Trading Strategy Optimization via Textual Gradient

Chaoqun Yang; Qian Wang; Fengbin Zhu; Xinyu Lin; Bingsheng He; Roger Zimmermann; Tat-Seng Chua

cs.AIOct 2, 2026

Trading Strategy Optimization via Textual Gradient

Authors: Chaoqun Yang, Qian Wang, Fengbin Zhu, Xinyu Lin, Bingsheng He, Roger Zimmermann, Tat-Seng Chua

Organizations: National University of Singapore

Abstract

Quantitative trading strategy design aims to discover trading programs from historical data that remain effective in future markets, which can be viewed as a black-box program optimization problem. LLM-based textual gradients offer a promising approach by providing explicit optimization directions for iterative strategy refinement. However, directly applying textual gradients faces two challenges: (1) optimization is myopic, underutilizing experience from previous evaluations; and (2) aggregate backtest feedback overlooks temporal robustness, potentially favoring strategies that perform well only in specific market periods. To address these challenges, we propose TradeGrad, an experience-guided textual-gradient framework for robust trading strategy optimization. TradeGrad leverages accumulated optimization experience to estimate textual gradients and employs multi-scale revisions for both strategy exploration and refinement. It further introduces the Cross-Period Robust Objective (CPRO), which emphasizes performance in unfavorable historical periods to promote temporal robustness. Experiments on cross-sectional and time-series strategy design in Chinese A-share and U.S. equity markets show that TradeGrad achieves the best in-sample and out-of-sample performance across all four settings. Notably, its Chinese cross-sectional strategy achieves 27.99% annualized return, 12.19% maximum drawdown, and a Sharpe ratio of 1.63, approximately 68% higher than the CSI 300 benchmark. Further analyses validate the proposed components and show consistent improvements in both in-sample and out-of-sample performance throughout optimization. The code is available at https://github.com/transcend-0/TradeGrad.

Figures & tables

Figure 1: (a) Trading Strategy as a Program. (b) Optimization for Generalizable Trading Strategy.

Figure 2: Overview of TradeGrad. TradeGrad iteratively samples archived strategies, estimates textual gradients from backtest feedback and optimization experience, updates strategies at varying revision scales, and evaluates and archives the resulting strategies.

		CSI 300					S&P 500
Setting	Method	IS Score $\uparrow$	OOS Score $\uparrow$	AR $\uparrow$	MDD $\downarrow$	SR $\uparrow$	IS Score $\uparrow$	OOS Score $\uparrow$	AR $\uparrow$	MDD $\downarrow$	SR $\uparrow$
Market Index Reference		1.63	45.01	17.69	15.66	0.97	5.44	39.38	20.76	18.90	1.26
Cross-Sectional Strategies
Classical	Momentum	0.52	9.19	5.76	28.49	0.35	9.02	2.37	-15.35	31.13	-1.58
	Reversal	0.28	14.51	17.80	28.59	0.74	5.12	26.76	0.56	11.59	0.11
	Volatility	7.56	60.69	18.59	8.74	1.41	4.41	6.93	-9.38	23.79	-0.79

Table 1: Overall performance comparison. Results are averaged over three runs; standard deviations are reported in Appendix A.3 . Best and second-best results are bolded and underlined, respectively.

Method	IS Score $\uparrow$	OOS Score $\uparrow$	AR $\uparrow$	MDD $\downarrow$	SR $\uparrow$
Full Method	$\mathbf{10.69}_{\pm 1.06}$	$\mathbf{65.65}_{\pm 12.15}$	$\mathbf{27.99}_{\pm 4.19}$	$12.19_{\pm 4.65}$	$\mathbf{1.63}_{\pm 0.17}$
w/o Memory	$4.56_{\pm 0.42}$	$35.62_{\pm 7.07}$	$11.78_{\pm 1.73}$	$16.09_{\pm 5.97}$	$0.80_{\pm 0.08}$
w/o Sampling	$\underline{8.32}_{\pm 0.85}$	$46.72_{\pm 10.68}$	$12.98_{\pm 2.03}$	$\underline{11.35}_{\pm 4.38}$	$0.94_{\pm 0.10}$
w/o MSR (Diff)	$5.79_{\pm 0.54}$	$46.88_{\pm 12.10}$	$14.87_{\pm 1.14}$	$13.04_{\pm 5.14}$	$0.96_{\pm 0.10}$
w/o MSR (Rewrite)	$7.42_{\pm 0.77}$	$\underline{55.74}_{\pm 10.01}$	$\underline{15.98}_{\pm 2.39}$	$\mathbf{8.86}_{\pm 3.48}$	$\underline{0.98}_{\pm 0.11}$

Table 2: Ablation of experience-guided textual-gradient estimation. Results are mean

\pm

standard deviation over three runs. Best and second-best means are bolded and underlined, respectively.

