Explorations in Asset Returns with Examples in R
9 Pages Posted: 19 Sep 2015 Last revised: 20 Sep 2015
Date Written: September 18, 2015
Robustness of common estimates (mean, volatility) from returns data depends on the period. It is common that we have a dataset of daily returns but would like to estimate and optimise portfolio allocations for the monthly horizon or longer. Statistical projection is required. To develop an intuition it is recommended to rediscover the rules taught about asset returns, the most common is the scaling of volatility with the square root of time period and how that affects probability density. Building histograms with varying bucket size as well as constructing Q-Q plots by-step help to gain appreciation about densities and empirical nature of probability distribution.
This tutorial comes in three sections covering projection with asset returns, density histogram, and empirical Q-Q plot construction, for which R code is provided in the Appendix.
Keywords: asset returns, projection, histogram, percentiles, Q-Q plot, Normality tests, iid, R
JEL Classification: C5, G11
Suggested Citation: Suggested Citation