The problem of MAG meeting in a two-stage PA simulation in ADS

Dear friends
These days I am designing a 2-stage power-amplifier in 5.8 GHz in ADS , but when I test the S parameter , there is something strange about its MAG(max available gain ) , in low frequency , its MAG is -3040 while in 5.8 GHz its MAG is normal . when I test the first and second stage PA separately , the S-parameters and MAG of both PA are normal . The problem only exists when the PAs are put together. Is it the problem of arithmetic inside ADS ? Or the problem of my circuit ?
Could anyone help me , tell me where the problem is ? Thank you very much !

This is not a realistiv value so we should assume that it is a software error or you have a problem with your setup. It looks like you know what you are doing but I have to ask: did you place all the DC blocks at the input and RF blocks at the bias?

I have put all the DC_blocks and the RF_blocks . And I did the 1-tone and 2-tone simulations , the results were all normal . It's just the MAG which is unusual . So I am not sure whether could I do the next step to design the PCB or continued to find where the problem is .