We study the stellar and star formation properties of the host galaxies of 58 X-ray selected AGN in the GOODS portion of the Chandra Deep Field South ( CDF-S ) region at z \sim 0.5 - 1.4 .
The AGN are selected such that their rest-frame UV to near-infrared spectral energy distributions ( SEDs ) are dominated by stellar emission , i.e. , they show a prominent 1.6 \mu m bump , thus minimizing the AGN emission ’ contamination ’ .
This AGN population comprises approximately 50 % of the X-ray selected AGN at these redshifts .
Using models of stellar and dust emission we model their SEDs to derive stellar masses ( \mathcal { M } _ { * } ) and total ( UV+IR ) star formation rates ( SFR ) .
We find that AGN reside in the most massive galaxies at the redshifts probed here .
Their characteristic stellar masses ( \mathcal { M } _ { * } \sim 7.8 \times 10 ^ { 10 } { M } _ { \odot } and \mathcal { M } _ { * } \sim 1.2 \times 10 ^ { 11 } { M } _ { \odot } at median redshifts of 0.67 and 1.07 , respectively ) appear to be representative of the X-ray selected AGN population at these redshifts , and are intermediate between those of local type 2 AGN and high redshift ( z \sim 2 ) AGN .
The inferred black hole masses ( \mathcal { M } _ { BH } \sim 2 \times 10 ^ { 8 } { M } _ { \odot } ) of typical AGN are similar to those of optically identified quasars at similar redshifts .
Since the AGN in our sample are much less luminous ( L _ { 2 - 10 keV } < 10 ^ { 44 } { erg s } ^ { -1 } ) than quasars , typical AGN have low Eddington ratios ( \eta \sim 0.01 - 0.001 ) .
This suggests that , at least at intermediate redshifts , the cosmic AGN ’ downsizing ’ is due to both a decrease in the characteristic stellar mass of typical host galaxies , and less efficient accretion .
Finally there is no strong evidence in AGN host galaxies for either highly suppressed star formation ( expected if AGN played a role in quenching star formation ) or elevated star formation when compared to mass selected ( i.e. , IRAC-selected ) galaxies of similar stellar masses and redshifts .
This may be explained by the fact that galaxies with \mathcal { M } _ { * } \sim 5 \times 10 ^ { 10 } -5 \times 10 ^ { 11 } { M } _ { \odot } are still being assembled at the redshifts probed here .