Abstract
This document contains the results for the estimation of Value at Risk (VaR) based on linear and non-linear quantile regression techniques. In particular, several CAViaR (conditional autoregressive value at risk) models are implemented for this purpose. These models can replicate the empirical properties of asset returns without requiring distributional assumptions. In addition, these methods are compared with traditional VaR techniques for the Colombian peso exchange rate, a public debt market price index, and the Colombian stock price index, during the periods of December 2007 and November 2015. In general, the quantile regression-based techniques show a good performance with respect to the traditional models.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
