Newest papers

MintEval: Do LLMs Implement the Trading Strategy You Asked For? A Behavioural-Equivalence Benchmark for Natural-Language-to-Strategy Code

Siyu Wang, Yifan Wang, Yuecheng He · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.03080 · cited by 0 works (OpenAlex) · open-access PDF

Indonesian

Model aktif siliven bahasa yang akurat bergerak dari memproduksi sinyal perdagangan ke menulis kode yang mengeksekusi mereka. Mode kegagalan peran kedua adalah diam: kode yang dihasilkan berjalan, plot backtest, namun logika risiko spesifik yang diidentifikasi oleh pedagang tidak dinilai secara eksekusi. Benchmarks kode yang ada menguji keakuratan fungsional pada tes unit dan perencanaan perencanaan keuangan; tidak mengukur apakah implementasi berperilaku seperti strategi yang diminta. Kami memperkenalkan MintEval, sebuah benchmark di mana strategi referensi dihasilkan secara programatis dari perpustakaan blok bangunan yang dapat dikomposisi, terjemahkan kembali ke dalam instruksi pedagang koloktif, dan diimplementasikan kembali oleh model di bawah tes. Program implementasi yang dihasilkan dan kegagalan diimplementasikan dengan bar pada data pasar dan tugas yang identik bartik, dan

Machine translation (Meta M2M100); the English original follows.

Abstract

Large language models are moving from producing trading signals to writing the code that executes them. The failure mode of the second role is silent: generated code runs, a backtest plots, yet the risk logic that the trader described is not the logic being executed. Existing code benchmarks test functional correctness on unit tests and finance benchmarks test forecasting; neither measures whether an implementation behaves like the strategy that was asked for. We introduce MintEval, a benchmark in which reference strategies are generated programmatically from a library of composable building blocks, back-translated into colloquial trader instructions, and re-implemented by the model under test. Generated and reference programs are executed bar by bar on identical market data and frictions, and compared on their actions rather than on code similarity or profit: alpha is differenced away. MintEval v0 contains 800 tasks on BTCUSDT 15-minute data, stratified by an execution-measured state-span complexity tau that is decoupled from description length. Low-cost models reach a mean ActionMatch of at most 0.544 and reproduce at most 0.087 of tasks exactly; on a stratified subset of 200 tasks a frontier model (Claude Opus 5.5) reaches 0.889 and reproduces 0.575 exactly, yet still fails silently on 0.275 of tasks. Given a menu of building blocks, models identify the strategy almost perfectly, yet 79.2% of the implementations whose specification was read correctly diverge on more than 10% of active bars. The LLM judge of a recent strategy-generation benchmark, applied verbatim, accepts every one of these silent failures.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Benchmark (surveying), Computer science, Correctness, Code (set theory), Unit testing, Implementation

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange