Monte Carlo simulation with non-normal distributions
Posted by David Gordon on December 12, 2001 4:34 PM
I am using Excel to model Monte-Carlo simulation, and have been using the formula (NORMINV(RAND(),m,s) to provide normally distributed random returns.
How do I do the same with non-normal distributions, e.g fat-tailed?
Posted by Joe Was on December 12, 2001 4:49 PM
Try looking up these fat tailed functions:
Gammadist
Chidist
Ftest
What kind of data are you looking at and what type of results are you looking for? JSW
Posted by David on December 12, 2001 5:29 PM
thanks for those suggestions. I'm using portfolio data (Average return, standard deviation). Criticism has been made that such returns are not likely to be "normally" distributed, so I'd like to re-map allowing for non-normal distribution. Can you advise what Gamma and Chi are, and how they're applied?
Posted by Joe Was on December 12, 2001 6:02 PM
Now that I know what you want to do, you need to load the Statistical analysis tools addin to Excel if you have not done so. Then use the "Moving Average" tool, its what most people use.
The other worksheet functions I told you about are more or less tools to tell how good you data is and how will your sample data relates to the trends they represent. They may help you as well each can be found in Excel help and each give a worksheet example. You may also get the full Stat help by searching help with: About Statistical Analysis Tools. JSW
Posted by Mit Tit on April 04, 2002 6:16 PM
Does any one have a serial number for @RISK? Can be version 3.5, 4, or 4.5
