Quantile-Quantile (QQ) Plots: A Detailed Look
Posted On:
1. Context
In this short video from FRM Part 2 curriculum, we take a detailed look at Quantile-Quantile (QQ) Plots – how to interpret them, and use them to identify the correct distribution and/or outliers in the dataset. We build the intuition about QQ Plots right from scratch, and work our way through to the final four takeaways:
- If the reference distribution matches the empirical distribution, we get linear QQ plot, even if there is a linear transformation involved.
- A linear transformation changes the slope and intercept of the QQ plot. Use the slope of the QQ plot to determine the scale of the empirical distribution and the intercept to determine the location of the empirical distribution (both relative to the chosen reference distribution).
- If empirical distribution is a heavy tailed one (relative to the reference distribution), the QQ plot has steeper slope in the tails, with the central part of the plot still being linear.
- QQ plots give you an easy means of identifying outliers in a dataset.
The details of the readings in which this topic appears are given below:
| Area | Market Risk |
| Reading | Estimating Market Risk Measures |
| Reference | Kevin Dowd, Chapter 3. Estimating Market Risk Measures: An Introduction and Overview In Measuring Market Risk, 2nd Edition, (West Sussex, England: John Wiley & Sons, 2005). |