Browsing by Subject "High-frequency analysis"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Open Access Essays in empirical finance(Bilkent University, 2019-12) Serdengeçti, SüleymanThis thesis comprise three essays that investigate foreign exchange market volatility and its dynamics using high frequency exchange rate data. In the first essay, we decompose the jump component of USDTRY exchange rate volatility and investigate association of jump frequencies and sizes with portfolio ows, carry trade activity and proxies for heterogeneous expectations derived from foreign exchange rate forecasts, currency options and forecasts for key macro-economic variables. The findings of the essay show that portfolio ows, particularly bond ows significantly reduce size and frequency of jumps. Moreover, we observe significant increases in jump size and frequencies with increasing dispersion in beliefs in future exchange rate level and CPI. In the second essay, we study the dynamics of return and liquidity jumps for USDMXN, USDTRY and USDZAR exchange rates. The findings of the essay show that the duration between consecutive return jump arrivals are significantly reduced by average liquidity level in the same period. Furthermore, arrival rates of both liquidity and return jumps are significantly affected by market-wide risk and liquidity factors and key macroeconomic news releases. In the third essay, we investigate the trading volume and volatility nexus for USDTRY exchange rate by using local banks' foreign exchange transaction volume data. In this context, foreign currency denominated spot, forward and swap transactions in with local and foreign customers and between each other for intraday realized volatility of different trading sessions. The findings of this study reveal that positive contemporaneous relationship between trading volume and volatility is evident for local customers and in local trading sessions. Moreover, dispersion in expectations for future foreign exchange rate strengthens this relationship.Item Open Access Essays on foreign exchange(Bilkent University, 2022-09) Uzun, SevcanThis thesis investigates the foreign exchange market dynamics by using high fre-quency data. There is a vast literature on currency markets. However, we aimed to bring a new light on the foreign exchange market dynamics by investigating high frequency data for a set of mostly traded currencies, that includes both developed and emerging market currencies. In the first chapter, we focus on the commonality in liquidity in the foreign exchange market where we are able to contribute to the literature by using a comprehensive data set (14 currencies) with high frequency analysis. Our findings indicate that commonality in liquidity exist for foreign ex-change markets even beyond crisis periods and also monetary policy meetings of Federal Reserve (FOMC) have significant effect on commonality in liquidity. In the second chapter, we study the foreign exchange market for a large data set (14 currencies) where we analyzed the predictability of jumps in the foreign exchange market. We showed that different machine learning methodologies can be used for jump prediction as well as prediction of the direction of jumps in foreign exchange market where Multi Layer Perceptron (MLP), Support Vector Machine (SVM) and Random Forest methodologies have the highest accuracy rates. In our analysis, we are able to predict the occurrence of jumps as well as the direction of jumps in the foreign exchange market using state of art machine learning methodologies for high frequency data even for the Covid Pandemic period where volatility in the foreign exchange market is very high.