Notice: On March 31, it was **announced** that Statalist is moving from an email list to a **forum**. The old list will shut down on April 23, and its replacement, **statalist.org** is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
San K <devank@gmail.com> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: Autocorrelation doesn't seem to go away. What can I do here? |

Date |
Fri, 30 Mar 2012 17:51:20 +1100 |

Hello, I’m wondering if anyone can tell me what my next step should be or how bad my results are going to be if I ignore the autocorrelation problem and bad Hansen statistics. Basically this modelling is done on a utility data where consumption is metered on a roughly 90 day period. I’m trying to estimate the impact of the price on the consumption. I chose to go with GMM as I have endogeneity problems with the lagged consumption and two tier pricing. I’m using laglimits(2 2) of consumption to handle the first endogeneity. Also using actual RealTier1& RealTier2 as instrument for the weighted average price. I don’t think this causes any concern. RainDeviation, TempDeviation & EvapDeviation measure weather variables as deviation from their mean value for the meter reading period. I don’t think this causes any concern. Summer, autumn, winter & spring are variables measuring how much season is covered by each meter reading period. restrictionsL2 is some sort of restriction on use placed by the government. I have tried many different variations but nothing seems to solve the autocorrelation problem. I have attached the results. Any suggestion I should do? Any help is appreciated. Regards, devank . xtabond2 l(0/1).ConsDayAvgLN waitedAvgPrice waitedAvgPriceL1 waitedAvgPriceL2 waitedAvgPriceL3 RainDeviation TempDeviation EvapDeviation summer autumn winter spring restrictionsL2, noleveleq gmmstyle(ConsDayAvgLN, laglimits(2 2) equation(diff)) gmmstyle(RealTier1 RealTier2, laglimits(0 4) equation(diff) collapse) ivstyle(l(0/0).(RainDeviation TempDeviation EvapDeviation waitedAvgPriceL1 waitedAvgPriceL2 waitedAvgPriceL3), equation(diff)) ivstyle(summer autumn winter spring restrictionsL2 , equation(diff)) twostep ar(9) robust Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm. Dynamic panel-data estimation, two-step difference GMM ------------------------------------------------------------------------------ Group variable: subject Number of obs = 48800 Time variable : period Number of groups = 1952 Number of instruments = 46 Obs per group: min = 25 Wald chi2(13) = 1369.01 avg = 25.00 Prob > chi2 = 0.000 max = 25 ---------------------------------------------------------------------------------- | Corrected ConsDayAvgLN | Coef. Std. Err. z P>|z| [95% Conf. Interval] -----------------+---------------------------------------------------------------- ConsDayAvgLN | L1. | .4040915 .0171865 23.51 0.000 .3704065 .4377764 | waitedAvgPrice | .0011504 .0002147 5.36 0.000 .0007296 .0015712 waitedAvgPriceL1 | -.0006233 .0002843 -2.19 0.028 -.0011804 -.0000661 waitedAvgPriceL2 | -.0005686 .0003135 -1.81 0.070 -.0011831 .0000459 waitedAvgPriceL3 | -.0006479 .0002677 -2.42 0.015 -.0011725 -.0001233 RainDeviation | -.0099555 .0014745 -6.75 0.000 -.0128455 -.0070655 TempDeviation | .0003753 .0020412 0.18 0.854 -.0036253 .0043759 EvapDeviation | .0504452 .0049941 10.10 0.000 .040657 .0602335 summer | -.4569891 .1917348 -2.38 0.017 -.8327825 -.0811957 autumn | -.4950277 .1921709 -2.58 0.010 -.8716757 -.1183796 winter | -.5130355 .1954777 -2.62 0.009 -.8961648 -.1299063 spring | -.4546386 .1939049 -2.34 0.019 -.8346852 -.074592 restrictionsL2 | .0045096 .0045326 0.99 0.320 -.0043742 .0133933 ---------------------------------------------------------------------------------- Instruments for first differences equation Standard D.(RainDeviation TempDeviation EvapDeviation waitedAvgPriceL1 waitedAvgPriceL2 waitedAvgPriceL3) D.(summer autumn winter spring restrictionsL2) GMM-type (missing=0, separate instruments for each period unless collapsed) L2.ConsDayAvgLN L(0/4).(RealTier1 RealTier2) collapsed ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -26.06 Pr > z = 0.000 Arellano-Bond test for AR(2) in first differences: z = -7.07 Pr > z = 0.000 Arellano-Bond test for AR(3) in first differences: z = -2.38 Pr > z = 0.018 Arellano-Bond test for AR(4) in first differences: z = 13.79 Pr > z = 0.000 Arellano-Bond test for AR(5) in first differences: z = -4.87 Pr > z = 0.000 Arellano-Bond test for AR(6) in first differences: z = -8.13 Pr > z = 0.000 Arellano-Bond test for AR(7) in first differences: z = -2.26 Pr > z = 0.024 Arellano-Bond test for AR(8) in first differences: z = 11.77 Pr > z = 0.000 Arellano-Bond test for AR(9) in first differences: z = -2.04 Pr > z = 0.042 ------------------------------------------------------------------------------ Sargan test of overid. restrictions: chi2(33) =1045.99 Prob > chi2 = 0.000 (Not robust, but not weakened by many instruments.) Hansen test of overid. restrictions: chi2(33) = 283.74 Prob > chi2 = 0.000 (Robust, but can be weakened by many instruments.) Difference-in-Hansen tests of exogeneity of instrument subsets: gmm(ConsDayAvgLN, eq(diff) lag(2 2)) Hansen test excluding group: chi2(8) = 88.43 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(25) = 195.31 Prob > chi2 = 0.000 gmm(RealTier1 RealTier2, collapse eq(diff) lag(0 4)) Hansen test excluding group: chi2(23) = 175.38 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(10) = 108.36 Prob > chi2 = 0.000 iv(RainDeviation TempDeviation EvapDeviation waitedAvgPriceL1 waitedAvgPriceL2 waitedAvgPriceL3, eq(diff)) Hansen test excluding group: chi2(27) = 133.54 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(6) = 150.20 Prob > chi2 = 0.000 iv(summer autumn winter spring restrictionsL2, eq(diff)) Hansen test excluding group: chi2(28) = 264.04 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(5) = 19.71 Prob > chi2 = 0.001 * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**Re: st: Drop if beyond certain percentile** - Next by Date:
**st: gformula** - Previous by thread:
**st: Drop if beyond certain percentile** - Next by thread:
**st: gformula** - Index(es):