博学大鲤鱼 发表于 2013-12-3 16:12 
您好,我看了那个帖子了,但是困扰我的问题还是没有解决。我现在没搞明白的是我之前问的两个公式在工具变 ...
可能我理解的不是很对,我认为xtdpdsys的
maxldep(#) maximum lags of dependent variable for use as instruments
这个选项有点误导。比如看下面的例子,执行结果是一样的。
- . webuse abdata
- .
- . xtdpdsys n w k yr1980-yr1984, lags(1) maxldep(2)
- System dynamic panel-data estimation Number of obs = 891
- Group variable: id Number of groups = 140
- Time variable: year
- Obs per group: min = 6
- avg = 6.364286
- max = 8
- Number of instruments = 28 Wald chi2(8) = 5005.11
- Prob > chi2 = 0.0000
- One-step results
- ------------------------------------------------------------------------------
- n | Coef. Std. Err. z P>|z| [95% Conf. Interval]
- -------------+----------------------------------------------------------------
- n |
- L1. | .6004169 .0488289 12.30 0.000 .504714 .6961197
- |
- w | -.5614975 .057 -9.85 0.000 -.6732154 -.4497796
- k | .275974 .0259597 10.63 0.000 .2250939 .326854
- yr1980 | -.0407179 .0093706 -4.35 0.000 -.059084 -.0223518
- yr1981 | -.0891385 .0097512 -9.14 0.000 -.1082505 -.0700265
- yr1982 | -.0604416 .0105754 -5.72 0.000 -.081169 -.0397143
- yr1983 | -.0155483 .0146597 -1.06 0.289 -.0442807 .0131841
- yr1984 | .0059419 .0227527 0.26 0.794 -.0386526 .0505363
- _cons | 2.305954 .1896099 12.16 0.000 1.934325 2.677582
- ------------------------------------------------------------------------------
- Instruments for differenced equation
- GMM-type: L(2/3).n
- Standard: D.w D.k D.yr1980 D.yr1981 D.yr1982 D.yr1983 D.yr1984
- Instruments for level equation
- GMM-type: LD.n
- Standard: _cons
- . xtabond2 n l.n w k yr1980-yr1984, gmm(n,lag(2 3)) iv(yr1980-yr1984 w k, eq(dif)) h(2)
- Favoring speed over space. To switch, type or click on mata: mata set matafavor space, perm.
- Dynamic panel-data estimation, one-step system GMM
- ------------------------------------------------------------------------------
- Group variable: id Number of obs = 891
- Time variable : year Number of groups = 140
- Number of instruments = 28 Obs per group: min = 6
- Wald chi2(8) = 5065.82 avg = 6.36
- Prob > chi2 = 0.000 max = 8
- ------------------------------------------------------------------------------
- n | Coef. Std. Err. z P>|z| [95% Conf. Interval]
- -------------+----------------------------------------------------------------
- n |
- L1. | .6004169 .0485354 12.37 0.000 .5052892 .6955445
- |
- w | -.5614975 .0566574 -9.91 0.000 -.6725439 -.450451
- k | .275974 .0258037 10.70 0.000 .2253997 .3265482
- yr1980 | -.0407179 .0093143 -4.37 0.000 -.0589736 -.0224622
- yr1981 | -.0891385 .0096926 -9.20 0.000 -.1081356 -.0701413
- yr1982 | -.0604416 .0105118 -5.75 0.000 -.0810444 -.0398388
- yr1983 | -.0155483 .0145716 -1.07 0.286 -.044108 .0130114
- yr1984 | .0059419 .0226159 0.26 0.793 -.0383846 .0502683
- _cons | 2.305954 .1884703 12.24 0.000 1.936559 2.675349
- ------------------------------------------------------------------------------
- Instruments for first differences equation
- Standard
- D.(yr1980 yr1981 yr1982 yr1983 yr1984 w k)
- GMM-type (missing=0, separate instruments for each period unless collapsed)
- L(2/3).n
- Instruments for levels equation
- Standard
- _cons
- GMM-type (missing=0, separate instruments for each period unless collapsed)
- DL.n
- ------------------------------------------------------------------------------
- Arellano-Bond test for AR(1) in first differences: z = -6.28 Pr > z = 0.000
- Arellano-Bond test for AR(2) in first differences: z = -0.79 Pr > z = 0.427
- ------------------------------------------------------------------------------
- Sargan test of overid. restrictions: chi2(19) = 72.51 Prob > chi2 = 0.000
- (Not robust, but not weakened by many instruments.)
- Difference-in-Sargan tests of exogeneity of instrument subsets:
- GMM instruments for levels
- Sargan test excluding group: chi2(12) = 30.65 Prob > chi2 = 0.002
- Difference (null H = exogenous): chi2(7) = 41.86 Prob > chi2 = 0.000
- iv(yr1980 yr1981 yr1982 yr1983 yr1984 w k, eq(diff))
- Sargan test excluding group: chi2(12) = 37.86 Prob > chi2 = 0.000
- Difference (null H = exogenous): chi2(7) = 34.65 Prob > chi2 = 0.000
复制代码定义maxldep(2)其实是 L(2/3).n,即n的滞后2阶和3阶作为工具变量,而不只是2阶,这里或者
作者是针对l.n的最大滞后2阶说的。加入h(2)受刚才那个帖子的启发,谢谢。