FFA hedging in the supramax segment : how alterations of the Baltic supramax index have affected hedging efficiency

Abstract

This thesis studies how altering the composition of the Baltic Supramax Index (BSI) affects the hedging efficiency of forward freight agreements (FFA) traded with the index as underlying. We evaluate the hedging efficiency using hedged portfolios with both minimum variance hedge ratios and naïve hedge ratios, both within and across subperiods between 2006 and 2018. Bootstrapping techniques and bias-corrected and accelerated confidence intervals are utilised to investigate if the hedging efficiency is affected to a statistically significant degree when the composition of the index is revised. We find that forward freight agreements can significantly reduce the volatility of freight rates. However, we find no evidence suggesting that reduced weight of a constituent route in the underlying index induces decreased hedging efficiency, or vice versa. Nor do we find that overall hedging efficiency decreases when more routes are added. The cointegrated relationship between individual routes and the FFA time series seems to make changes to the index irrelevant with regard to hedging efficiency. The thesis provides a basis for further research of the hedging efficiency of freight derivatives. Primarily, the topic of this thesis should be revisited when data points from more dimensions of the shipping cycle become available. Moreover, the impact of changes in the underlying asset to real market participants is an interesting continuation. We believe our findings are especially important to the producers of freight indices, as their relevance in terms of hedging efficiency is paramount in order to secure volume and quality in the derivatives market. For charterers, shipowners, and other market participants, our findings are interesting with regards to risk management, specifically in understanding how alterations of an underlying asset have historically affected hedging efficiency. The thesis supplements the rather limited literature on changes in hedging efficiency when the structure of the underlying asset is altered. First, because no such research has been conducted on the FFA market in its current form in recent years. Secondly, because we access trial data of the 10TC FFA before it went live, we can study the differences between overlapping time series of FFA prices with different versions of the BSI as underlying – not only those separated by the date of their alteration. This allows us to compare the hedging performance of two FFAs isolated from time-varying effects. Furthermore, it allows us to evaluate the effect of basis risk caused by differing technical specifications and that caused by a geographical diversification effect due to the addition of more constituent routes.