Calendar anomalies Non-linearity Market efficiency Asymmetric volatility Rolling windows
This paper investigates two calendar anomalies in an emerging African
market. Both the day of the week and month of the year effects are examined
for Ghana. The latter is an interesting case because i) it operates for only
three days per week during the sample period and ii) the increased focus that
African stock markets have received lately both from academics and
practitioners. We employ rolling techniques to asses the affects of policy and
institutional changes. This allows deviations from the linear paradigm. We
finally employ non-linear models from the GARCH family in a rolling
framework to investigate the role of asymmetries. Contrary to a January
return pattern in most markets, an April effect is found for Ghana. The
evidence also shows the presence of the day of the week effects with
asymmetric volatility performing better than the benchmark linear estimates.
This seasonality though disappears when only the latest information is used
(time-varying asymmetric GARCH). Our approach provides a new
framework for investigating this well-known puzzle in finance.
This is a working paper. It is also available at: http://ideas.repec.org/p/lbo/lbowps/2006_13.html.