AN ALGORITHM COMPETITION - FIRST-ORDER ITERATIONS VERSUS NEWTON-BASEDTECHNIQUES

The development and use of forward-looking macromodels in policy makin g institutions has proceeded at a much slower pace than what was predi cted in the early 1980s. An important reason for this is that research ers have not had access to robust and efficient solution techniques fo r solving nonlinear forward-looking models. This paper discusses the p roperties of alternative algorithms for solving MULTIMOD, the IMF's mu lticountry model of the world economy. Relative to traditional first-o rder algorithms in use today, we find that Newton-based techniques are considerably faster and much less prone to simulation failure. (C) 19 98 Elsevier Science B.V. All rights reserved.