Super-horizon perturbations and preheating

It has recently been claimed by Bassett et al that preheating after inflation may affect the amplitude of curvature perturbations on large scales, undermining the usual inflationary prediction. We analyze the simplest model, and confirm the results of Jedamzik and Sigl and of Ivanov that in linear perturbation theory the effect is negligible. However the dominant effect is second-order in the field perturbation and we show that this too is negligible, and hence conclude that preheating has no significant influence on large-scale perturbations in this model. We briefly discuss the likelihood of an effect in other models.