Existence of the Semi-month Effect on Stock Returns at the Nairobi Securities Exchange

Abstract

Thesemi-montheffectpurportsthatstockreturnsin the firsthalfofthe tradingmonthis significantly higherthanthesecondhalfof the month. The objectiveof this studywas toestablishwhetherthere isexistenceof thesemimontheffectsonstockreturnsattheNairobi SecuritiesExchange(NSE).Thisstudy useda descriptiveresearchdesign.Thepopulationof interestwasallthelistedcompaniesforequitystocksattheNSEasatDecember2013.The samplepopulationwascompanieslistedcontinuouslyintheNSE- 20shareIndex.Thedataused inthisstudy constitutedindividualsharepricesfortheperiodstartingfromJanuary2011to December2013.Thedailyreturnforeachfirmstockwasdeterminedasthechangeinmarket priceofthestockplusanydividendreceived, expressedasafractionofinitialstockprice. Descriptivestatisticsused weremeanandstandarddeviation. Regressionanalysiswasusedtoassess the semi-month effect on stock returns. Themeanreturnsforthefirsthalfofthemonthforthe3yearperiodis0.0674%with a standard deviation of 0.7645%. For thesecondhalfthemean return was0.0304% with a standard deviation of 0.6091%.Thecoefficientβ1in the regression model was-0.000369with ap-valueof0.465038.The difference between the returns of the second half and the first half of the month had a negative effect on stock returns. Thep-value for the significance of the regression coefficient 0.465038 isgreaterthan0.05thusthecoefficientisnotsignificant.Fstatisticwas found to be 0.534291witha p-valueof0.465038.Thispvalueisalsogreaterthan0.05hencethe overallfitofthemodelispoor.Fromthefindingsofthisstudy thereisnoevidenceofthesemimontheffectonstockreturnsintheNSE. Therefore investors should not expect returns from one half of the month to be statistically different from returns in the other half.