{"paper_id":"2507.03963","title":"Quantum Stochastic Walks for Portfolio Optimization: Theory and Implementation on Financial Networks","authors":["Yen Jui Chang","Wei-Ting Wang","Yun-Yuan Wang","Chen-Yu Liu","Kuan-Cheng Chen","Ching-Ray Chang"],"paper_date":"2025-07-05","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2507.03963v2","page_url":"https://thequant.space/flowcharts/quantum-stochastic-walks-for-portfolio-optimization-theory/","scores":{"empirical_rigor":8,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a quantum stochastic walk (QSW) optimizer that embeds assets in a weighted graph and derives portfolio weights from the walk's stationary distribution.","quote":"We propose a quantum stochastic walk (QSW) optimizer that embeds assets in a weighted graph: nodes represent securities while edges encode the return-covariance kernel. Portfolio weights are derived from the walk's stationary distribution.","summary":"Financial markets are noisy yet contain a latent graph-theoretic structure that can be exploited for superior risk-adjusted returns. We propose a quantum stochastic walk (QSW) optimizer that embeds assets in a weighted graph: nodes represent securities while edges encode the return-covariance…","strategy":"Quantum Stochastic Walk (QSW) optimizer"},"assumptions":{"status":"available","items":[{"assumption":"The QSW dynamics converge to a unique stationary state for any ω>0.","type":"model","stated":"explicit","quote":"if we set cij = Gij where G is the Google matrix (2), then for any ω > 0 the quantum dynamical semigroup generated by Eq. (4) is primitive: a unique, full-rank stationary state ρ∞exists and ρ(t) →ρ∞exponentially fast."},{"assumption":"Daily returns are computed over a training window of length T.","type":"data","stated":"explicit","quote":"We encode financial data on the graph by taking n financial assets with daily returns {ri,t} over a training window of length T."},{"assumption":"Portfolios are rebalanced quarterly.","type":"execution","stated":"explicit","quote":"All strategies are evaluated in a common backtesting framework with quarterly rebalancing."},{"assumption":"Turnover above roughly 200% per year is considered prohibitively expensive.","type":"execution","stated":"explicit","quote":"annual turnover above roughly 200% is often considered prohibitively expensive because of spread and market-impact costs, whereas turnover below about 50% is usually viewed as readily implementable."},{"assumption":"The universe consists of 100 top‑S&P 500 constituents selected point‑in‑time to avoid survivorship bias.","type":"data","stated":"explicit","quote":"we draw 100 constituents from the S&P 500 members active at that time; at each quarterly rebal‑ance we remove any delisted names"},{"assumption":"Self‑loops encode asset‑specific Sharpe information and edges penalize correlation.","type":"model","stated":"explicit","quote":"we map expected returns and covariances to a weighted, directed graph in which self-loops encode asset-specific Sharpe information and inter-node edges penalize correlation"}]},"datasets":{"status":"available","items":[{"name":"S&P 500 daily adjusted close prices","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":"1988-01-01 to 2024-12-31","universe":"S&P 500 constituents","quote":"Historical daily adjusted close prices from 1988-01-01 to 2024-12-31 are obtained from Yahoo Finance.","known_id":"yahoo-finance"}],"known":[{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"His-torical daily adjusted close prices from 1988-01-01 to 2024-12-31 are obtained from Yahoo Finance.","mentions":2,"used_in":2}],"sample":{"start":"1990","end":"2024","market":"S&P 500 stocks","asset_class":"equities","frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":null,"detail":null,"quote":"High-turnover strategies face substantial transaction costs, price impact during execution, and capacity constraints that can erode theoretical performance gains in practice","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"no","evidence_card_quoted":true},"card_detail":null,"card_quote":"High-turnover strategies face substantial transaction costs, price impact during execution, and capacity constraints that can erode theoretical performance gains in practice","meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"QSW lifts out‑of‑sample Sharpe ratio by up to 27% while cutting annual turnover from 480% (mean‑variance) to 2‑90%.","kind":"performance","quote":"six scenario portfolios calibrated on 1- and 2-year windows lift the out-of-sample Sharpe ratio by up to 27% while cutting annual turnover from 480% (mean-variance) to 2-90%."},{"finding":"A grid search identifies a robust sweet spot (α,λ≲0.5, ω∈[0.2,0.4]) delivering Sharpe ≈0.97 at ≤5% turnover and HHI ≈0.01.","kind":"performance","quote":"A $5^{4}=625$-point grid search identifies a robust sweet spot, $α,λ\\lesssim0.5$ and $ω\\in[0.2,0.4]$, that delivers Sharpe $\\approx0.97$ at $\\le 5\\%$ turnover and Herfindahl-Hirschman index $\\sim0.01$."},{"finding":"Across 31,350 back‑tests, the best‑per‑draw QSW beats re‑optimised mean‑variance on Sharpe in 54% of cases and always wins on trading efficiency (median turnover 36% vs 351%).","kind":"performance","quote":"the best-per-draw QSW beats re-optimised mean-variance on Sharpe in 54% of cases and always wins on trading efficiency, with median turnover 36% versus 351%."},{"finding":"Overall QSW raises annualized Sharpe by 15% and cuts turnover by 90% relative to classical optimisation, while respecting UCITS 5/10/40 rule.","kind":"performance","quote":"Overall, QSW raises the annualized Sharpe ratio by 15% and cuts turnover by 90% relative to classical optimisation, all while respecting the UCITS 5/10/40 rule."},{"finding":"In a 34‑year, 30‑universe dynamic experiment QSW achieves ~23% CAGR, Sharpe near 1.0, and turnover around 200% (vs 300‑400% for MPT).","kind":"performance","quote":"... The dynamic optimizer accepts a higher average turnover (around 200% per year) than in the static Phase 1 experiments"}],"reported_performance":{"strategy":"Quantum Stochastic Walk (QSW) optimizer","sharpe":0.9846,"sharpe_benchmark":0.86,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":19.27},"out_of_sample":{"method":"walk-forward","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":29,"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":true,"code_url":"https://github.com/aceest/quantum-stochastic-walk-for-Portfolio-Optimization-Theory-and-Implementation-on-Financial-Networks","note":null},"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":"2026-11-14T17:34:02Z"},"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/"}