Usporedite metode

	Testiranje pouzdanosti vrijednosti u riziku (VaR)	HAR-RV model ostvarene volatilnosti	Regresija običnih najmanjih kvadrata (OLS)
Područje	Financije	Financije	Ekonometrija
Obitelj	Regression model	Regression model	Regression model
Godina nastanka	1998	2009	2019
Tvorac	Kupiec (1995); Christoffersen (1998); Engle & Manganelli (DQ test)	Fulvio Corsi	Wooldridge (textbook treatment); classical least squares
Vrsta	Statistical hypothesis tests on VaR violation sequences	Linear time-series regression for volatility	Linear regression
Temeljni izvor	Kupiec, P. H. (1995). Techniques for Verifying the Accuracy of Risk Measurement Models. The Journal of Derivatives, 3(2), 73-84.	Corsi, F. (2009). A Simple Approximate Long-Memory Model of Realized Volatility. Journal of Financial Econometrics, 7(2), 174–196.	Wooldridge, J. M. (2019). Introductory Econometrics: A Modern Approach (7th ed.). Cengage Learning. ISBN: 978-1337558860
Drugi nazivi	VaR backtest, Kupiec test, Christoffersen test, Dynamic Quantile test	HAR-RV, heterogeneous autoregressive realized volatility, Corsi HAR model, HAR-RV Modeli (Heterogeneous Autoregressive Realized Volatility)	ordinary least squares, classical linear regression, linear regression, en küçük kareler regresyonu
Srodne	3	5	5
Sažetak	VaR backtesting is a family of statistical tests that validate a risk model by comparing its Value-at-Risk forecasts against realised losses. It builds on Kupiec's (1995) unconditional coverage test, Christoffersen's (1998) conditional coverage test, and the Engle-Manganelli Dynamic Quantile (DQ) test.	The HAR-RV model, introduced by Fulvio Corsi in 2009, forecasts realized volatility by decomposing it into daily, weekly, and monthly components. It is a simple linear regression that mirrors how market participants with different investment horizons react to volatility, and it naturally captures the long-memory behaviour of volatility.	Ordinary Least Squares is the classical linear regression method that explains a continuous outcome as a linear combination of predictors. It estimates the coefficients by minimising the sum of squared residuals, and under the Gauss-Markov assumptions these estimates are the best linear unbiased estimator (BLUE).