Page 45 - Contributed Paper Session (CPS) - Volume 8

CPS2176 Chiraz KARAMTI et al.
Tableau 1. Interpretation of time scales

Horizon [2 , 2 +1 ] = 1, 2, … ,7
Scale Component Frequency resolution
Scale 1 d1 10-20 mn
Scale 2 d2 20-40 mn
Scale 3 d3 40-80 mn
Scale 4 d4 80-160 mn
Scale 5 d5 160-320 mn
Scale 6 d6 320-640 mn
Scale 7 d7 640-1280 mn≈ 1
Approximation A7 >1280

The wavelet coefficients and scale coefficients of the exchange rate level time
series derive a Mallat decomposition algorithm using Daubechies least
asymmetric (LA) wavelet filter of length L=8 are shown in Fig. 1.

Fig 1. Decomposed wavelet sub-time series components (d) and
Approximation (a) of crude oil

The outcomes of estimating the parameters of the model given in Eq. (5) using
the MODWT transformed exchange rate (EUR/USD) return series are reported
1
in Table 2 .

1
The MODWT-ARMA (2,0) model is the best fitting model for the mean equation according to the
conventional information criteria (AIC and BIC). Given the amount of results, the mean equation
34 | I S I W S C 2 0 1 9