About this question

AR(p) Order Selection via AIC

Medium · time_series · Quant Researcher interview question · ar_model, aic, order_selection, model_selection, time_series

Autoregressive (AR) models are fundamental for modeling covariance-stationary time series like asset returns, spread signals, or volatility residuals. Selecting the optimal lag order p is critical to balance model fit and complexity, a trade-off managed by the Akaike Information Criterion (AIC). This method provides a principled, data-driven approach to prevent underfitting or overfitting in financial forecasting models. Task Implement the function ar_order_selection_aic(series, max_lag) to fin