Measuring high-frequency phase shifts between stock markets
|Author:||Mgr. Jakub Cieslar|
|Year:||2019 - winter|
|Leaders:|| Mgr. Lukáš Vácha Ph.D.
|Work type:|| Finance, Financial Markets and Banking
|Awards and prizes:|
|Abstract:||Thesis investigates frequency-related phase relationships among returns of five
m a jor 5-minute European stock m arket indexes and compares relative phases
on high frequencies, w ith focus on dynamics between developed and developing
stock m arkets from 2008 to 2015 using. Using continuous and discrete wavelet
transform we find significant phase relationships among the considered indexes,
particularly we spot very strong relationship between the developed ones with
no significant phase difference on any investigated frequency. Furtherm ore we
observe significant lag of developing m arkets behind developed ones, particularly on horizons between 20 and 80 minutes. We also observe th a t the relationships is fading throughout the examined period, w ith increased variance of
the relative phases and diminishing significance of some phase differences. The
results indicate th a t either less developed m arkets are becoming more effective
or it can be a sign of decreasing inter-dependencies (e.g. lower common trends).
This thesis contributes to the literature by examining noisy financial tim e series
on highest frequencies and shows relevance of the m ethod on simulated signals
w ith high degree of noise.