Section7 The January Anomaly(1).ppt

The January Anomaly:effects of low share price,transaction costs,and bid-ask bias,Bhardwaj and BrooksJOF Jun,1992,Abstract,Introduction,Share price may explain abnormal January returns after control for firm sizeTransaction costs,degree of neglect,misassessment of risk and infrequent trading might explainThis paper documents that low share price stocks earn abnormal returns in January before transaction costs,When transaction and bid-ask bias are considered,no positive abnormal returns for 1977-1986.5*5 matrix of portfolios based on market value quintiles and pre-specified stock price ranges over 1967-1986Second,derive brokerage commissions and bid-ask spread,Third,buy-and-hold strategyFinally,a more recent time period,1977-1986,1 The January price effect,A.Data and methodologyReturns on 25 portfolios,cross-classified by market valueEquity market value of firm size:S1 to S5Price:P1 to P5RJ1,RJ5 and RJ in January1967-1986,B.Results,January returns exhibit an inverse relation with stock priceReturns seem to have little or no relation to firm sizeF statisticsSupports the contention that the January anomaly is primarily a price effect and not a size effects,C.Discussion,A low price phenomenon,several reasonsSelling pressure and buying pressureTransaction costs,low liquidity Buy-and-hold strategy,2.Transaction Costs,Both before and after transaction-cost returns are examinedA.Estimates of bid-ask spreads and commissionsPanel A of table 2 provides descriptive information on the six bid-ask spread estimate samples,bid-ask spreads are inversely related to share price,commission,B.Estimation of holding period returns,C.Holding period returns results,Before considering transaction costs,lower price stocks appear to outperform higher price stocks only After transaction-cost returns are positively related stock price,D Excess return results,When the time horizon reaches two years,the lower price stocks appear to be underperforming the higher price stocks,E.Supplemental Tests,Same for the period 1977-1986After-transaction-cost excess returns are statistically higher for lower price stocks than higher price stocks in the 1967-1976 periodBut the 1982-1986 results are sufficient to question the prior findings of a small-firm or low-price effects,3.The bid-ask Bias in computed returns,Computed returns on lower price stocks also may be overstated due to a turn-of-the-year bid-ask effectThe bias would be largest for lowest price stocks where bid-ask spreads are known to be largest.,A.Description of SampleB.Distribution of bid,ask and mixed transactionsThe evidence of the likelihood of bids exceeding the likelihood of asks at year-end and the greater probability of asks in early January,C.Estimate of Bid-Ask Bias,4.Conclusions,Over 1967-1988 period,the January anomaly is found to be a low-price phenomenon rather than a small firm effectAfter consideration of transaction cost during 1982-1986 period,the low-price stocks appear to be dominated by high-price stocks,A turn-of-year switch from excess probability of bids in late Dec.to asks in early Jan.induces a positive bid-ask bias in returns at about 1%.The bias is significant for low-price stocksThe January anomaly of low-price stocks outperforming high-price stocks cannot be used to earn abnormal returns,