What determines the PEP-to-average power ratio of an unprocessed single-sideband phone signal?
Why In SSB the RF envelope is a direct copy of the audio waveform, so the ratio between the highest instantaneous peak (PEP) and the long-term average power is set by the peak-to-average ratio of the talker's voice. Speech is very spiky, with brief loud syllables separated by quiet intervals, so an unprocessed SSB signal typically averages only a fraction of its PEP. That is why speech processing, which compresses the dynamic range, raises average power without raising PEP.
Watch out Carrier suppression is tempting, but a properly generated SSB signal has essentially no carrier to begin with, and any residual carrier does not control the voice envelope's peak-to-average behavior. Amplifier gain scales peak and average power together, leaving the ratio unchanged.
The envelope is your voice: peak-to-average comes from how you talk, and processing evens it out.