Time Series for Actuaries

Explore the fundamentals of time series for actuaries, covering stationary properties, auto and partial auto correlation, white noise, trends, seasonality, ARMA models, integration, volatility modeling, and practical R code.

Explore time series using cash, stock prices, and bond prices to see how observations over time reveal dependencies, enabling description, modeling, forecasting, out-of-control checks, and connections with other time series.

Explore stationary properties, including strictly and weakly stationary time series with constant mean and variance and lag-dependent covariance, and the Markov property, where the present predicts the future.

Explore the auto covariance function, its link to the mean function and variance, and how the auto correlation function and partial auto correlation function guide order regressive and moving average.

Explore white noise with zero mean and a spike at zero in the auto correlation function, and see how alternating, trending, and seasonal time series differ in decay.

Explore the ARIMA framework, detailing p, d, q, and how white noise underpins auto regressive, moving average, and integrated components. See how differencing restores stationarity.

Learn to test for stationary, transform nonstationary time series, and fit the prima model using differencing and trend removal, with diagnostic tests to forecast future data.

Explain how GARCH models measure volatility by modeling variance as omega plus alpha times lagged squared returns plus beta times lagged volatility, capturing volatility clustering.