Bubble, Date-Stamping, Fractional Integration, Long Memory, NSE, Sup-ADF
This study evaluates a single bubble episode in the Nigerian Stock Exchange (NSE) by utilizing monthly data on nominal and real all-share index (ASI) from January 2010 to December 2017. Analysis of data based on Sup Augmented Dickey-Fuller (SADF) test for bubble detection suggests non-existence of a bubble in the NSE between 2010 and 2017. Though there is an indication of one explosive episode in September 2011 at which the DickeyFuller statistic lied above the critical values sequence line. However, it is not a bubble but a short deviation from trend. The study also estimates a time-varying long memory parameter, using a fractionally-integrated autoregressive model to check the robustness of the SADF test and it provided further evidence on the absence of a bubble. These ﬁndings showed that the behaviour of stock prices is not driven by a bubble in the Nigerian Stock Exchange (NSE). The study, therefore recommends a time-to-time bubble diagnostic check on the exchange so that symptoms of a bubble can be early detected and managed to avoid losses that may result from the bust.
CBN Journal of Applied Statistics
Ilyasu, Jamilu and Saba, Ndayezhin D.
"Testing for Single Bubble Episode in the Nigerian Stock Market: an empirical investigation,"
CBN Journal of Applied Statistics (JAS): Vol. 10
, Article 2.
Available at: https://dc.cbn.gov.ng/jas/vol10/iss1/2