Robustness of Detrended Cross-correlation Analysis Method Under Outliers Observations

The computation of the bivariate Hurst exponent constitutes an important technique to test power-law cross-correlation time series. For this objective, detrended analysis (DCCA) method represents most used one. In paper, we prove robustness DCCA method, where trend is estimated using polynomial fitting, estimate when series are corrupted by outliers observations. On other hand, give exact order and a regression region for computing function obtain least square estimator exponent. Our theoretical results shown simulation study on two-fractional Gaussian noise process Additionally, our applied financial empirical findings accompanied interpretations.