Galaxy bias from galaxy-galaxy lensing in the DES Science Verification Data

We present a measurement of galaxy-galaxylensing around a magnitude-limited ($i_{AB} < 22.5$) sample of galaxiesselected from the Dark Energy Survey Science Verification (DES-SV) data. We split these lenses into three photometric-redshift binsfrom 0.2 to 0.8, and determine the product of the galaxybias $b$ and cross-correlationcoefficient between the galaxyand dark matter overdensity fields $r$ in each bin, using scales above 4 Mpc/$h$ comoving, where we find the linear bias model to be valid given our current uncertainties. We compare our galaxybias results from galaxy-galaxylensing with those obtained from galaxy clustering(Crocce et al. 2016) and CMB lensing (Giannantonio et al. 2016) for the same sample of galaxies, and find our measurements to be in good agreement with those in Crocce et al. (2016), while, in the lowest redshift bin($z\sim0.3$), they show some tension with the findings in Giannantonio et al. (2016). Our results are found to be rather insensitive to a large range of systematic effects. We measure $b\cdot r$ to be $0.87\pm 0.11$, $1.12 \pm 0.16$ and $1.24\pm 0.23$, respectively for the three redshift binsof width $\Delta z = 0.2$ in the range $0.2 photometric-redshiftalgorithm BPZ. Using a different code to split the lens sample, TPZ, leads to changes in the measured biases at the 10-20% level, but it does not alter the main conclusion of this work: when comparing with Crocce et al. (2016) we do not find strong evidence for a cross-correlationparameter significantly below one in this galaxysample, except possibly at the lowest redshift bin($z\sim 0.3$), where we find $r = 0.71 \pm 0.11$ when using TPZ, and $0.83 \pm 0.12$ with BPZ, assuming the difference between the results from the two probes can be solely attributed to the cross-correlationparameter.