Analysing Bollinger Bands in Relation to Energy Spreads
The British University in Dubai (BUiD)
This thesis examines the signalling ability of Bollinger Bands when analysed with energy spreads. The study evaluates the profitability of preliminary trading systems over different spread constructions to identify the Bollinger Band signalling efficiency. The test is thus an analysis of the weak form Efficient Market Hypothesis as proposed by Fama (1970). Earlier research has illustrated varied conclusions. The research into Bollinger Bands has illustrated that Bollinger Bands trading systems are unable to yield economic profits and perform worse than other Technical Analysis methods. The spread research, on the other hand, clearly identifies that there are inefficiencies and large potential for automated trading systems, also with Bollinger Bands or comparable methods. This study has found that Bollinger Bands generate inefficient signals, are not a good predictor of spread movements. On average the performance between 1995 and 2009 has generated a loss. On this basis a “contra mean reversing” (trending) approach of the highest loss giving settings has been deeper analysed to find considerable signalling efficiency, thus potential trading profits. The evaluation incorporates conventional energy spreads like the Crack and Frac spreads and created additional spreads. A created spread of the end products of the crack spread, heating oil and unleaded gasoline, generated the best results indicating the need for further research into this spread. The profitability of the contra Bollinger Band system indicates that the Efficient Market Hypothesis should be rejected in the weak form.
DISSERTATION WITH DISTINCTION
bollinger bands, energy spreads, trading systems, efficient market hypothesis