It is possible to model stacked holograms, but the details depend on what exactly you are trying to accomplish.  Do you want the individual holograms to be recorded and played back in such a way that there is negligible interaction or crosstalk between them?  Or are you wanting to exploit crosstalk in some way?

Based on the model you posted, the holograms are thin and therefore have no angular and/or spectral selectivity.  In fact, no attempt is made by OpticStudio to calculate diffraction efficiencies.  However, the user can assign strengths to the various orders.  It is important to note that all defined source wavelengths will have the same strengths applied to the orders, but the diffraction angles will change based on changes in wavelength.  So, in your case I used this flexibility to split the light equally between the zero and -1 orders. 

To see how this works, I separated your holograms and turned off Fresnel reflection at the substrate interfaces just to keep things simpler (you can always stack them closer and also turn Fresnel reflection back on once you are satisfied that the holograms themselves are behaving as desired).

In this case, there is interaction between the holograms because, for example, diffracted light that is reflected from the second hologram will subsequently diffract off of the first hologram.  The model is attached. 

If you don’t want this interaction, then you can likely put each hologram into its own configuration and then simply overlay the configurations, which is a bit artificial, but if it accomplishes what you want then it should work (I think...).  Alternatively, if you want a more realistic model, then you should consider using the volume hologram capability which employs Kogelnik theory to determine the angular and spectral selectivity curves, scaled in accordance with the calculated diffraction efficiency (see: Simulating diffraction efficiency of a volume holographic grating using Kogelnik’s method).  In this case, you should probably design each hologram individually in sequential mode and use the Diffraction Efficiency analysis tools to make sure the selectivity curves are sufficient to minimize crosstalk.

Once satisfied with the hologram designs, they can be replicated in non-sequential mode.