This document discusses the application of time series analysis in finance. It provides an overview of financial time series and their characteristic properties. Key points include:
- Financial time series have high frequency values which leads to high volatility that changes over time. Systematic factors create trends and cycles in the data.
- Under the efficient market hypothesis, prices fully reflect all available information and changes are unpredictable. The martingale model suggests the best forecast of tomorrow's price is today's price.
- For the martingale model to work properly, simple returns should follow a lognormal rather than normal distribution to ensure non-negative values. Taking the log of returns results in a normal distribution.
- Tables of statistics like mean, standard