Climate models are capable of producing features similar to tropical cyclones, but typically display strong biases for many of the storm physical characteristics due to their relatively coarse resolution compared to the size of the storms themselves. One strategy that has been adopted to circumvent this limitation is through the use of a hybrid downscaling technique, wherein a large set of synthetic tracks are created by seeding disturbances in the large-scale environment. Here, we evaluate the ability of this technique at reproducing many of the characteristics of the recent North Atlantic hurricane activity as well as its sensitivity to the choice of the reanalysis dataset used as boundary conditions. In particular, we show that the geographical and intensity distributions are well reproduced, but that the technique has difficulty capturing the large difference in activity observed between the most recent active and quiescent phase. Although the signal is somewhat reduced compared to observation, the technique also detects a significant decrease in the intensification rate of hurricanes near the coastal US during the active phase compared to the quiescent phase. Finally, the influence of the El Nino Southern Oscillation on hurricane activity is generally well captured as well, but the technique fails to reproduce the increase in activity over the western part of the basin during Modoki El Ninos.