{"paper_id":"2601.07792","title":"Non-Convex Portfolio Optimization via Energy-Based Models: A Comparative Analysis Using the Thermodynamic HypergRaphical Model Library (THRML) for Index Tracking","authors":["Javier Mancilla","Theodoros D. Bouloumis","Frederic Goguikian"],"paper_date":"2026-01-12","paper_type":"software-tool","paper_url":"https://arxiv.org/pdf/2601.07792v1","page_url":"https://thequant.space/flowcharts/non-convex-portfolio-optimization-via-energy-based-models-a/","scores":{"empirical_rigor":7,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a novel THRML-based approach for cardinality-constrained index tracking that samples portfolios from a Boltzmann distribution using GPU-accelerated block Gibbs sampling and introduces three key innovations.","quote":"We implement three key innovations: (1) dynamic coupling strength that scales inversely with market volatility (VIX), adapting diversification pressure to market regimes; (2) rebalanced bias weights prioritizing tracking quality over momentum for index replication; and (3) sector-aware post-processing ensuring institut","summary":"Portfolio optimization under cardinality constraints transforms the classical Markowitz mean-variance problem from a convex quadratic problem into an NP-hard combinatorial optimization problem. This paper introduces a novel approach using THRML (Thermodynamic HypergRaphical Model Library), a…","strategy":"THRML (Thermodynamic HypergRaphical Model Library)"},"assumptions":{"status":"available","items":[{"assumption":"Cardinality constraint limits the number of holdings to a fixed K (30 stocks).","type":"model","stated":"explicit","quote":"Target cardinality (K) 30 stocks"},{"assumption":"All selected assets are equally weighted after selection.","type":"execution","stated":"explicit","quote":"After selection, we apply equal weights to selected assets."},{"assumption":"Transaction costs are a flat 10 basis points per trade.","type":"execution","stated":"explicit","quote":"Transaction cost 10 bps per trade"},{"assumption":"Training and test periods are non‑overlapping with a walk‑forward methodology.","type":"data","stated":"explicit","quote":"Training Period: January 2010 – December 2022 (13 years) • Test Period: January 2023 – December 2025 (3 years)"},{"assumption":"Liquidity is measured by normalized average trading volume.","type":"model","stated":"explicit","quote":"Liquidityi = AvgVolumei / maxj AvgVolumej"},{"assumption":"Dynamic coupling strength depends on VIX levels.","type":"model","stated":"explicit","quote":"dynamic coupling strength that scales inversely with market volatility (VIX)"}]},"datasets":{"status":"available","items":[{"name":"Historical price data","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":"January 2010 to December 2025","universe":"100 large‑cap U.S. equities (S&P 500 constituents)","quote":"Historical price data from January 2010 to December 2025 was obtained via Yahoo Finance. Prices were adjusted for splits and dividends, with missing values forward-filled.","known_id":"yahoo-finance"},{"name":"VIX data","provider":null,"kind":"macro","access":"unknown","period":"January 2010 to December 2025","universe":null,"quote":"VIX data for dynamic coupling was obtained for the same period."}],"known":[{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"TABLE II ASSET UNIVERSE COMPOSITION BY SECTOR Sector Number of Stocks Technology 23 Healthcare 15 Financials 16 Consumer Discretionary 10 Consumer Staples 8 Energy 6 Industrials 10 Communication Services 6 Utilities 4 Real Estate 2 Total 100 Historical price data from January 2010 to December 2025 was obtained via Yahoo Finance.","mentions":1,"used_in":1}],"sample":{"start":"2023","end":"2025","market":"S&P 500 stocks","asset_class":"equities","frequency":"monthly"},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"Flat 10 basis points per trade applied in backtest results.","quote":"Transaction cost 10 bps per trade","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"Flat 10 basis points per trade applied in backtest results.","card_quote":"Transaction cost 10 bps per trade","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"THRML achieves a tracking error of 4.31%, 24% lower than the next‑best baseline (NSGA‑II at 5.66%).","kind":"performance","quote":"Tracking Error: THRML’s 4.31% TE is 24% lower than the next-best baseline (NSGA‑II at 5.66%)"},{"finding":"THRML generates a total return of 128.63% versus the S&P 500 index return of 79.61%, yielding an information ratio of 1.85.","kind":"performance","quote":"THRML generates 49 percentage points of excess return (128.63% vs 79.61% index), yielding an Information Ratio of 1.85"},{"finding":"The Sharpe ratio of THRML is 1.87, exceeding all baselines by more than 34%.","kind":"performance","quote":"The Sharpe ratio of 1.87 exceeds all baselines by 34%+ margin."},{"finding":"Diebold‑Mariano tests show statistical significance at p < 0.0001 for all baseline comparisons.","kind":"statistical","quote":"All comparisons are statistically significant at the p < 0.0001 level, rejecting the null hypothesis that THRML and baselines have equal tracking accuracy."},{"finding":"Maximum drawdown for THRML is -17.09%, the shallowest among all methods.","kind":"performance","quote":"THRML achieves the shallowest maximum drawdown (-17.09%) while delivering the highest returns"}],"reported_performance":{"strategy":"THRML (Thermodynamic HypergRaphical Model Library)","sharpe":1.87,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":128.63,"max_drawdown_pct":17.09},"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":83,"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":"2026-11-14T17:16:30Z"},"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/"}