.76 0.94
Q (1−8) p-value 0.11 0.18 0.11

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) See Section 3.3

Table B4: NAB Quarterly Survey
Expected Business Conditions (3-month Outlook)
Regressor Average index Trading Profitability Employment
ΔGDPt−2 3.81
(1.32)***		 3.40
(1.66)**		 – 3.10
(0.88)***
ΔGDPt−3 3.70
(1.25)***		 4.60
(1.45)***		 3.06
(1.14)***		 –
ΔJVt−1 0.89
(0.17)***		 0.86
(0.20)***		 0.79
(0.17)***		 –
ΔJVt−2 0.48
(0.16)***		 0.67
(0.22)***		 0.69
(0.18)***		 –
ΔJVt−3 −0.47
(0.22)**		 – – –
ΔHWt−1 – – – 4.26
(1.34)***
ΔAOt−1 – – – 0.22
(0.11)**
ΔAOt−2 – – – 0.36
(0.11)***
ΔAOt−4 0.37
(0.12)***		 0.32
(0.15)**		 0.34
(0.13)**		 0.24
(0.10)**
ΔERt−2 0.46
(0.21)**		 – – 0.37
(0.18)**
ΔERt−3 0.45
(0.20) **		 0.49
(0.26)*		 – 0.67
(0.17)***
CASHt−1 – – – 3.31
(1.20)***
CASHt−2 – – – −4.72
(1.17)***
CASHt−3 6.96
(1.46)***		 5.66
(1.85)***		 6.70
(1.55)***		 –
CASHt−4 −7.90
(1.52)***		 −6.19
(1.82)***		 −7.16
(1.51)***		 –
Rho 0.31
(0.17)*		 0.49
(0.15)***		 0.52
(0.14)***		 0.48
(0.17)***
Constant 3.73
(3.71)		 −0.13
(5.66)		 3.54
(4.56)		 7.43
(2.60)***
Observations 47 47 47 47
R2 0.94 0.92 0.91 0.92
Relevant R2(a) 0.91 0.86 0.84 0.81
Q (1−8) p-value 0.19 0.30 0.12 0.94

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) See Section 3.3

Table B5: NAB Quarterly Survey
Expected Business Conditions (12-month Outlook)
Regressor Average index Trading Profitability Employment
ΔJVt−1 – – 0.64
(0.15)***		 –
ΔHWt−4 3.03
(1.63) *		 – – –
ΔAOt−1 0.23
(0.11)**		 0.36
(0.13)***		 – –
ΔAOt−2 0.29
(0.13)**		 – – 0.29
(0.13)**
ΔAOt−3 0.39
(0.12)***		 0.29
(0.13)**		 – 0.34
(0.13)**
ΔAOt−4 0.26
(0.11)***		 – – 0.42
(0.12)***
CASHt – 5.86
(1.97)***		 – –
CASHt−1 −2.56
(0.75)***		 −6.87
(1.92)***		 −1.31
(0.64)**		 –
CASHt−4 – – – −1.54
(0.44)***
Rho 0.85
(0.10)***		 0.77
(0.10)***		 0.78
(0.10)***		 0.65
(0.13)***
Constant 14.26
(7.17)*		 8.80
(6.52)		 11.15
(6.14)*		 11.17
(4.00)***
Observations 44 47 47 44
R2 0.90 0.89 0.87 0.89
Relevant R2(a) 0.60 0.55 0.50 0.72
Q (1−8) p-value 0.42 0.52 0.12 0.45

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) See Section 3.3

Table B6: Bivariate OLS Regressions – NAB Quarterly Survey
Actual Business Conditions
Regressors(a) ΔGDP ΔEmployment ΔGOS
p-values
Average index 0.209 0.000*** 0.349
Residuals 0.693 0.003*** 0.827
Trading conditions 0.289 0.000*** 0.481
Residuals 0.838 0.086* 0.993
Profitability 0.214 0.000*** 0.453
Residuals 0.990 0.217 0.968
Employment 0.215 0.000*** 0.138
Residuals 0.069* 0.923 0.144

