{"paper_id":"2510.17165","title":"Trading with the Devil: Risk and Return in Foundation Model Strategies","authors":["Jinrui Zhang"],"paper_date":"2025-10-20","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2510.17165v1","page_url":"https://thequant.space/flowcharts/trading-with-the-devil-risk-and-return-in-foundation-model/","scores":{"empirical_rigor":6,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes an extension to CAPM that separates systematic risk from a shared foundation model and uses uncertainty disentanglement (e.g., MC dropout) to measure epistemic risk.","quote":"In this paper, we propose an extension to the Capital Asset Pricing Model (CAPM) that disentangles the systematic risk introduced by a shared foundation model—potentially capable of generating alpha if the underlying model is genuinely predictive—from the idiosyncratic risk attributable to custom fine-tuning","summary":"Foundation models - already transformative in domains such as natural language processing - are now starting to emerge for time-series tasks in finance. While these pretrained architectures promise versatile predictive signals, little is known about how they shape the risk profiles of the trading…","strategy":"TimesFM-v2"},"assumptions":{"status":"available","items":[{"assumption":"Aleatory Collapse: observable aleatory uncertainty is arbitraged away, so uncertainty estimator converges to epistemic uncertainty.","type":"market","stated":"explicit","quote":"Assumption 3.1. Aleatory Collapse. In a competitive market, any readily observable aleatory uncertainty is swiftly arbitraged away, due to its’ correlation with return."},{"assumption":"The empirical study restricts to single‑asset trading to avoid portfolio aggregation effects.","type":"execution","stated":"explicit","quote":"we restrict our scope to single-asset trading. This choice avoids the added complexity of portfolio aggregation effects"}]},"datasets":{"status":"available","items":[{"name":"U.S. equity NBBO","provider":"Interactive Brokers API","kind":"order-book","access":"commercial","period":"2023-01-03 to 2025-01-31","universe":"S&P 500 and Russell 2000 constituents","quote":"US Equity: One-second aggregates of the National Best Bid and Offer (NBBO) for a selection of S&P 500 and Russell 2000 constituents, collected through the ibkr API"},{"name":"Binance crypto BBO","provider":"Binance exchange","kind":"order-book","access":"public","period":"2023-01-01 to 2025-01-31","universe":"FDUSD‑quoted pairs","quote":"Cryptos: Best Bid and Offer quotes for FDUSD‑quoted pairs on Binance","known_id":"binance"}],"known":[{"id":"binance","name":"Binance","provider":"Binance","kind":"market-prices","access":"public","quote":"Data: We draw upon high-frequency market data from both U.S. equities and the Binance cryptocurrency exchange, detailed in Table 1: • US Equity: One-second aggregates of the National Best Bid and Offer (NBBO) for a selection of S&P 500 and Russell 2000 constituents, collected through the ibkr API1.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":"U.S. equities","asset_class":"equities","frequency":"intraday"},"data_resolution":"intraday-bars"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":false,"detail":null,"quote":"Our backtest assumes orders are filled at the next tick’s best ask—a convention frequently described as a tick‑by‑tick backtest or immediate exec","source":"front_matter","rule":"R3","candidates":{"front_matter":"no","evidence_card":"no","evidence_card_quoted":true},"card_detail":null,"card_quote":"Our backtest assumes orders are filled at the next tick’s best ask—a convention frequently described as a tick‑by‑tick backtest or immediate exec","meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Isolating epistemic and aleatory risk yields deeper insights into performance limits, model degradation, and refinement avenues for foundation‑model strategies.","kind":"theoretical","quote":"Our experiments show that isolating these distinct risk factors yields deeper insights into the performance limits of foundation‑model‑based strategies, their model degradation over time, and potential avenues for targeted refinements."},{"finding":"Foundation Sharpe ratios (PML slopes) are unstable for conventional baselines but more evident for pretrained architectures where shared signals dominate risk‑return.","kind":"performance","quote":"of “shared risk” is not universally applicable—rather, it becomes most evident in pretrained architectures where common signals genuinely dominate the risk–return profile."},{"finding":"A rolling‑window analysis shows a gradual decay of alpha as foundation‑model edges become widely recognized, with observed Sharpe ratios falling below theoretical upper bounds.","kind":"performance","quote":"... the overall downward trend remains consistent, underscoring a gradual but persistent decay in alpha once a foundation model’s edge becomes widely recognized."}],"reported_performance":{"strategy":"TimesFM-v2","sharpe":3.87,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"walk-forward","headline_out_of_sample":true}},"independent_evidence":{"status":"available","supports":[{"paper":{"paper_id":"2601.13770","title":"Look-Ahead-Bench: a Standardized Benchmark of Look-ahead Bias in Point-in-Time LLMs for Finance","paper_date":"2026-01-20","authors":["Mostapha Benhenda"],"empirical_rigor":6,"page_url":"https://thequant.space/flowcharts/look-ahead-bench-a-standardized-benchmark-of-look-ahead-bia/"},"question":"Do foundation‑model strategies exhibit alpha decay over time?","this_paper_finding":"A rolling‑window analysis shows a gradual decay of alpha as foundation‑model edges become widely recognized, with observed Sharpe ratios falling below theoretical upper bounds.","other_paper_finding":"Standard foundation models show severe alpha decay, exceeding –15 percentage points between the in‑sample and out‑of‑sample periods, indicating strong look‑ahead bias.","other_paper_quote":"standard LLMs exhibit significant lookahead bias, with alpha decay exceeding -15 percentage points between periods, while Pitinf models maintain stable performance.","note":"Both report decay: A shows gradual alpha loss with Sharpe below theoretical bounds; B shows a –15 pp drop from in‑sample to out‑of‑sample, supporting A.","confidence":0.748}],"contradicts":[],"qualifies":[],"compared_with":18,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:24:01Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}