Skip to contents

High-performance implementation of the Modified HP Filter for decomposing time series into trend and cyclical components. Based on the methodology of Choudhary, Hanif & Iqbal (2014) which uses generalized cross-validation to automatically select the optimal smoothing parameter lambda.

Details

The standard Hodrick-Prescott (1997) filter decomposes a time series \(y_t\) into trend \(g_t\) and cycle \(c_t\) components by minimizing:

$$\sum_{t=1}^{T}(y_t - g_t)^2 + \lambda \sum_{t=1}^{T-2}[(g_{t+2} - g_{t+1}) - (g_{t+1} - g_t)]^2$$

where \(\lambda\) is the smoothing parameter that controls the trade-off between trend smoothness and cycle fit.

The Modified HP Filter (McDermott, 1997) selects \(\lambda\) optimally using generalized cross-validation (GCV). The GCV criterion is:

$$GCV(\lambda) = \frac{SSR(\lambda)}{T} \left[ 1 + \frac{2}{T\lambda} \right]$$

where \(SSR(\lambda)\) is the sum of squared residuals. The optimal \(\lambda\) minimizes this GCV criterion.

Main Functions

References

Choudhary, M.A., Hanif, M.N., & Iqbal, J. (2014). On smoothing macroeconomic time series using the modified HP filter. Applied Economics, 46(19), 2205-2214.

Hodrick, R.J., & Prescott, E.C. (1997). Postwar US business cycles: An empirical investigation. Journal of Money, Credit and Banking, 29(1), 1-16.

McDermott, C.J. (1997). Note on the modified Hodrick-Prescott filter. IMF Working Paper No. 97/108.

Author

Maintainer: Muhammad Yaseen myaseen208@gmail.com (ORCID)

Other contributors: