NICEFA Logo
nicefa.
Library
←Module

The Autoregressive (AR) Model: Definition, Properties, and Conditions for Stationarity

Master the AR(p) model with our rigorous guide on stationarity conditions, characteristic polynomials, and the underlying stochastic mechanics for math students.

Visualizing...

Our institutional research engineers are currently mapping the formal proof for The Autoregressive (AR) Model: Definition, Properties, and Conditions for Stationarity.

Apply for Institutional Early Access →

The Formal Theorem

A stochastic process {Yt}t∈Z \{Y_t\}_{t \in \mathbb{Z}} follows an Autoregressive model of order p p , denoted as AR(p) AR(p) , if it satisfies the stochastic difference equation:
Yt=c+ϕ1Yt−1+ϕ2Yt−2+⋯+ϕpYt−p+ϵt=c+∑i=1pϕiYt−i+ϵt \begin{aligned} Y_t &= c + \phi_1 Y_{t-1} + \phi_2 Y_{t-2} + \dots + \phi_p Y_{t-p} + \epsilon_t \\ &= c + \sum_{i=1}^{p} \phi_i Y_{t-i} + \epsilon_t \end{aligned}
where ϵt∼WN(0,σ2) \epsilon_t \sim WN(0, \sigma^2) denotes white noise. The process is covariance-stationary if and only if all roots of the characteristic polynomial Φ(z)=1−ϕ1z−ϕ2z2−⋯−ϕpzp=0 \Phi(z) = 1 - \phi_1 z - \phi_2 z^2 - \dots - \phi_p z^p = 0 lie strictly outside the unit circle in the complex plane, i.e., ∣z∣>1 |z| > 1 .

Analytical Intuition.

Imagine a pendulum tethered by a series of elastic bands, each pulling it toward its previous positions with varying intensities defined by the weights ϕi \phi_i . The process Yt Y_t is a conversation with its own past: current values are simply a weighted sum of history plus a random 'nudge' from ϵt \epsilon_t . If the influence of the distant past fades rapidly enough—mathematically guaranteed when roots lie outside the unit circle—the system remains within a stable corridor of variance. If the influence is too strong (roots on or inside the circle), the 'pendulum' loses its anchor, drifting into the infinite abyss of non-stationarity. We are essentially solving for the equilibrium of memory: how much of yesterday should survive into today to ensure the system doesn't explode? By restricting the polynomial Φ(z) \Phi(z) , we constrain the system's 'memory' to be dissipative, ensuring that shocks ϵt \epsilon_t eventually vanish rather than permanently altering the mean level of the series.
CAUTION

Institutional Warning.

Students often conflate the stationarity condition ∣ϕ1∣<1 |\phi_1| < 1 for AR(1) AR(1) with the general condition for AR(p) AR(p) . For higher p p , individual coefficients ϕi \phi_i can exceed unity while the process remains stationary, provided the roots of the characteristic polynomial lie outside the unit circle.

Academic Inquiries.

01

Why is the constant term c c excluded from the stationarity condition?

Stationarity concerns the existence of a finite, time-invariant mean and autocovariance structure. The constant c c merely shifts the level of the process (the mean μ=c/(1−∑ϕi) \mu = c / (1 - \sum \phi_i) ) without affecting the stability of the autoregressive roots.

02

What happens if a root lies exactly on the unit circle?

The process is non-stationary and exhibits a 'unit root.' It is non-mean-reverting, meaning a shock to ϵt \epsilon_t has a permanent impact on the level of the series, often requiring differencing to achieve stationarity.

03

Does a stationary AR(p) AR(p) process always have an MA(∞) MA(\infty) representation?

Yes, if the stationarity condition holds, the polynomial Φ(L) \Phi(L) is invertible, allowing us to write Yt=Φ(L)−1ϵt=∑j=0∞ψjϵt−j Y_t = \Phi(L)^{-1} \epsilon_t = \sum_{j=0}^{\infty} \psi_j \epsilon_{t-j} , where the sequence ψj \psi_j is absolutely summable.

Standardized References.

  • Definitive Institutional SourceHamilton, J. D., 'Time Series Analysis', Princeton University Press.

Related Proofs Cluster.

Advanced

The Matrix Formulation of the General Linear Model: Y = Xβ + ϵ and its Fundamental Assumptions

Master the matrix formulation of the General Linear Model, \( Y = X\beta + \epsilon \), and its fundamental assumptions. Rigorous yet intuitive content for BSc Math/Stats students.

Foundational

Derivation of the Ordinary Least Squares (OLS) Estimator: β̂ = (X'X)⁻¹X'Y

Master the OLS estimator derivation: \( \hat{\beta} = (X'X)^{-1}X'Y \). Explore the geometric orthogonality, matrix calculus, and Gauss-Markov foundations.

Foundational

Proof of Unbiasedness of the OLS Estimator: E(β̂) = β

Master the rigorous proof of OLS estimator unbiasedness, \( E(\hat{\boldsymbol{\beta}}) = \boldsymbol{\beta} \). Understand critical assumptions, geometric intuition, and common pitfalls for robust linear modeling.

Foundational

Derivation of the Variance-Covariance Matrix of the OLS Estimator: Var(β̂) = σ²(X'X)⁻¹

A rigorous derivation of the Variance-Covariance matrix for the OLS estimator, exploring the geometric impact of data configuration on statistical precision.

Institutional Citation

Reference this proof in your academic research or publications.

NICEFA Visual Mathematics. (2026). The Autoregressive (AR) Model: Definition, Properties, and Conditions for Stationarity: Visual Proof & Intuition. Retrieved from https://www.nicefa.org/library/general-linear-models-/the-autoregressive--ar--model--definition--properties--and-conditions-for-stationarity

Dominate the Logic.

"Abstract theory is just a movement we haven't seen yet."

Subscribe for Full ProofsEarly Access