Abstract
Since the collapse of the Metallgesellschaft AG due to hedging losses in 1993, energy practitioners have been concerned with the ability to hedge long-dated linear and non-linear oil liabilities with short-dated futures and options. This paper identifies a model-free non-parametric approach to extrapolating futures prices and implied volatilities. When we expand the analysis to implementing hedge portfolios for long-dated futures or option contracts over the time period 2007–2017, we utilize the useful benchmark of hedge ratios arising from Schwartz and Smith. With respect to the empirical consequences of hedging long-dated futures and options with their short-dated counterparts, we find that the long-term tracking errors are, on average, quite close to zero, but there is increasing risk entailed in attempting to do so, as the hedge-tracking errors for both futures and option contracts increase with time-to-maturity.
Highlights
Edwards and Canter (1995) summarize the misfortunes of the German conglomerateMetallgesellschaft AG:“In late 1993 and early 1994 MG Corporation, the U
The straightforward conclusion of this section is the ease and feasibility of obtaining a proxy for long-dated futures and option contracts using their short-term counterparts, especially when one takes advantage of the smoothing of these curves using the asymptotic condition imposed on the slopes of the futures and option contracts
On average, the long-term errors are quite close to zero, there is increasing risk entailed in attempting to do so, as the estimation errors for both futures and option contracts increase with time-to-maturity
Summary
A component of the motivation for our paper pertains to the matter of numerous agents’ need to hedge long-dated futures contracts—i.e., beyond those of the “liquidity term” provided by active futures contracts. If such liquidity terms are defined as T ≤ 2 years, it is clear that real assets in the oil sector have economic lives well beyond these liquidity terms. With respect to the empirical conclusions of our work pertaining to hedging longdated futures and options with their short-dated counterparts, we find that the long-term tracking errors are, on average, quite close to zero, but there is increasing risk entailed in. We posit the need to hedge longer-dated exposures using short-dated instruments T ≤ 2 years
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
