factor_augmented_var – Factor-augmented VAR (Bernanke-Boivin-Eliasz 2005).#

Back to family axis | Back to L4 | Browse all options

Operational op under axis family, sub-layer L4_A_model_selection, layer l4. Standalone callable: mf.functions.favar_fit.

Function signature#

mf.functions.favar_fit(
    X: np.ndarray | pd.DataFrame,
    y: np.ndarray | pd.Series,
) -> FAVARFitResult

Parameters#

name

type

default

constraint

description

X

`np.ndarray

pd.DataFrame`

—

—

y

`np.ndarray

pd.Series`

—

—

Returns#

FAVARFitResult — frozen dataclass with fit results.

Attribute

Type

Description

.n_factors

int

Number of PCA factors extracted from X.

.n_lags

int

VAR lag order p.

.predict(X)

np.ndarray

Predictions for new data X, shape (n_samples,).

.summary()

str

Table: factor count and lag order.

Behavior#

Two-stage estimator: PCA factors from the predictor panel + VAR(params.n_lag) on (factors, target). Captures dynamic interactions between latent factors and the target series.

Useful for monetary-policy studies where the factors stand in for unobserved economic state.

When to use

Monetary-policy / macro-state studies; diffusion-index VAR baselines.

In recipe context#

Set params.family = "factor_augmented_var" in the relevant layer to activate this op within a recipe:

# Layer L4 recipe fragment
params:
  family: factor_augmented_var

References#

  • macroforecast design Part 2, L4: ‘forecasting model is the layer where every authoring iteration ends – pick family, tune, repeat.’

  • Bernanke, Boivin & Eliasz (2005) ‘Measuring the Effects of Monetary Policy: A Factor-Augmented Vector Autoregressive Approach’, QJE 120(1).