The need for more sophisticated models, the desire to offer more individualized products, and the requirement for more accurate value at risk calculations, all suggest that the need for efficiency and accuracy in Monte Carlo simulation (MCS) will increase in the coming years. This paper focuses on two methods for increasing the efficiency of simulation methods. The performance of MCS is compared with the performance of simulation using quasirandom Sobol sequences. This comparison is performed both directly and also with a new stratified sampling variance reduction (VR) procedure. It is found that Sobol sequences are superior to pseudorandom without VR and comparable with VR.