GTBpy Documentation

GTBpy is a toolbox that implements bubble tests from the following papers:

• PWY (SADF test): Phillips, Wu, and Yu (2011)

• PSY (GSADF test): Phillips, Shi, and Yu (2015)

• WHL test: Whitehouse, Harvey, and Leybourne (2022)

The package also includes two tests for evaluating forecast performance:

• Equal MAFE and MSFE test: Harvey, Leybourne, and Newbold (1997)

• Forecast encompassing test: Harvey, Leybourne, and Newbold (1998)

Additionally, GTBpy introduces a novel method for constructing a search index using Google Trends data. This method is inspired by the following study:

• Google Trends: Stig V. Møller, Thomas Pedersen, Erik C. M. Schutte, Allan Timmermann (2023)

GTBpy is a work in progress, and a lot remains to be done before we reach version 1.0. The source code is available on GitHub.

Contents:

• GTBpy
  • bubble_indicators module
  • forecast_functions module
  • google_trends_functions module

Credits

GTBpy was created by Mohammad Ebrahimi.