by Cairong Zhao, Ding Qi, Dongsheng Li, Guosheng Hu, Junyao Gao, Kaitao Song, Xiaohua Wang, Zifan Song
Continue on SelfMadeQuant
Suggested foundations for trading research and honest backtests and text and language models in finance. Specialist prerequisites vary by paper.
Use only information available at the simulated decision time.
Separate model selection from the final evaluation.
Understand how text is converted into numerical inputs.
Check whether labels measure the financial construct of interest.
Reconcile signals, executions, positions and P&L.
Include the costs that can erase a paper strategy.
Test whether a text-derived signal survives an honest financial evaluation.
Choose a reproducible strategy, specify the trading rule, and test what remains after costs and estimation uncertainty.