Kaluba, Mwangele2022-01-052022-01-052021https://hdl.handle.net/10539/32598A research report submitted in partial fulfilment of the requirements for the degree of Master of Commerce (Business Finance) to the Faculty of Commerce, Law and Management, School of Economics and Finance, University of the Witwatersrand, Johannesburg, 2021Strategies based on return continuation have been shown to return a premium unexplained by common risk factors. These strategies are collectively called momentum. Momentum strategies experienced dramatic losses following the volatility episode during the 2008 and 2009 financial crisis. This study therefore tests if volatility and/or market state have any explanatory power for momentum payoff through time. This is the first study to examine the time-varying nature of momentum payoff in South Africa. Previous studies have focused on cross-sectional tests of momentum. These studies are primarily concerned with the existence of momentum on the JSE, the strength of the anomaly and any interaction with other known anomalies. Conversely, this study tests if momentum payoffs through time change as lagged values of volatility change, where volatility is simply the lagged 12-month market volatility. The study also examines whether momentum payoff is affected by the state of the market. Market state is defined using the return of the 6-month market return. A positive market state is one where the prior 6-month return has not been negative. The converse will mean that the market is in a negative state. As a robustness measure, this study considers both equal weighting and value weighted strategies on lookback periods covering 3, 6, 9 and 12 months of prior returns. Furthermore, the study also includes size balanced momentum as well as different window periods for defining volatility and market state. For the JSE, it would appear that when volatility increases, this is not followed by reduced momentum payoffs. This is true regardless of weighting scheme or lookback period. Furthermore, even when momentum is considered on size balanced portfolios, the results do not change across the board. Consequently, momentum strategies that reverse to a loser minus winner payoff in periods of high volatility do not outperform a standard momentum strategy. The reasons for the insignificance of volatility are unclear. One potential factor could be the consistent drop in the payoff to momentum. The changes in volatility simply do not follow this overall trend. In addition, on closer inspection, when lagged volatility increased in 2008, the payoff to momentum did not decrease immediately following this. In fact, the worst return of the period came immediately before lagged volatility started to increase. These results make momentum in South Africa an even greater mystery than before. The results of this paper are an indication that although momentum has been known to exist in multiple countries, the factors that drive momentum are different in each country. Further research can aid to clarify the momentum anomaly in South Africa, with one potential avenue being momentum and liquidityenThesis