Objective	IS OV $\uparrow$	IS Score $\uparrow$	OOS Score $\uparrow$	AR $\uparrow$	MDD $\downarrow$	SR $\uparrow$
CPRO	$10.69_{\pm 1.06}$	$\mathbf{10.69}_{\pm 1.06}$	$\mathbf{65.65}_{\pm 12.15}$	$\mathbf{27.99}_{\pm 4.19}$	$\mathbf{12.19}_{\pm 4.65}$	$\mathbf{1.63}_{\pm 0.17}$
CPRO w/o Freq Gate	$10.99_{\pm 1.07}$	$\underline{8.10}_{\pm 0.79}$	$\underline{40.24}_{\pm 8.68}$	$10.43_{\pm 1.51}$	$\underline{12.66}_{\pm 4.81}$	$0.73_{\pm 0.07}$
Mean-Period Objective	$37.72_{\pm 4.18}$	$4.40_{\pm 0.40}$	$23.37_{\pm 5.06}$	$6.08_{\pm 0.93}$	$18.70_{\pm 7.13}$	$0.44_{\pm 0.05}$
Aggregate Objective	$5.11_{\pm 0.56}$	$0.74_{\pm 0.07}$	$35.63_{\pm 7.64}$	$\underline{13.12}_{\pm 1.96}$	$16.89_{\pm 6.38}$	$\underline{0.77}_{\pm 0.09}$

Table 3: Ablation of the CPRO. IS OV reports each variant’s own objective value and is not directly comparable across rows. Results are mean

\pm

standard deviation over three runs. Best and second-best means are bolded and underlined, respectively.

Figure 3: Effect of different hyperparameters: Memory window for summarizing optimization experience, sample size for textual-gradient estimation, and probability of large-scale revision.

Figure 4: Optimization trajectories of different methods. Points denote the strategies generated at each optimization step, and curves show their 10-step moving averages (MA(10)).

Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

	CSI 300					S&P 500
Method	IS Score	OOS Score	AR	MDD	SR	IS Score	OOS Score	AR	MDD	SR
Cross-Sectional Strategies
Island Algorithm	0.02	0.52	2.01	11.26	0.06	1.12	0.13	3.37	15.03	0.44
QuantEvolve	0.31	0.94	0.84	11.04	0.02	0.03	0.03	4.02	18.54	0.24
Vanilla TextGrad	0.01	0.97	1.21	11.50	0.04	0.10	0.74	2.23	10.09	0.27
Ours	1.06	12.15	4.19	4.65	0.17	2.67	8.39	0.94	1.50	0.11

Appendix

Table 4: Standard deviations of the main results over three independent runs.

Figure 5: Evolution of a final strategy along its ancestral lineage. Labels indicate program IDs and IS Scores.

Algorithmic Trading

Algorithmic trading uses computational models to generate signals, allocate portfolios, and execute orders across financial markets.

See all 149 papers

EVOQUANT: Self-Evolving Verifier-Guided Strategy Optimization for Robust Quantitative Trading
Jul 14, 2026Jie Mao, Changlun Li, Xiang Li +6Large Language Model-Guided Optimization Language Model Self-Improvement
The Hong Kong University of Science and Technology (Guangzhou) · Paradoox AI Research
Strat-LLM: Stratified Strategy Alignment for LLM-based Stock Trading with Real-time Multi-Source Signals
May 7, 2026Wenliang Huang, Zengyi YuLLM Evaluation LLM Alignment
School of Economics Zhejiang University of Technology Hangzhou, China · Faculty of Education East China Normal University Shanghai, China
EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents
Sep 15, 2026Sehee Kim, Yumin Choi, Minki Kang +1LLM Agent Self-Improvement Prompt Optimization
KAIST

Jul 14, 2026J/K move · Enter open · S save

EVOQUANT: Self-Evolving Verifier-Guided Strategy Optimization for Robust Quantitative Trading

Jie Mao, Changlun Li, Xiang Li, Qiqi Duan+5

The Hong Kong University of Science and Technology (Guangzhou) · Paradoox AI Research

Large Language Model-Guided Optimization Language Model Self-Improvement Algorithmic Trading

Open paper AskPDF arXiv

Quantitative strategy optimization remains largely manual, requiring domain experts to identify weak signals, tune risk-control rules, and repeatedly validate iterative revisions. Large language models can accelerate this process, but directly relying on them to rewrite trading strategies often introduces hallucinated edits, strategy drift, and backtest overfitting. We propose EVOQUANT, a self-Evolving Verifier-guided framework for strategy Optimization in Quantitative trading. Our method utilizes LLMs to deeply diagnose performance bottlenecks, generates semantically controlled candidate edits, selects the best strategy through a multi-stage verification pipeline, and distills optimization experience into reusable knowledge for continual self-improvement. We evaluate our method using seven representative strategies: four from the A-share market and three from the Crypto market. Experimental results show that our method significantly improves the Sharpe ratio across all tested strategies: the average test Sharpe increases from -0.298 to 0.538, and the best-performing strategy achieves a 199% relative improvement. Ablation studies and stress tests under stricter conditions further validate the effectiveness and robustness of the framework. Overall, this work transforms quantitative strategy optimization from costly manual trial and error into an automated and verifiable iterative paradigm, offering a new path for applying large language models to financial strategy research.

Abstract

Algorithmic Trading

Explore similar work

EVOQUANT: Self-Evolving Verifier-Guided Strategy Optimization for Robust Quantitative Trading

Strat-LLM: Stratified Strategy Alignment for LLM-based Stock Trading with Real-time Multi-Source Signals

EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents