In a forecasting model, where y is a function of x$1,\backslash$times $2,\backslash$times $3,$ and lagged values of y$,$ we have a problem creating forecasts because the model you have estimated includes only contemporaneous values of x$1,\backslash$times $2,$ and $\backslash$times $3.$ Which of the following is not a way to handle this problem?

$$ We could create AR models of each of these variables and use forecasts from these univariate models in our forecasts

$$ We could lag each of these variables by one period

$$ We could lag each of these variables by more than one period

$$ Sample from the historical distribution of values of the variables and insert those values into equation for the out of sample forecast terms.