Analysis of Liquidity vs. Sharpe

Questions

Questions:

  • The AiLA products often benefit significantly from that allocations are modelled individually for contracts within the same commodity.
  • However, it has also been observed that applying the aggregated contract weights to the prompt commodity contract often result in only a small impact on performance.
  • How consistent is this effect if applied to different AiLA commodity indices, and are any typical patterns observed?

Approach:

  • Construct a generic approach aggregating the individual contract weights and applying these weights to the prompt commodity futures contract, e.g. rolling front month.
  • Compare the AiLA strategy performance from the individual vs aggregated prompt contract weights, for a set of different AiLA index products.

Individual: SR = 2.3, Risk = 7.8%

Prompt: SR = 1.8, Risk = 9.7%

𝜌(Individual, Prompt) = 93%

Prompt Contract

  • Allocation weights of individual monthly contracts were summed for a given commodity futures, e.g. monthly contact weights for the CME WTI futures market.
  • The aggregated allocation weights were then applied to the prompt contract of the corresponding commodity, which here was typically rolled in the middle of the month before the contract month.
  • The prompt version of the AiLA product was then compared to the original version, i.e. the latter being based on individual contract allocations.
  • The comparison was made for a set of 16 different AiLA products, which do have overlap in terms of asset composition, however, also with significant differences.
  • Due to the netting of long and short weights the prompt version of the portfolio typically tends to allocate a lower percentage of capital than the individual version.

The tendency that the prompt portfolio version allocate a lower percentage of capital depends on the amount of offsetting long and short weights, where a long-only portfolio, for example, would result no difference.

Prompt Contract

  • In terms of product returns, the prompt portfolio version tends to have a higher annual risk than the individual version, which comprises more deferred contracts.
  • The Prompt/Individual ratio of annual risk, where the period 2017 until 2022 was used in the analysis, show values above 1 for all 16 products, however, to a varying degree depending on the portfolio composition.
  • The annual average returns on the other hand do not show such a systematic difference between the prompt and individual portfolios, which results in a difference in terms of Sharpe ratios which follow the annual risk pattern.
  • The prompt vs individual performance correlation is within a range between 80% to 100% for all 16 AiLA products with a median correlation of 95%.

Conclusions

  • A set of 16 AiLA products was used to study the difference between applying the AiLA allocations obtained for individual commodity contracts, to the corresponding contracts vs only the prompt contract, i.e. the rolling front commodity contract.
  • The results indicate a tendency of obtaining a similar performance between the two versions, with the prompt version typically resulting in a slightly lower Sharpe ratio than the individual version.
  • The main reason behind the Sharpe ratio difference appears related to the prompt version typically having a higher annual risk than the individual version, which comprises more deferred contracts.
  • The results suggest that despite a significant value in using separate allocation strategies for individual contracts, i.e. diversity w.r.t. allocation decisions, aggregating the final contract weights and applying those aggregated weights to the prompt contract has a relatively small impact on performance.