I am using simulations to compare two economic models and want to understand their impact on returns (i.e., the percentage change in prices). I have employed common random numbers for these simulations and understand how to compare the models based on their mean returns following the methods described here: https://www.researchgate.net/publication/228039569_Comparing_Systems_via_Simulation However, I am also interested in comparing these models in terms of the variance of their returns, as this is a crucial measure of risk. The simulated data (returns) are highly skewed and not normally distributed, which complicates the comparison. While I am aware that I could compare the simulations pairwise and report the distribution of test statistics (Levene’s test seems ideal) for a large number of replications, I believe there might be a more effective method that leverages all the data. Could someone please provide guidance on how to approach this comparison or suggest any readings that could help? Any advice would be greatly appreciated.
Asked
Active
Viewed 26 times
