Investing - Theory, News & General • How to use log-normal probability distribution function to simulate correlated four-fund in Monte Carlo

Hi Guys,

I've got a Monte-Carlo simulation that I use to tell myself how much to save, and I've been told that log-normal returns are a better fit for stock returns than normal distributions.Is the way to do log-normal simulation of stock returns like like thus:

• take logarithm of each four fund returns for each year -- for a single example of a 5% return, log(1.05) = 0.0487
• fit these "logarithmic returns" to correlated normal distributions
• predict "logarithmic returns" per the fitted normal distributions
• take exponential of the predicted logarithmic returns to recover the "1.05"-like predicted values

I have access to these:

• for correlations, https://www.mathworks.com/help/stats/co ... lts_1_corr
• for multivariate random normal distribution generator, https://www.mathworks.com/help/stats/mvnrnd.html
• for log normal probability distribution function, https://www.mathworks.com/help/stats/lognpdf.html

Thanks,

-b4xt3r

Statistics: Posted by B4Xt3r — Sat Jun 01, 2024 4:11 pm — Replies 2 — Views 171

---