Analysis

[1] "景気ウォッチャー調査:九州:季節調整値:景気の現状判断(方向性)DI:内閣府"
      Jan  Feb  Mar  Apr  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec
2002 34.7 36.0 43.0 44.5 47.6 40.9 39.3 39.9 40.2 36.2 38.6 37.5
2003 37.7 38.3 40.1 37.3 39.3 41.4 44.7 47.1 49.3 51.9 47.4 48.9
2004 50.8 49.0 48.6 52.5 50.4 50.3 51.4 46.3 44.9 45.4 45.8 46.0
2005 49.5 48.3 46.8 47.7 50.3 52.5 51.6 53.1 52.3 52.2 54.6 60.5
2006 55.1 58.1 55.7 50.8 50.2 47.0 46.1 47.4 51.3 51.5 50.8 49.9
2007 49.6 48.1 46.5 46.7 45.0 44.7 43.7 43.4 40.4 43.1 40.7 40.7
2008 36.0 34.4 33.2 31.8 30.0 28.1 27.2 25.1 28.2 20.2 21.9 18.2
2009 20.8 20.9 25.8 28.9 34.6 40.1 37.5 41.1 43.2 41.1 34.9 41.4
2010 43.8 44.7 44.2 47.2 44.4 46.0 46.0 44.5 44.2 48.1 46.9 47.6
2011 48.0 50.1 28.4 30.0 35.6 46.7 48.4 47.7 48.0 50.6 49.6 51.4
2012 44.3 46.8 47.8 49.1 45.2 42.8 41.3 44.5 43.6 42.4 45.2 47.4
2013 50.9 53.7 53.0 52.7 54.8 53.0 52.6 53.3 55.6 55.3 58.0 56.3
2014 56.6 53.5 54.6 37.4 43.7 50.6 49.2 47.7 49.8 47.8 46.4 47.8
2015 45.5 51.1 52.6 52.6 51.1 51.7 51.5 50.2 48.9 50.4 46.3 48.5
2016 47.5 47.3 45.1 31.3 38.7 41.9 47.5 50.1 46.1 50.4 52.3 51.7
2017 51.0 50.0 47.5 49.6 49.1 51.4 48.4 49.7 51.3 50.8 52.9 55.6
2018 49.9 48.9 50.0 50.0 47.8 46.1 45.7 47.1 50.5 48.6 47.7 45.7
2019 43.9 48.0 44.9 44.4 41.8 44.8 42.3 40.0 45.5 35.4          
  • 民主党政権


Call:
lm(formula = value ~ ID)

Residuals:
    Min      1Q  Median      3Q     Max 
-16.049  -1.581   1.229   3.228   6.213 

Coefficients:
            Estimate Std. Error t value            Pr(>|t|)    
(Intercept) 42.97274    1.63806  26.234 <0.0000000000000002 ***
ID           0.08200    0.07138   1.149               0.258    
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 5.017 on 37 degrees of freedom
Multiple R-squared:  0.03444,   Adjusted R-squared:  0.008348 
F-statistic:  1.32 on 1 and 37 DF,  p-value: 0.258



    Two-sample Kolmogorov-Smirnov test

data:  lm_residuals and rnorm(n = length(lm_residuals), mean = 0, sd = sd(lm_residuals))
D = 0.17949, p-value = 0.5622
alternative hypothesis: two-sided



    Durbin-Watson test

data:  value ~ ID
DW = 0.94544, p-value = 0.00007803
alternative hypothesis: true autocorrelation is greater than 0



    studentized Breusch-Pagan test

data:  value ~ ID
BP = 0.098468, df = 1, p-value = 0.7537



    Box-Ljung test

data:  lm_residuals
X-squared = 11.572, df = 1, p-value = 0.0006694
  • 第二次安倍内閣~


Call:
lm(formula = value ~ ID)

Residuals:
     Min       1Q   Median       3Q      Max 
-17.6326  -1.7066   0.8595   2.4784   8.5490 

