Partial Cointegration

A collection of time series is partially cointegrated if a linear combination of these time series can be found so that the residual spread is partially autoregressive - meaning that it can be represented as a sum of an autoregressive series and a random walk. This concept is useful in modeling certain sets of financial time series and beyond, as it allows for the spread to contain transient and permanent components alike. Partial cointegration has been introduced by Clegg and Krauss (2016) <>, along with a large-scale empirical application to financial market data. The partialCI package comprises estimation, testing, and simulation routines for partial cointegration models in state space. Clegg et al. (2017) <> provide an in in-depth discussion of the package functionality as well as illustrating examples in the fields of finance and macroeconomics.


Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.1.0 by Jonas Rende, 7 months ago

Browse source code at

Authors: Matthew Clegg [aut], Christopher Krauss [aut], Jonas Rende [cre, aut]

Documentation:   PDF Manual  

GPL-2 | GPL-3 license

Imports zoo, parallel, ggplot2, grid, MASS, TTR, data.table, glmnet, methods, Rcpp, FKF

Depends on partialAR

Suggests egcm, knitr, rmarkdown

Linking to Rcpp

See at CRAN