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) In each regression the regressors are four lags of the dependent variable and four lags of the index or residuals in the far left-hand column. Each p-value summarises an F-test for the joint significance of the lagged index or residuals.
Table B7: Bivariate OLS Regressions – NAB Quarterly Survey
Expected Business Conditions (3-month Outlook)
Regressors(a) ΔGDP ΔEmployment ΔGOS
p-values
Average index 0.047** 0.000*** 0.344
Residuals 0.416 0.672 0.991
Trading conditions 0.094* 0.000*** 0.382
Residuals 0.561 0.543 0.953
Profitability 0.040** 0.000*** 0.385
Residuals 0.304 0.481 0.896
Employment 0.015** 0.002*** 0.216
Residuals 0.424 0.124 0.660

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) In each regression the regressors are four lags of the dependent variable and four lags of the index or residuals in the far left-hand column. Each p-value summarises an F-test for the joint significance of the lagged index or residuals.
Table B8: Bivariate OLS Regressions – NAB Quarterly Survey
Expected Business Conditions (12-month Outlook)
Regressors(a) ΔGDP ΔEmployment ΔGOS
p-values
Average index 0.210 0.020** 0.434
Residuals 0.549 0.174 0.761
Trading conditions 0.029** 0.004*** 0.320
Residuals 0.548 0.699 0.993
Profitability 0.188 0.004*** 0.271
Residuals 0.915 0.270 0.787
Employment 0.038** 0.012** 0.136
Residuals 0.267 0.105 0.813
Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) In each regression the regressors are four lags of the dependent variable and four lags of the index or residuals in the far left-hand column. Each p-value summarises an F-test for the joint significance of the lagged index or residuals
Table B9: Probability of a Recession – Roy Morgan Consumer Confidence
P(recessiont=1)=F(α0 + α1xt−k)
k=lagged quarters
xt variables 1 2 3 4 5 6 7 8
Average index
Pseudo R2(a)
t−stat		 0.160
−3.25***		 0.238
−3.44***		 0.191
−3.34***		 0.110
−2.87***		 0.023
−1.51		 0.012
−1.11		 0.020
−1.37		 0.012
−1.09
Residuals
Pseudo R2
t−stat		 0.016
1.14		 0.031
−0.51		 0.014
−1.05		 0.046
−1.85*		 0.002
0.35		 0.008
0.81		 0.002
−0.44		 0.040
−1.70*
Index 1
Pseudo R2
t−stat		 0.150
−2.96***		 0.142
−2.93***		 0.117
−2.83***		 0.048
−2.09**		 0.000
−0.13		 0.000
−0.15		 0.003
−0.54		 0.000
−0.19
Residuals
Pseudo R2
t−stat		 0.013
1.03		 0.000
−0.04		 0.015
−1.09		 0.037
−1.68*		 0.014
1.03		 0.007
0.72		 0.000
−0.20		 0.006
−0.70
Index 2
Pseudo R2
t−stat		 0.13
−2.90***		 0.28
−3.26***		 0.235
−3.19***		 0.084
−2.57***		 0.004
−0.68		 0.006
−0.75		 0.019
−1.33		 0.012
−1.09
Residuals
Pseudo R2
t−stat		 0.013
1.04		 0.015
−1.11		 0.060
−2.11**		 0.051
−1.95*		 0.004
0.59		 0.007
0.75		 0.004
−0.58		 0.053
−1.94*
Index 3
Pseudo R2
t−stat		 0.193
−3.20***		 0.280
−3.27***		 0.186
−3.19***		 0.098
−2.66***		 0.026
−1.57		 0.012
−1.08		 0.011
−1.03		 0.008
−0.88
Residuals
Pseudo R2
t−stat		 0.002
0.44		 0.008
−0.90		 0.009
−0.93		 0.015
−1.17		 0.000
−0.09		 0.000
0.039		 0.002
−0.43		 0.040
−1.84*
Index 4
Pseudo R2
t−stat		 0.072
−2.51**		 0.196
−3.06***		 0.120
−3.07***		 0.168
−3.03***		 0.082
−2.53**		 0.064
−2.31**		 0.053
−2.13**		 0.054
−2.14**
Residuals
Pseudo R2
t−stat		 0.016
1.23		 0.015
−1.20		 0.026
−1.56		 0.062
−2.27**		 0.002
−0.45		 0.003
−0.57		 0.005
−0.66		 0.016
−1.20
Index 5
Pseudo R2
t−stat		 0.053
−2.31**		 0.042
−2.07**		 0.030
−1.74*		 0.015
−1.24		 0.000
−0.24		 0.001
0.29		 0.004
−0.63		 0.000
0.03
Residuals
Pseudo R2
t−stat		 0.002
0.40		 0.005
0.68		 0.000
−0.01		 0.000
−0.22		 0.017
1.16		 0.064
2.10**		 0.000
−0.01		 0.006
−0.71

