Abstract—Present paper objective is to analyze how trends can be detected and interpreted using a GARCH (1, 1) model. Through VAR calculation along 2 distinct ways, the whole concept is to find out if trends can be identified in the return series of an index. Then this information can be used with an estimated probability to complement the VAR in order to get better anticipation of possible losses in a stressed environment. In addition, back testing on CVaR reliability compared to the VaR will be run as well.
Index Terms—CVaR, estimated shortfall, GARCH, model comparison, portfolio management, trend analysis, VAR.
M. Maïza, A. Clavel, N. Hatem, and C. Recommandé are with the ECE Paris, Paris, 75015 France (e-mail: maiza@ece.fr; clavel@ece.fr; hatem@ece.fr; recommand@ece.fr).
[PDF]
Cite: M. Maïza, A. Clavel, N. Hatem, and C. Recommandé, "VaR Trend Analysis from Discretized and Continuous Approaches," International Journal of Trade, Economics and Finance vol.6, no.4, pp. 218-222, 2015.