What do records of actual positions suggest about commodity futures trading success? In the June 2013 version of their paper entitled “Determinants of Trader Profits in Commodity Futures Markets”, Michaël Dewally, Louis Ederington and Chitru Fernando examine actual daily closing positions of energy futures traders to determine how profitability relates to differences in risk taking, trading strategy and information/skill. They assign each trader in their sample to one of eleven categories: refiners, independent producers, pipelines and marketers, large energy consumers, commercial banks, energy traders, hedge funds, households, investment banks and dealers, market makers and others. Using end-of-day open interest for 382 traders reporting positions in NYMEX crude oil, heating oil and gasoline futures markets via the CFTC’s Large Trader Reporting System during June 1993 through March 1997, they find that:
- On average over the sample period, data account for 78%, 68% and 81% of open interest in crude oil, heating oil and gasoline futures contracts, respectively.
- On average over the sample period (not counting intra-day trades):
- Hedger categories lose money (average annualized gross return -4.2%). Traders who are long (short) when hedgers categories are net short (long) outperform those aligned with hedgers.
- Speculator categories make money (average annualized gross return 12.1%). Hedge funds drive this profitability, apparently by trading against hedgers (exploiting hedging pressure), generating an average gross annualized return of 17.8%. Both hedge funds and speculators overall are significantly profitable after reducing average daily profits by 0.03% to 0.05% to account for round-trip transaction fees of $15 per contract.
- Market makers lose money (average annualized gross return -9.0%), not counting their earnings from the bid-ask spread (or, as stated, intra-day trades).
- While less important than hedging pressure, profitability of long positions varies inversely with commodity inventories and directly with price volatility (due to costs of commodity storage).
- After controlling for hedging pressure and storage costs, futures contract prices exhibit no short-term (one-month) momentum.
- On a category by category basis, the shedding/assumption of risk and the taking/providing of liquidity generally explain profitability. There is no evidence that any categories are more informed or skillful than others. However, within categories, some traders exhibit skill (lack of skill) via persistent profits (losses).
- There is no evidence that larger traders do better than smaller ones, but those who change their positions more frequently tend to outperform on a gross basis.
- Traders who hold spread positions tend to outperform those who hold mostly long or short positions. For example, traders who are always half long and half short make on average 10% more per year gross than those who are always all long or all short.
In summary, evidence indicates that energy futures speculators generate good profits primarily by assuming risk from, and providing liquidity to, hedgers.
Cautions regarding findings include:
- Data are 16-20 years old and may not reflect the current commodity futures trading environment.
- As noted in the paper, traders essentially self-classify, potentially introducing inconsistencies.
- Data do not include cash levels or intra-day trades, and profitability calculations mostly ignore transaction fees and bid-ask spreads. Results therefore do not represent net portfolio-level return on capital.
- Statistical tests and associated interpretations assume that the distribution of trader performances is tame. To the extent the distribution is wild, this approach breaks down.
