Inconsistency between time and freq. domain results in pss plot

I ran pss with shooting on an oscillator and plotted the output in time and frequency domain(spectrum). The phase shown in the spectrum is inconsistent with the phase seen on the time domain plot(with sine or cosine as reference). In the attached figure, the figure on the left is the time domain plot and is visibly close to being a cosine with about +90 degree phase. The phase shown in the spectrum is -35.11 degrees. The figure on the right shows the spectrum output from Results->Plot->Main Form as well as DFT of the time domain waveform taken from the calculator. The two have the same magnitude, but different phase. This can also be seen in the snapshot of the table below(Columns 2 and 3 are magnitude and phase from . Results->Plot->Main Form and columns 5 and 6 are from DFT.)

If I use harmonic balance, I don't see this inconsistency. Would appreciate a fix or any input on this.