Dear Stata users,

I am estimating a dynamic model of y = f(y_1 x1, x2, x3) using GMM. the coefficient of the dynamic parameter (y_1) is negative and these results are strange to me. My concern is whether this coefficient can be negative in the first place? Theoretically, I have good reasons to estimate a dynamic model. See results below
xtabond2 lnfdi1 l.lnfdi1 BIT GDP TRADE RENTS D.EXR yr*, gmm(l.lnfdi1 l(0/1).BIT l(0/2).GDP l(0/1).TRADE l(0/1).RENTS l(0/1).D.EXR, collapse lag(1 1)) iv(BIT GDP TRADE RENTS D.EXR yr*, eq(level)) twostep robust small orthog
Code:
Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: country                         Number of obs      =       730
Time variable : year                            Number of groups   =        48
Number of instruments = 47                      Obs per group: min =         4
F(23, 47)     =    215.64                                      avg =     15.21
Prob > F      =     0.000                                      max =        16
------------------------------------------------------------------------------
             |              Corrected
      lnfdi1 |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      lnfdi1 |
         L1. |  -.0317584   .0287833    -1.10   0.275    -.0896629    .0261462
             |
         BIT |   .1747198   .2629956     0.66   0.510     -.354359    .7037986
         GDP |   .0031572   .1793907     0.02   0.986    -.3577303    .3640447
       TRADE |   .0438124   .3197854     0.14   0.892    -.5995128    .6871376
       RENTS |  -.1092118   .1148513    -0.95   0.347    -.3402628    .1218391
             |
         EXR |
         D1. |   .1979432   .0635437     3.12   0.003     .0701097    .3257767
             |
        yr_1 |          0  (omitted)
        yr_2 |   5.296599   3.717287     1.42   0.161    -2.181618    12.77482
        yr_3 |   5.487772   3.774281     1.45   0.153    -2.105101    13.08065
        yr_4 |   5.844388    3.81474     1.53   0.132     -1.82988    13.51866
        yr_5 |    5.20261   3.690366     1.41   0.165    -2.221449    12.62667
        yr_6 |   5.367215   3.749457     1.43   0.159     -2.17572    12.91015
        yr_7 |   5.311162    3.74973     1.42   0.163    -2.232322    12.85464
        yr_8 |   5.885764   3.936006     1.50   0.142     -2.03246    13.80399
        yr_9 |   5.005554   3.749457     1.34   0.188    -2.537381    12.54849
       yr_10 |   5.400537   3.796573     1.42   0.161    -2.237182    13.03826
       yr_11 |   6.080197   3.794042     1.60   0.116    -1.552432    13.71282
       yr_12 |   5.533739   3.798446     1.46   0.152    -2.107748    13.17523
       yr_13 |   5.410425   3.894485     1.39   0.171    -2.424268    13.24512
       yr_14 |   5.487684   3.767903     1.46   0.152    -2.092359    13.06773
       yr_15 |   5.740524   3.851853     1.49   0.143    -2.008406    13.48945
       yr_16 |   5.600427   3.906189     1.43   0.158    -2.257812    13.45867
       yr_17 |   5.115525   3.666877     1.40   0.170     -2.26128    12.49233
       _cons |          0  (omitted)
------------------------------------------------------------------------------
Instruments for orthogonal deviations equation
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    L.(L.lnfdi1 BIT L.BIT GDP L.GDP L2.GDP TRADE L.TRADE RENTS L.RENTS D.EXR
    LD.EXR L2D.EXR) collapsed
Instruments for levels equation
  Standard
    BIT GDP TRADE RENTS D.EXR yr_1 yr_2 yr_3 yr_4 yr_5 yr_6 yr_7 yr_8 yr_9
    yr_10 yr_11 yr_12 yr_13 yr_14 yr_15 yr_16 yr_17
    _cons
  GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(L.lnfdi1 BIT L.BIT GDP L.GDP L2.GDP TRADE L.TRADE RENTS L.RENTS D.EXR
    LD.EXR L2D.EXR) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z =  -4.97  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -1.43  Pr > z =  0.152
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(23)   =  19.29  Prob > chi2 =  0.684
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(23)   =  21.16  Prob > chi2 =  0.571
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(10)   =  11.40  Prob > chi2 =  0.327
    Difference (null H = exogenous): chi2(13)   =   9.77  Prob > chi2 =  0.713
  iv(BIT GDP TRADE RENTS D.EXR yr_1 yr_2 yr_3 yr_4 yr_5 yr_6 yr_7 yr_8 yr_9 yr_10 yr_11 yr_12 yr_13 yr_14 yr_15 yr_16 yr_17, eq(level)
> )
    Hansen test excluding group:     chi2(3)    =   2.49  Prob > chi2 =  0.476
    Difference (null H = exogenous): chi2(20)   =  18.67  Prob > chi2 =  0.543