This paper demonstrates an easy way to obtain a substantial gain in precisi
on in the popular GHK simulator, and the RIS simulator that extends it, of
multivariate normal probabilities. On average, this gain translates into a
40% saving in computational time. It is also shown that the simulator can b
e used to compute probabilities over wedge-shaped areas. In this case, the
RIS simulator based on the legit distribution has a substantial advantage o
ver the GHK simulator. (C) 2000 Elsevier Science S.A. All rights reserved.
JEL classification: C15; C40.