The key benefit of using a non-parametric density estimation method like Kernel Density Estimation (KDE) is that it creates a smooth, continuous distribution from the limited historical data. This allows for flexible percentile selection (B).

With only 80 data points, using a direct percentile method (like taking the 2nd smallest return for a 97.5% VaR) is very sensitive to individual data points and doesn't give a very precise estimate. KDE interpolates between the data points, creating a much smoother distribution from which you can read off any percentile you want with greater precision than the raw data would allow.

A is incorrect. KDE doesn't increase the actual number of data points. It uses the existing data to estimate a continuous density function.

C is incorrect. KDE doesn't eliminate outliers. The density estimation will still be influenced by extreme values, though their impact is smoothed out across the distribution.

D is incorrect. Correlation is not directly related to this approach. KDE focuses on the distribution of a single variable (returns in this case), not the relationship between multiple variables.