LLM agents (hawkish, dovish, debate) outperform a deterministic z-score rule agent in Sharpe ratio for commodity ETF portfolios by 0.04-0.044, with advantage concentrated in the soft-landing sub-period and preserved up to 30bp trading costs.
arXiv preprint arXiv:2603.11408 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
q-fin.PM 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Macro Economists in the Machine: A Multi-Agent LLM Framework for Commodity-Related ETF Portfolio Construction
LLM agents (hawkish, dovish, debate) outperform a deterministic z-score rule agent in Sharpe ratio for commodity ETF portfolios by 0.04-0.044, with advantage concentrated in the soft-landing sub-period and preserved up to 30bp trading costs.