Online supplement to 'A panel clustering approach to analyzing bubble behavior

This is the online supplement to the working paper 'A panel clustering approach to analyzing bubble behavior', available at https://ink.library.smu.edu.sg/soe_research/2591/This online supplement has six sections. Section A collects together technical lemmas that are used for membership estimation in the first stage. Section B collects the lemmas needed for post-clustering panel estimation and the bubble detection methods, specifically the post-clustering panel t- and J-tests. Section C collects results and proofs for selecting the number of groups. Section D extends the two-stage algorithm and the corresponding post-clustering statistics to the mixed-root panel autoregressive model with purely stationary, unit, and purely explosive roots. Section E overviews experimental designs and reports simulation findings. Section F contains tables.