Coefficients:
            Estimate Std. Error t value             Pr(>|t|)    
(Intercept) 52.69575    0.93422  56.406 < 0.0000000000000002 ***
ID          -0.09408    0.01955  -4.811           0.00000695 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 4.191 on 80 degrees of freedom
Multiple R-squared:  0.2244,    Adjusted R-squared:  0.2147 
F-statistic: 23.15 on 1 and 80 DF,  p-value: 0.000006951



    Two-sample Kolmogorov-Smirnov test

data:  lm_residuals and rnorm(n = length(lm_residuals), mean = 0, sd = sd(lm_residuals))
D = 0.12195, p-value = 0.5785
alternative hypothesis: two-sided



    Durbin-Watson test

data:  value ~ ID
DW = 0.81426, p-value = 0.0000000005199
alternative hypothesis: true autocorrelation is greater than 0



    studentized Breusch-Pagan test

data:  value ~ ID
BP = 0.06469, df = 1, p-value = 0.7992



    Box-Ljung test

data:  lm_residuals
X-squared = 26.589, df = 1, p-value = 0.0000002516
  • 白川日銀総裁


Call:
lm(formula = value ~ ID)

Residuals:
    Min      1Q  Median      3Q     Max 
-14.162  -4.870   1.169   5.764   9.182 

Coefficients:
            Estimate Std. Error t value             Pr(>|t|)    
(Intercept) 28.10544    1.71234  16.414 < 0.0000000000000002 ***
ID           0.41304    0.04964   8.321      0.0000000000202 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 6.493 on 57 degrees of freedom
Multiple R-squared:  0.5485,    Adjusted R-squared:  0.5406 
F-statistic: 69.24 on 1 and 57 DF,  p-value: 0.00000000002019



    Two-sample Kolmogorov-Smirnov test

data:  lm_residuals and rnorm(n = length(lm_residuals), mean = 0, sd = sd(lm_residuals))
D = 0.16949, p-value = 0.3674
alternative hypothesis: two-sided



    Durbin-Watson test

data:  value ~ ID
DW = 0.46748, p-value = 0.00000000000005572
alternative hypothesis: true autocorrelation is greater than 0



    studentized Breusch-Pagan test

data:  value ~ ID
BP = 2.8961, df = 1, p-value = 0.08879



    Box-Ljung test

data:  lm_residuals
X-squared = 36.385, df = 1, p-value = 0.000000001619
  • 黒田日銀総裁~


Call:
lm(formula = value ~ ID)

Residuals:
     Min       1Q   Median       3Q      Max 
-17.6315  -1.7102   0.8705   2.5150   8.5496 

Coefficients:
            Estimate Std. Error t value             Pr(>|t|)    
(Intercept) 52.41168    0.96888   54.09 < 0.0000000000000002 ***
ID          -0.09406    0.02104   -4.47            0.0000265 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 4.265 on 77 degrees of freedom
Multiple R-squared:  0.206, Adjusted R-squared:  0.1957 
F-statistic: 19.98 on 1 and 77 DF,  p-value: 0.00002654



    Two-sample Kolmogorov-Smirnov test

data:  lm_residuals and rnorm(n = length(lm_residuals), mean = 0, sd = sd(lm_residuals))
D = 0.13924, p-value = 0.4302
alternative hypothesis: two-sided



    Durbin-Watson test

data:  value ~ ID
DW = 0.8107, p-value = 0.0000000008805
alternative hypothesis: true autocorrelation is greater than 0



    studentized Breusch-Pagan test

data:  value ~ ID
BP = 0.2344, df = 1, p-value = 0.6283



    Box-Ljung test

data:  lm_residuals
X-squared = 25.895, df = 1, p-value = 0.0000003604
  • 特記その他
  1. 時系列データの特徴(誤差構造、負数の有無その他等)に関わらず線形回帰を求めている。よってあくまでも対象とした期間における線形回帰そしてその残差の傾向を確認しているのみであり結果の外挿は出来ない。
  2. 民主党政権:2009-09-16~2012-12-25
  3. 白川方明氏の日銀総裁就任期間:2008-04-09~2013-03-19