ABSTRACT
High frequency data usually refers to financial data collected daily, by the hour, by the minute or even by the second. Ultra High Frequency data refers to tip-by-tip trading data. Models of high frequency financial data have developed quickly since the beginning of the 1990s, and they are now widely used in theoretical and empirical studies of microstructures of financial markets. Generally, because of continuity of impact of information on changes in prices, discrete models will cause loss of information; the lower the frequency of data, the more will be the loss of information. Low frequency data cannot reveal real-time dynamic characteristics of financial assets prices; high frequency data contain more and abundant intraday volatility information and, therefore, high frequency data is a more effective indicator of how information influences the markets. In research based on high frequency data, one important issue is description of the statistical laws and characteristics of the trading patterns. The ‘Calendar Effect’ is one of the most important findings in this field.