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) Proposed by Estrella (1998)
Table B10: Probability of a Recession – Westpac-Melbourne Institute Consumer Confidence
P(recessiont=1)=F(α0 + α1xt−k)
K=lagged quarters
xt variables 1 2 3 4 5 6 7 8
Average index
Pseudo R2(a)
t−stat		 0.141
−3.23***		 0.202
−3.38***		 0.144
−3.20***		 0.095
−2.81***		 0.031
−1.75*		 0.008
−0.91		 0.004
−0.60		 0.001
−0.32
Residuals
Pseudo R2
t−stat		 0.012
0.99		 0.010
−0.92		 0.015
−1.09		 0.023
−1.36		 0.002
−0.38		 0.000
−0.01		 0.001
0.29		 0.003
−0.48
Index 1
Pseudo R2
t−stat		 0.108
−2.96***		 0.098
−2.85***		 0.073
−2.54**		 0.038
−1.93*		 0.006
−0.79		 0.000
−0.22		 0.001
0.25		 0.004
0.65
Residuals
Pseudo R2
t−stat		 0.003
−0.47		 0.006
−0.70		 0.018
−1.19		 0.028
−1.48		 0.002
−0.40		 0.001
−0.33		 0.003
0.50		 0.007
0.76
Index 2
Pseudo R2
t−stat		 0.15
−3.28***		 0.317
−3.32***		 0.237
−3.33***		 0.152
−3.17***		 0.030
−1.72*		 0.008
−0.88		 0.004
−0.64		 0.001
−0.33
Residuals
Pseudo R2
t−stat		 0.000
−0.08		 0.052
−2.04**		 0.077
−2.43**		 0.092
−2.59***		 0.009
−0.87		 0.000
−0.20		 0.000
−0.16		 0.000
−0.02
Index 3
Pseudo R2
t−stat		 0.208
−3.26***		 0.287
−3.11***		 0.174
−2.37***		 0.100
−2.68***		 0.031
−1.68*		 0.010
−0.98		 0.005
−0.69		 0.003
−0.55
Residuals
Pseudo R2
t−stat		 0.000
−0.17		 0.035
−1.65*		 0.015
−1.09		 0.008
−0.80		 0.000
−0.26		 0.000
0.02		 0.000
−0.11		 0.002
−0.35
Index 4
Pseudo R2
t−stat		 0.046
−2.08**		 0.16
−3.04***		 0.139
−2.98***		 0.134
−2.94***		 0.063
−2.28**		 0.032
−1.72*		 0.014
−1.18		 0.018
−1.30
Residuals
Pseudo R2
t−stat		 0.004
0.57		 0.027
−1.47		 0.016
−1.12		 0.042
−1.78*		 0.015
−1.09		 0.001
−0.34		 0.003
0.47		 0.000
−0.072
Index 5
Pseudo R2
t−stat		 0.055
−2.35**		 0.053
−2.28**		 0.032
−1.81*		 0.018
−1.37		 0.008
−0.91		 0.000
−0.03		 0.001
−0.24		 0.001
0.33
Residuals
Pseudo R2
t−stat		 0.000
−0.19		 0.007
−0.87		 0.000
−0.21		 0.000
−0.03		 0.002
−0.45		 0.021
1.36		 0.001
0.23		 0.004
−0.66

