Are commodity indices after their launch still rewarding for traditional portfolios?

This thesis examines the diversification benefits of 1st, 2nd, and 3rd generation commodity indices for a traditional portfolio. A broad set of asset allocation strategies is employed in order to augment each benchmark portfolios that consists of 4 stock and 4 bond indices, with one of the seven commodity indices. This study examines the weekly return observations from February 2000 until May 2018, but due to backfilled data points, the main focus is on the actual performance starting in December 2010 until the end. The in-sample analyses, such as the spanning tests, and the other allocation strategies seem to exaggerate the benefits of commodities for portfolios as the out-of-sample fail to reproduce the results. First, 1st and 2nd generation indices fail to add value to a portfolio for the latest period. Second, heterogeneous performances of 3rd generation indices indicate that only a selection, but not all are superior to the previous generations or able to outperform the benchmark portfolio. Consequently, opposed to most of the previous studies, investments in 3rd generation commodity indices do not always yield beneficial performances.
Keywords: Generation commodity indice, Asset allocation strategies, Spanning tests, Out-of-sample analysis, In-sample analysis,