{"paper_id":"2407.17645","title":"Hopfield Networks for Asset Allocation","authors":["Carlo Nicolini","Monisha Gopalan","Jacopo Staiano","Bruno Lepri"],"paper_date":"2024-07-24","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2407.17645v1","page_url":"https://thequant.space/flowcharts/hopfield-networks-for-asset-allocation/","scores":{"empirical_rigor":6,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes applying modern Hopfield networks to portfolio optimization, introducing two architectures – Hopfield Pooling and Hopfield Encoder – as alternatives to LSTM and Transformer models.","quote":"We present the first application of modern Hopfield networks to the problem of portfolio optimization.","summary":"We present the first application of modern Hopfield networks to the problem of portfolio optimization. We performed an extensive study based on combinatorial purged cross-validation over several datasets and compared our results to both traditional and deep-learning-based methods for portfolio…","strategy":"Hopfield Pooling (HOP-POOL)"},"assumptions":{"status":"available","items":[{"assumption":"Asset returns are computed as log‑returns, assumed to be more Gaussian‑like.","type":"data","stated":"explicit","quote":"We compute asset returns as log-returns rti = log Pt,i −log Pt−1,i instead of simple returns, as they are leaner to work with and tend to be better distributed towards a Gaussian distribution (Campbell et al., 1998)."},{"assumption":"Only long‑only portfolios are considered; weights are non‑negative and sum to one.","type":"market","stated":"explicit","quote":"For long-only portfolios asset weights are all positive and sum to 1."},{"assumption":"Transaction costs and management fees are ignored in the reported results.","type":"execution","stated":"explicit","quote":"Importantly, in order to focus on specific behaviours of our methods, in the following Results Section we have ignored transaction costs and management fees."},{"assumption":"Cross‑validation uses a purging and embargo period of 21 days (≈ one business month).","type":"statistical","stated":"explicit","quote":"he purging and embargoing periods we have set 21 days, corresponding approximately to one business month."},{"assumption":"Within each CV fold the training set is split 80%/20% for training and validation.","type":"statistical","stated":"explicit","quote":"Within each cross-validation training fold we further subdivided the training set into two distinct consecutive sets in proportions of 80% training and 20% validation, in order to keep track of the loss function and trigger early stopping."},{"assumption":"Asset weights sum to one, allowing negative weights only in short‑portfolio settings (not used here).","type":"model","stated":"explicit","quote":"One could also deal with short portfolios, having positions with the only constraint of summing to 1, thus allowing negative weights."}]},"datasets":{"status":"available","items":[{"name":"ETF","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":"2006-02-06 to 2023-12-29","universe":"4 ETFs (AGG, VIX, DBC, VTI)","quote":"The first dataset (named ETF in the text) consists of the four ETFs used by Zhang et al. (2020), although over a different temporal interval. The period analyzed spans from February 6th 2006 to December 29th 2023.","known_id":"yahoo-finance"},{"name":"STOCKS","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":"2003-01-02 to 2022-12-30","universe":"66 Nasdaq securities","quote":"As the second dataset (named STOCKS in the text) we have obtained the spot prices of 66 securities from the Nasdaq index, from January 2nd 2003 to December 30th 2022.","known_id":"yahoo-finance"},{"name":"FAMAFRENCH","provider":"Fama‑French data library","kind":"market-prices","access":"public","period":"1973-01-01 to 2023-11-30","universe":"10 industry portfolios","quote":"As the intermediate size dataset (named FAMAFRENCH in the text), we considered the daily returns from 10 industry portfolios starting from January 1st 1973 until November 30th 2023. Data were obtained from the Fama‑French data library.","known_id":"french-library"},{"name":"COMMODITIES","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":"2003-01-02 to 2023-12-29","universe":"19 commodity futures (gold to gasoline and soybean)","quote":"from 2003 to 2023. The dataset named COMMODITIES included 19 commodities from gold to gasoline and soybean.","known_id":"yahoo-finance"}],"known":[{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"All the data have been obtained from Yahoo Finance.","mentions":2,"used_in":1},{"id":"french-library","name":"Kenneth French Data Library","provider":"Kenneth R. French","kind":"market-prices","access":"public","quote":"Data were obtained from the Fama-French data library.2 Finally, we have run experiments on commodities future prices downloaded from Yahoo finance,3 from 2003 to 2023.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":"multi-asset","frequency":"daily"},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":false,"detail":null,"quote":"Importantly, in order to focus on specific behaviours of our methods, in the following Results Section we have ignored transaction costs and management fees.","source":"front_matter","rule":"R3","candidates":{"front_matter":"no","evidence_card":"no","evidence_card_quoted":true},"card_detail":null,"card_quote":"Importantly, in order to focus on specific behaviours of our methods, in the following Results Section we have ignored transaction costs and management fees.","meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Modern Hopfield networks achieve performance on par or better than state‑of‑the‑art LSTM and Transformer models, with faster training times and better stability.","kind":"performance","quote":"Compared to state-of-the-art deep-learning methods such as Long-Short Term Memory networks and Transformers, we find that the proposed approach performs on par or better, while providing faster training times and better stability."},{"finding":"Hopfield Encoder outperforms both LSTM and Hopfield Pooling on the COMMODITIES dataset according to Sharpe‑ratio statistical tests.","kind":"statistical","quote":"the Hopfield Encoder performs better than both LSTM and Hopfield Pooling in the COMMODITIES dataset."},{"finding":"On the ETF and STOCKS datasets, deep‑learning methods (LSTM, HOP‑POOL, HOP‑TRA) achieve statistically significantly higher Sharpe ratios than traditional MVO and HRP methods.","kind":"performance","quote":"The superiority of the end-to-end deep learning approaches (LSTM, HOP-POOL, HOP-TRA) are also clear in the STOCKS dataset, again with statistically significant higher Sharpe ratios between the deep learning methods against the traditional ones."},{"finding":"For the COMMODITIES and FAMAFRENCH datasets, all deep‑learning methods underperform compared to traditional approaches, likely due to poor generalization across multiple regimes.","kind":"negative","quote":"Here all the deep learning methods fail in comparison to traditional approaches. We believe that this result stems from a poor generalization effect in return series where multiple regimes are present."},{"finding":"A simple MHN model often surpasses more intricate architectures such as LSTMs, achieving superior performance with significantly shorter training times.","kind":"performance","quote":"Our results reveal that a simple MHN often surpasses more intricate architectures such as LSTMs, achieving superior performance within significantly shorter training times."}],"reported_performance":{"strategy":"Hopfield Pooling (HOP-POOL)","sharpe":1.252,"sharpe_benchmark":0.827,"annual_return_pct":13,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"cross-validation","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":41,"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":"2026-11-14T18:05:44Z"},"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/"}