Second Harmonic Generation: components of the SHG susceptibility tensor that are expected to be zero

Hello community. I have just started learning how to use The Exciting code. I am interested in second harmonic generation. While doing the related tutorial with GaAs, I attempted to calculate different components of the SHG susceptibility. Given that the group symmetry of GaAs is -43m, all components are zero except for d14. However, after submitting this input file:

<input>

   <title>GaAs SHG</title>

   <structure speciespath="/home/francis/Software/exciting/tutorials/exciting/species">
      <crystal>
         <basevect>5.3435 5.3435 0.0000</basevect>
         <basevect>5.3435 0.0000 5.3435</basevect>
         <basevect>0.0000 5.3435 5.3435</basevect>
      </crystal>
      <species speciesfile="Ga.xml" rmt="2.0">
         <atom coord="0.00 0.00 0.00"/>
      </species>
      <species speciesfile="As.xml" rmt="2.0">
         <atom coord="0.25 0.25 0.25"/>
      </species>
   </structure>

   <groundstate
      do="fromscratch"
      rgkmax="7.0"
      ngridk="11 11 11"
      xctype="LDA_PW"
      nempty="10"
      >
   </groundstate>

   <properties>
      <momentummatrix/>
      <shg
         wmax="0.3"
         wgrid="400"
         swidth="0.004"
         etol="1.d-4"
         scissor="0.0423"
         tevout="true"
         >
         <chicomp>1 1 1</chicomp>
      </shg>
   </properties>

</input>

I did not obtain zero for d_{11} ; however, the values range from -0.1 to 0.1. In reality, would I never obtain zero due to numerical factors? I conducted several numerical experiments by increasing ngridk​. I noticed that by increasing that parameter, \chi^{(2)}_{111} gets reduced; however, it is not yet zero. What am I missing? I am lost. Please, could someone in the community provide insights into this? I would really appreciate it.