Date of Award
Master of Arts (MA)
Economics and Finance
As a basic principle in statistics, a larger sample size is preferred whenever possible. Nonetheless, in the financial world, especially equities and currencies trading, including all available data poses great challenges due to the noise present in the volatility estimation. In his paper I examine the Two Time Scales Realized Volatility estimator by Zhang, Mykland, and Ait-Sahalia (2005b) and I find that it not only provides a more efficient estimator than a basic estimator of the integrated volatility of returns, but it also consistently estimates the microstructure noise present in the latent efficient return process. I find that by using this approach, it is possible to compare the efficiency of the prices of securities with lower transaction costs traded against those with higher transactions costs.
Romero, Aristides, "Microstructure Noise: The Use of Two Scales Realized Volatility for the Noisy High-Frequency Data and its Implications for Market Efficiency and Financial Forecasting" (2016). All Graduate Plan B and other Reports. 826.