One side and two sided filters

Estimation ML/Bayesian estimation
1

What is the main difference between on sided and two sided filters?

For example one sided hp filter versus two sided hp filter.

2

Hi eisamabodia,

we use the one-sided HP filter to detrend data because it is purely backward looking. Thus, at each observation it only assumes current and past states influencing the current observations. This is in line with the backward looking nature of our solution of a DSGE model.
The two-sided HP filter on the other hand assumes that observations at t+k, k>0, influence the current observation t. This is against the way we write our DSGE models in recursive form where only current and past states and shocks influence the current solution.

Best

1 Like
3

Thank you so much for your clear answer.