LRM with time series data – the static model

Summary

Setup

  • Time series data on a dependent variable yy and k1k1 explanatory variables x2,,xkx2,,xk .
  • y1,,yTy1,,yT and x1,j,,xT,jx1,j,,xT,j for j=2,,kj=2,,k are stationary processes.
  • We formulate a linear regression model

yt=β1+β2xt,2++βkxt,k+εt,t=1,,Tyt=β1+β2xt,2++βkxt,k+εt,t=1,,T

Exogeneity in time series models

  • We say that the xx -variables are exogenous if

E(x)=0,t=1,,TE(x)=0,t=1,,T

  • where E(x)E(x) is the conditional expectation of ytyt given all data on the explanatory variables at all points in time.
  • If the xx -variables are exogenous then

E(x)=β1+β2xt,2++βkxt,k,t=1,,TE(x)=β1+β2xt,2++βkxt,k,t=1,,T

  • and

yt=E(x)+εt,t=1,,Tyt=E(x)+εt,t=1,,T

  • or

εt=ytE(x),t=1,,Tεt=ytE(x),t=1,,T

Static and dynamic models

  • The model is static if only observations made at time tt affect E(x)E(x) . If past values affect E(x)E(x) then the model is dynamic .

Homoscedasticity in time series models

  • We say that the error terms are homoscedastic if

Var(εt|x)=σ2,t=1,,nVar(εt|x)=σ2,t=1,,n

  • Otherwise it they are heteroscedastic.

Autocorrelation

  • We say that the error terms are autocorrelated if, conditionally on xx ,

Cov(εt,εs)0foranytsCov(εt,εs)0foranyts

  • We say that we have no autocorrelation if Cov(εt,εs)=0foranytsCov(εt,εs)=0foranyts

Gauss-Markov assumptions in time series models

  • All data is stationary
  • The explanatory variables are exogenous
  • The error terms are homoscedastic
  • There is no autocorrelation