Abstract

Context. The recent discovery of three giant planets orbiting the extremely metal-poor stars HIP 11952 and HIP 13044 have challenged theoretical predictions of the core-accretion model. According to this, the metal content of the protoplanetary disk from which giant planets are formed is a key ingredient for the early formation of planetesimals prior to the runaway accretion of the surrounding gas. Aims. We reanalyzed the original FEROS data that were used to detect the planets to prove or refute their existence, employing our new reduction and analysis methods. Methods. We applied the cross-correlation technique to FEROS spectra to measure the radial velocity variation of HIP 13044 and HIP 11952. We reached a typical precision of similar to 35 m s(-1) for HIP 13044 and similar to 25 m s(-1) for HIP 11952, which is significantly superior to the uncertainties presented previously. Results. We found no evidence of the planet orbiting the metal-poor extragalactic star HIP 13044. We show that given our radial velocity precision, and considering the large number of radial velocity epochs, the probability for a non-detection of the radial velocity signal recently claimed is lower than 10(-4). Finally, we also confirm findings that the extremely metal-poor star HIP 11952 does not contain a system of two gas giant planets. These results reaffirm the expectations from the core-accretion model of planet formation.