The Wall Street Journal has an article today about the limitations of using Monte Carlo simulations for retirement portfolio planning. The article properly notes that the real issue is not the simulations, but the fundamental assumptions that these models often make. Most Monte Carlo simulations that were being used by financial planners assumed that market returns adhered to a bell-curve-shaped distribution. Of course, this means very thin tails on the distribution, i.e. an extremely low probability of an "extreme" event such as the 2008-2009 market collapse.
Of course, these planners and the simulations they used were not the only ones improperly assuming a bell-curve-shaped distribution. Most experts also ignored the possibility of "fatter tails" on the distribution. In other words, extreme events might not be as rare as once thought. For a great book on this subject, check out The Black Swan by Nassim Nicholas Taleb.