Modeling extreme risks in commodities and commodity currencies

被引:9
|
作者
Fuentes, Fernanda [1 ]
Herrera, Rodrigo [2 ]
Clements, Adam [3 ]
机构
[1] Univ Talca, Fac Ingn, Camino Niches Km 1, Curico, Chile
[2] Univ Talca, Fac Rde Econ & Negocios, Av Lircay S-N, Talca, Chile
[3] Queensland Univ Technol, Sch Econ & Finance, Brisbane, Qld, Australia
关键词
Commodity currency; BEKK; Hawkes model; Value at risk; REAL EXCHANGE-RATE; VALUE-AT-RISK; POINT PROCESS; MARKETS; PRICES; MACROECONOMY; UNCERTAINTY; RETURNS; RATES; TIME;
D O I
10.1016/j.pacfin.2018.06.003
中图分类号
F8 [财政、金融];
学科分类号
0202 ;
摘要
This paper examines extreme co-movements between the Australian and Canadian currencies, often known as commodity currencies, and gold and oil markets respectively. Here, two main approaches based on extreme value theory are compared in the context of explaining the co movements between the markets in times of market instability. On the one hand, the intensity of the extreme events is represented by self-exciting marked point processes using a multivariate extension of the Hawkes-POT model, while contemporaneous co-movements are characterized utilizing a more traditional multivariate volatility model, the DBEKK-EVT model. It is found that intensity and volatility follow similar paths through time. The Hawkes-POT model reveals the unidirectional influence of the commodity on the currency, consistent with previous literature. Hawkes-POT model produces slightly more accurate Value at Risk results in the in-sample period, while the results are mixed in the backtesting period. Overall it seems as though the simpler multivariate volatility based approach produce forecasts of extreme risk that are comparable to the more complex Hawkes model.
引用
收藏
页码:108 / 120
页数:13
相关论文
共 50 条