Time‑series momentum buzz
A high‑engagement thread broke down time‑series momentum—one of hedge funds' go‑to methods—with a 23‑page paper and implementation notes, and a follow‑up 'To Trend or to Revert' post framed momentum vs mean‑reversion from a portfolio perspective. Both are being shared as turnkey research blueprints for quant projects. (x.com) (x.com)
The Journal of Financial Economics article by Tobias Moskowitz, Yao Hua Ooi and Lasse Heje Pedersen documented persistent trend effects across 58 liquid futures and forwards and reported positive predictability at horizons of roughly one to twelve months. (papers.ssrn.com) The Quant Science thread shared an executable 20‑day example (momentum = price[-1] / price[-20] - 1) and recommended yfinance for data ingestion while linking to paid educational products alongside the walkthrough. (threadreaderapp.com) Quant Science’s public code footprint includes a vectorbt_backtesting repository and other notebooks that students can fork to reproduce basic signal construction and volatility‑targeting wrappers. (github.com) Jungle Rock’s posts reframed signal choice as a portfolio tradeoff between “slow” trend capture and “fast” mean‑reversion detection, arguing trend allocations have historically outperformed in high‑volatility, non‑QE regimes and pointing followers to their Wifey ETF dashboard that overweights trend exposure. (threadreaderapp.com) An EDHEC‑linked note and conference listings cite an author who formalized the “trend vs revert” portfolio perspective in May 2020 and referenced industry scale — managed futures assets above roughly $300 billion — as context for practical portfolio implementations. (coursesidekick.com) Open‑source reproductions and project repos (examples on GitHub that reproduce Moskowitz et al.’s pipeline and an equities‑focused production‑grade TSM repo) provide turnkey data pipelines, survivorship‑bias filters, and volatility normalisation examples that are ready to be adapted for student projects. (github.com) Academic and practitioner reviews highlight two operational risks for such blueprints: partial reversals at longer horizons documented in the original study and turnover/transaction‑cost leakage that can materially shrink gross edge in real execution. (papers.ssrn.com)
