haku: @indexterm TIME SERIES / yhteensä: 413
viite: 13 / 413
| Tekijä: | Precup, O.V. Iori, G. |
| Otsikko: | Cross-correlation measures in the high-frequency domain |
| Lehti: | European Journal of Finance
2007 : APR/JUN, VOL. 13:3-4, p. 319-331 |
| Asiasana: | statistical methods time series |
| Vapaa asiasana: | high-frequency correlation Fourier method Epps effect |
| Kieli: | eng |
| Tiivistelmä: | Standard correlation measures cannot be directly applied to raw data because time series (here as: t-s.) are not homogeneous on a high-frequency scale. Therefore the t-s. must be homogenized through interpolation (as: int-n). Another option is to employ methods (as: mhds.) capable of handling raw non-synchronous t-s. This paper compares two traditional mhds. using int-n with an alternative method applied directly to the actual t-s. The three mhds. are tested on simulated data and actual trades t-s. |
SCIMA