Thank you, DVD, for taking the time to not only view my work on the Monte Carlo simulation example but also to provide strong feedback on it. Let’s address your concerns about my model.

Please keep in mind that my model only expresses risk management challenges from a normal market conditions standpoint. If the market starts to move out of the parameters set by the model, it might be disastrous for some. That’s why monitoring features and constant development and updates are essential for large-sized portfolios. Additionally, this model should be built on top of another model that can be automatically triggered in case a major event takes place. I have built such systems for large corporations.

Now, let’s address the mean component in my calculations. In my opinion, using the mean as a component to calculate volatility is the best option in a normal environment. Using the mode or median can distort the data and create outliers in the long run, especially if your data points are imbalanced. Although you can fix these problems with different techniques, such as using SMOTE or Condensed Nearest Neighbor techniques, these methods might not work well with fixed and unchangeable data, such as stock market returns.

True, returns fluctuate daily and might go up or down significantly. However, the model takes the returns of a certain period and works within that range of volatility. Hence, the mean should reflect that range, translating into VAR and CVAR numbers that make sense based on the given returns for that period.

I hope I have answered your queries effectively. If not, please connect with me or DM me on LinkedIn, and I will be more than happy to collaborate to find ways to make this model perform better.