One of our best 2016 MBA theses on residual momentum and investor sentiment

Residual momentum and investor sentiment on the Johannesburg Stock Exchange (JSE)

By: Louis Egbert Viljoen, 2017

Source: http://hdl.handle.net/2263/59799

ABSTRACT:

Momentum has been described as the premier financial market anomaly (Fama & French, 2008), but styles based on this phenomenon tend to suffer intermittent crashes (Barroso & Santa-Clara, 2015). The study investigated a variation on momentum that considers only firm-specific returns, determined from the residual remaining after deducting returns attributable to common risk factors, when selecting portfolio constituents. This prevents concentrated exposure to common risk factors in any one portfolio. The method is known as residual momentum and has shown great promise to improve risk-related returns.

Investor sentiment is another financial market phenomenon and is often explained by means of the same behavioural factors as momentum. The study also considered the effect of investor sentiment on momentum in order to document the effect on the JSE, to shed further light on the driving factors behind the phenomena, and to explore practical investment opportunities.

Equally weighted conventional momentum and residual momentum portfolios were constructed from the largest 160 stocks on the JSE on a quarterly basis over the last 27 years in order to compare the styles’ performances. In addition, momentum returns were compared across different sentiment states, defined based on the level of investor sentiment as proxied by a consumer confidence index orthogonalised to various macroeconomic variables.

Residual momentum was found to provide better risk-adjusted returns than conventional momentum on the JSE. Investor sentiment showed an effect on momentum styles, with residual momentum most profitable following pessimistic formation periods and conventional momentum most profitable following non-pessimistic periods.

Keywords:

  • Anomalies
  • Momentum
  • Investor sentiments

Another one of our best theses for 2016 was investment strategies for OECD and B RICS currencies

By: Blomeyer, Gregg

Source: http://repository.up.ac.za/handle/2263/59882

Abstract:

In this paper a graphical time-series approach was used to analyse style-based investment strategies for currencies. The styles investigated included momentum, volatility and value, and particular focus was given to understanding whether differences exist in the results between the currencies of developed versus emerging countries. The results showed that differences between emerging and developed currencies were statistically significant for each of the styles studied and that the classification of countries’ currencies, as either developed or emerging, was therefore necessary in analyses. Momentum was confirmed to exist in currencies, with a reversion to the mean in the long-term; optimal returns were achieved with the least momentum (quintile five) currencies, using a 10-month look-back period (formation period), three-month look-to period and a two-month holding period. Volatility as a style started out as a particularly good trading strategy, but the results show that the style has been traded-out from around the time of the global financial crisis in 2007 to 2008. Returns from the value style have persisted, with the greatest returns achieved with those currencies most under-valued according to the Big Mac index. The relative strength of the base currency used in the analysis, in this case the U.S. dollar, was found to have a significant impact on the success of the various style-based investment strategies.