Serial Correlation and Stationarity in Financial Time Series
Summary
The document introduces serial correlation, or autocorrelation, as a way to describe dependence between observations in a time series. It reviews expectation, variance, standard deviation, covariance, and correlation, then applies these ideas to time series. For a second-order stationary series, the mean and variance stay constant, while autocovariance depends on lag rather than calendar time. The autocorrelation function normalizes autocovariance by the series variance, and sample versions estimate these quantities from observed data.
The note explains why identifying dependence can help with forecasting, simulation, and risk management in trading research. It is conceptual rather than a worked analysis: formulas are referenced but several are missing from the supplied text, and no dataset, diagnostic procedure, or empirical result is provided. The discussion also does not establish that autocorrelation implies a profitable strategy; practical use depends on appropriate stationarity assumptions and further testing.
Key ideas
- Expectation and variance describe a random variable’s center and spread, while covariance and correlation describe how two variables move together.
- Second-order stationarity requires a constant mean and variance and autocovariance that depends only on lag.
- Autocorrelation is autocovariance normalized by variance and can be examined across different lags.
- Sample autocovariance and autocorrelation estimate serial dependence from observed time-series values.
- Serial dependence can inform forecasts, simulations, and risk analysis, but the note provides no empirical trading validation.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.