Autoregressive model
A time-series model that explains a variable by its own past values; an AR(p) model uses p lags. Its estimates and tests are valid only if the series is covariance stationary.
Where you will meet it
Autoregressive model appears in 3 learning outcomes on the CFA syllabus. A few of them:
- Explain what decides the choice between a linear and a log-linear trend for a given series, and assess where trend models fall short
“Serial correlation makes the trend model's inference unreliable and suggests an autoregressive model instead.”
- Explain why a time series needs to be covariance stationary and what it means when a series lacks stationarity
“Autoregressive models give valid estimates and tests only on stationary series.”
- Set out how an order-p autoregressive (AR) model is built and, from its fitted coefficients, work out forecasts one and two periods ahead
“An autoregressive model of order p explains a series by its own previous p values: x(t) = b0 + b1x(t − 1) + … + bpx(t − p) + ε.”
See also
Covariance · Covariance-stationary
Practise it
Practise autoregressive model until it is automatic: questions that explain every wrong option.
Take the free diagnostic →