AD8099 Spice Model Behavior


I am trying to simulate an AD8099 cascaded amplifier (please see the attached). Each stage amplifies x10, and the overall gain should be x100.

First I built the circuit using the FB pins of AD8099 as in the attached circuit and did an AC analysis, but the obtained gains at each stage were x7.3 and x62.3, which are way smaller than (I believe) what they should be. Then I reconnected the feedback resistors to the Vout (gave up using FB pins) and finally got gains I expected.

This happens in both MultiSim and LTSpice (although the gains are different when using FB pins).

Is this the simulation only issue? Or does this happen even in a real circuit?

Thanks in advance!