The merger of two spiral galaxies is believed to be one of the main channels for the production of elliptical and early-type galaxies. In the process, the system becomes an (ultra) luminous infrared galaxy, or (U)LIRG, that morphs to a quasar, to a K+A galaxy, and finally to an early-type galaxy. The time scales for this metamorphosis are only loosely constrained by observations. In particular, the K+A phase should follow immediately after the QSO phase during which the dust and gas remaining from the (U)LIRG phase are expelled by the AGN. An intermediate class of QSOs with K+A spectral signatures, the post-starburst QSOs or PSQ, may represent the transitional phase between QSOs and K+As. We have compiled a sample of 72 {bona fide} $z<0.5$ PSQ from the SDSS DR7 QSO catalogue. We find the intermediate age populations in this sample to be on average significantly weaker and metal poorer than their putative descendants, the K+A galaxies. The typical spectral energy distribution of PSQ is well fitted by three components: starlight; an obscured power-law; and a hot dust component required to reproduce the mid-IR fluxes. From the slope and bolometric luminosity of the power-law component we estimate typical masses and accretion rates of the AGN, but we find little evidence of powerful radio-loud or strong X-ray emitters in our sample. This may indicate that the power-law component originates in a nuclear starburst rather than in an AGN, as expected if the bulk of their young stars are still being formed, or that the AGN is still heavily enshrouded in dust and gas. We find that both alternatives are problematic and that more and better optical, X-ray, and mm-wave observations are needed to elucidate the evolutionary history of PSQ.