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) Proposed by Estrella (1998)
Table B11: Bivariate OLS Regressions – Roy Morgan
Regressors(a) ΔGDP ΔEmployment ΔHousehold spending ΔRetail trade
Joint significance tests (p-values)
Average index 0.016** 0.012** 0.020** 0.018**
Residuals 0.527 0.010** 0.750 0.162
Index 1 0.177 0.039** 0.035** 0.048**
Residuals 0.545 0.042** 0.239 0.345
Index 2 0.027** 0.085* 0.023** 0.075*
Residuals 0.770 0.058* 0.433 0.104
Index 3 0.000*** 0.003*** 0.005*** 0.020**
Residuals 0.234 0.050* 0.436 0.280
Index 4 0.155 0.095* 0.479 0.222
Residuals 0.105 0.224 0.934 0.647
Index 5 0.394 0.151 0.020** 0.002***
Residuals 0.807 0.695 0.279 0.104
Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a)In each regression the regressors are four lags of the dependent variable and four lags of the index or residuals in the far left-hand column. Each p-value summarises an F-test for the joint significance of the lagged index or residuals.
Table B12: Bivariate OLS Regressions – Westpac-Melbourne Institute
Regressors(a) ΔGDP ΔEmployment ΔHousehold spending ΔRetail trade
Joint significance tests (p-values)
Average index 0.008*** 0.017** 0.072* 0.042**
Residuals 0.318 0.010** 0.089* 0.229
Index 1 0.282 0.099* 0.242 0.200
Residuals 0.741 0.135 0.685 0.610
Index 2 0.000*** 0.037** 0.027** 0.125
Residuals 0.013** 0.015** 0.251 0.429
Index 3 0.000*** 0.002*** 0.016** 0.039**
Residuals 0.606 0.188 0.637 0.658
Index 4 0.175 0.139 0.664 0.081*
Residuals 0.421 0.026** 0.134 0.692
Index 5 0.162 0.153 0.034** 0.024**
Residuals 0.595 0.237 0.644 0.585

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) In each regression the regressors are four lags of the dependent variable and four lags of the index or residuals in the far left-hand column. Each p-value summarises an F-test for the joint significance of the lagged index or residuals.
Table B13: Employment Equation
Augmented by filtered consumer confidence indices
Filtered index (residual series) Roy Morgan Westpac-Melbourne institute
Joint significance tests (p-values)(a)
Average index 0.077* 0.204
Index 1 0.019** 0.098*
Index 2 0.022** 0.424
Index 3 0.376 0.596
Index 4 0.162 0.322
Index 5 0.959 0.461

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels.
(a) In each regression four lags of the residuals from our earlier regressions of sentiment on economic variables are added to a baseline error-correction model of full-time equivalent employment. The baseline equation on its own explains close to 70 per cent of variation in the dependent variable over the sample period. Each p-value summarises an F-test for the joint significance of the lagged residuals.
Table B14: Employment Equation
Augmented by filtered NAB business conditions indices
  Actual business conditions
 		 Expected business conditions (3-month outlook)
 		 Expected business conditions (12-month outlook)
 
Filtered index
(Residual series)		 Joint significance tests (p-values)(a)
 
Average index 0.078* 0.274 0.610
Trading 0.144 0.196 0.786
Profitability 0.170 0.265 0.740
Employment 0.912 0.060* 0.032**

Notes: *, **, *** denote significance at the 10%, 5% and 1% levels
(a) In each regression four lags of the residuals from our earlier regressions of sentiment on economic variables are added to a baseline error-correction model of full-time equivalent employment. The baseline equation on its own explains nearly 80 per cent of variation in the dependent variable over the sample period. Each p-value summarises an F-test for the joint significance of the lagged residuals.