1. Core inflation model
Start with a simple model of the previous year's inflation representing the core inflation.
p* = 0.468 + 1.197pt-1 - 0.088pt-2 - 0.217pt-3 + 0.048pt-4
(A1)
To test the suitability of model by carrying out tests for autocorrelation, heteroskedasticity and mis-specification.
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 12.552 [0.0280] * and F-Form(5, 102) = 2.5748 [0.0309] *
Testing for Heteroscedastic errors
Chi-squared(8) = 18.501 [0.0178] * and F-Form(8, 98) = 2.4239 [0.0196] *
RESET test for adding Y2
RESET F( 1,106) = 0.34961 [0.5556]
p* = 0.450 + 1.229pt-1 - 0.076pt-2 - 0.191pt-3 - 0.349pt-4 + 0.486pt-5 - 0.140pt-6 + 0.031pt-7 - 0.0448pt-8
(A2)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 29.3 [0.0000] ** and F-Form(5, 94) = 6.9993 [0.0000] **
Testing for Heteroscedastic errors
Chi-squarerd(16) = 23.569 [0.0994] and F-Form(16, 82) = 1.4306 [0.1482]
RESET test for adding Y2
RESET F( 1, 98) = 0.57972 [0.4483]
p* = 0.464 + 1.237pt-1 - 0.293pt-2
(A3)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 15.106 [0.0099] ** and F-Form(5, 100) = 3.2523 [0.0092] **
Testing for Heteroscedastic errors
Chi-squared(4) = 12.077 [0.0168] * and F-Form(4, 100) = 3.1476 [0.0175] *
RESET test for adding Y2
RESET F( 1,104) = 0.11816 [0.7317]
p* = 0.426 + 1.234pt-1 - 0.082pt-2 - 0.206pt-3 - 0.330pt-4 + 0.498pt-5 - 0.166pt-6
(A4)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 7.7057 [0.1732] and F-Form(5, 96) = 1.4752 [0.2051]
Testing for Heteroscedastic errors
Chi-squared(12) = 22.23 [0.0350] * and F-Form(12, 88) = 1.9007 [0.0449] *
RESET test for adding Y2
RESET F( 1,100) = 0.33303 [0.5652]
2. General Model
pt = a + pt* + Si=01 bix1t-i + Si=04 bix2t-i + Si=012 bix3t-i
pt = 0.042849 + 1.00011pt* + 3.6829x1t - 1.5933x1t-1 - 0.00019950x2t + 0.00039301x2t-1 - 0.000041654x2t-2 - 0.0000077147x2t-3 + 0.0000012834x2t-4 + 0.43954x3t - 0.63818x3t-1 - 1.0104x3t-2 - 0.00397x3t-3 + 1.0327x3t-4 + 0.018703x3t-5 + 0.75805x3t-6 + 0.093992x3t-7 - 1.2022x3t-8 + 0.18240x3t-9 + 0.52426x3t-10 - 0.19963x3t-11 + 0.059784x3t-12
(A5)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 7.9896 [0.1568] and F-Form(5, 77) = 1.2815 [0.2805]
Testing for Heteroscedastic errors
Chi-squared(29) = 28.923 [0.4691] and F-Form(29, 52) = 0.69077 [0.8573]
RESET test for adding Y2
RESET F( 1, 81) = 0.20154 [0.6547]
pt = a + pt* + Si=01 bix1t-i + Si=04 bix2t-i + Si=08 bix3t-i
pt = 0.044409 + 0.99629pt* + 0.92594x1t + 0.31964x1t-1 - 0.00018294x2t + 0.00030178x2t-1 - 0.000020155x2t-2 + 0.000022311x2t-3 - 0.0000022703x2t-4 + 0.43243x3t - 0.54418x3t-1 - 0.93395x3t-2 - 0.10061x3t-3 + 1.0996x3t-4 +0.10439x3t-5 + 0.66770x3t-6 +0.099075x3t-7 - 0.76516x3t-8
(A6)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 6.1857 [0.2886] and F-Form(5, 85) = 1.0328 [0.4037]
Testing for Heteroscedastic errors
Chi-squared(25) = 26.754 [0.3683] and F-Form(25, 64) = 0.84299 [0.6749]
RESET test for adding Y2
RESET F( 1, 89) =1.2673e-006 [0.9991]
pt = a + pt* + Si=01 bix1t-i + Si=04 bix2t-i + Si=04 bix3t-i
pt = -0.096080 + 0.99517pt* + 0.039207x1t + 0.94839x1t-1 - 0.000156x2t + 0.000273x2t-1 - 0.000022x2t-2 + 0.00003254x2t-3 - 0.0000266x2t-4 + 0.38992x3t - 0.9017x3t-1 -0.18939x3t-2 - 0.18939x3t-3 + 1.4065x3t-4
(A7)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 4.5131 [0.4781] and F-Form(5, 91) = 0.77866 [0.5676]
Testing for Heteroscedastic errors
Chi-squared(21) = 24.246 [0.2813] and F-Form(21, 74) = 0.99629 [0.4782]
RESET test for adding Y2
RESET F( 1, 95) = 0.037809 [0.8462]
pt = a + pt* + b1x1t + b2x2t + b3x3t
pt = -0.0061309 + 1.0096pt* + 0.48898x1t + 0.0000899x2t - 0.074345x3t
(A8)
Testing for Error Autocorrelation from lags 1 to 5
Chi-squared(5) = 3.5088 [0.6221] and F-Form(5, 100) = 0.65898 [0.6554]
Testing for Heteroscedastic errors
Chi-squared(7) = 19.246 [0.0075] ** and F-Form(7, 97) = 2.9386 [0.0078] **
RESET test for adding Y2
RESET F( 1,104) = 0.29011 [0.5913]
Email: gjs@swann39.freeserve.co.uk