Monte Carlo Simulation as a Financial Planing Tool
❑ Odds-On Imperfection: Monte Carlo Simulation - WSJ.com
The Wall Street Journal (Laise, 05/02/2009) has an interesting article about Monte Carlo simulation in financial planning, for example, to estimate the odds of reaching retirement financial goals.
These tools typically “run a portfolio through hundreds or thousands of potential market scenarios” to estimate potential outcome. But the chance is little to highlight a scenario like the current market slide, because such simulations “often assign minuscule odds to extreme market events.”
The report provides a simple example of a typical Monte Carlo procedure in retirement-planning. “The user enters information about his age, earnings, assets, retirement-plan contributions, investment mix and other details. The calculator crunches the numbers on hundreds or thousands of potential market scenarios, guided by assumptions about inflation, volatility and other parameters.” Finally, results of the simulation are summarized to get a “success rate,” which show the percentage of market scenarios in which the investor had money remaining at the end of his estimated life span.
The problem is that simulations are only as good as their underlying assumptions. Critics argue that many Monte Carlo models have mistakenly assume that market returns fall along a bell-curve-shaped (normal) distribution, while in fact extreme market cases happen a lot more often. Therefore, market risks were underestimated.
To address the issue, the report argue that (1) Monte Carlo models should allow users to choose a bell-curve-shaped distribution or a “fat-tailed” distribution, and (2) they should run enough iterations (“tens of thousand or hundreds of thousands”) to help “gauge” extreme events at the tail end of the distribution.”
These are great ideas, but I like to add my 2cents as well. Assumptions do not fall automatically out of the blue sky. They should be justified by empirical tests of historical data. Better yet, they can be dynamically assessed and revised as priors in ongoing iterative Bayesian procedures.
Comments
You can actually call the Bankruptcy Court in the vicinity where the bankruptcy was filed and conduct a search based either on the case number of the bankruptcy, the name of the person or company who filed it or the social security number or tax identification number of the involved parties.
Posted by: Financial Solutions | July 15, 2009 11:49 AM