{"paper_id":"2504.12266","title":"Semiparametric Dynamic Copula Models for Portfolio Optimization","authors":["Savita Pareek","Sujit K. Ghosh"],"paper_date":"2025-04-16","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2504.12266v1","page_url":"https://thequant.space/flowcharts/semiparametric-dynamic-copula-models-for-portfolio-optimizat/","scores":{"empirical_rigor":6,"math_complexity":8,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a semiparametric dynamic copula model that combines non‑parametric copula estimation with parametric marginal distributions to allow all parameters to evolve over time for portfolio optimization.","quote":"In this study, a semiparametric model is proposed that combines non‑parametrically estimated copulas with parametrically estimated marginals to allow all parameters to dynamically evolve over time.","summary":"The mean-variance portfolio model, based on the risk-return trade-off for optimal asset allocation, remains foundational in portfolio optimization. However, its reliance on restrictive assumptions about asset return distributions limits its applicability to real-world data. Parametric copula…","strategy":"copula-based portfolio optimization (copula_cov_3constraint)"},"assumptions":{"status":"available","items":[{"assumption":"Portfolio weights are constrained to be long‑only, sum to one, and meet an equal‑weighted‑portfolio (EWO) return constraint.","type":"execution","stated":"explicit","quote":"The optimization is subject to long‑only, full investment, and an additional EWO return constraint requiring the portfolio’s expected return to exceed the average return of the individual assets."},{"assumption":"A rolling window of approximately one year of daily returns is used to dynamically adjust portfolio weights.","type":"model","stated":"explicit","quote":"The third component incorporates a rolling window framework (typically 250 trading days, adjusted for the specific market’s calendar) with daily or weekly rebalancing to account for changing market dynamics."},{"assumption":"Asset marginal distributions follow the Skewed Generalized t (SGT) family.","type":"statistical","stated":"explicit","quote":"Marginal distributions were modeled using the Skewed Generalized T family."},{"assumption":"The empirical beta copula (EBC) is used for dependence modeling due to its computational efficiency.","type":"model","stated":"explicit","quote":"In this work, we adopt the EBC due to its greater computational efficiency compared to ECBC."},{"assumption":"Daily equity prices are adjusted for corporate actions (dividends, splits) before computing log returns.","type":"data","stated":"explicit","quote":"... and compute log returns as diff(log(Ad(get(ticker)))), ensuring consistency by accounting for corporate actions such as dividends and stock splits."},{"assumption":"The analysis period includes both stable periods and extreme market stress (e.g., COVID‑19 crisis).","type":"market","stated":"explicit","quote":"The empirical analysis spans April 1 2018 to March 31st 2025, capturing both stable periods and episodes of extreme market stress, including the COVID‑19 crisis."}]},"datasets":{"status":"available","items":[{"name":"US, India, Hong Kong equity price data","provider":"Yahoo Finance via quantmod package","kind":"market-prices","access":"public","period":"2018-04-01 to 2025-03-31","universe":"20 large‑cap stocks per market (S&P 500, NIFTY 50, top‑traded HKEX stocks)","quote":"We evaluate the proposed portfolio optimization framework using stock price data from three major markets: the United States (developed), India (developing), and Hong Kong (representing the Far East Asian financial hub). The empirical analysis spans April 1 2018 to March 31st 2025","known_id":"yahoo-finance"}],"known":[{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"Daily equity prices are retrieved from Yahoo Finance via the quantmod package in 6 Dynamic Copula Models for Portfolio Optimization PAREEK, S., GHOSH S.K. −0.004 0.000 0.002 (a) NFLX vs COST Mean Return Mean Return Jan−20 Jan−22 Jan−24 NFLX COST 0.01 0.02 0.03 0.04 (b) Volatility Standard Deviation Jan−20 Jan−22 Jan−24 −0.10 0.00 0.10 0.20 (c) Sharpe Ratio Sharpe Ratio Jan−20 Jan−22 Jan−24 0.2 0.3 0.4 0.5 (d) Rolling","mentions":1,"used_in":1}],"sample":{"start":"2018-04","end":"2025-03","market":"US, India, Hong Kong equities (20 large‑cap stocks per marke","asset_class":"equities","frequency":"weekly"},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":false,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"no","evidence_card":"unclear","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The copula‑based portfolio outperforms the sample‑covariance portfolio, achieving a 28.9% higher mean return and a 5.1% higher Sharpe ratio in the Hong Kong market during the COVID‑19 period.","kind":"performance","quote":"Hong Kong market exhibits the largest relative gains, with improvements of 28.9% in mean return and 5.1% in Sharpe ratio, highlighting the method’s advantage in capturing asymmetric and tail‑dependent relationships."},{"finding":"Across all three markets, the copula approach achieved higher average Sharpe ratios and returns in a substantial proportion of rolling windows (e.g., 54.5% of US windows for return, 59.6% of India windows for Sharpe).","kind":"performance","quote":"Across all three markets, the copula approach achieved higher average Sharpe ratios and returns in a substantial proportion of rolling windows, for instance, outperforming in 54.5% (US), 69.7% (India), and 30.3% (Hong Kong) of windows in"},{"finding":"Weekly rebalancing with the copula‑based method yields higher gains than daily rebalancing, especially in volatile periods.","kind":"performance","quote":"The copula‑based method demonstrates superior performance in most scenarios, particularly under weekly rebalancing."}],"reported_performance":{"strategy":"copula-based portfolio optimization (copula_cov_3constraint)","sharpe":11.1307,"sharpe_benchmark":11.1923,"annual_return_pct":0.1204,"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":"2508.14999","title":"Investment Portfolio Optimization Based on Modern Portfolio Theory and Deep Learning Models","paper_date":"2025-08-20","authors":["Maciej Wysocki","Paweł Sakowski"],"empirical_rigor":8,"page_url":"https://thequant.space/flowcharts/investment-portfolio-optimization-based-on-modern-portfolio/"},"question":"Which rebalancing frequency leads to better portfolio performance?","this_paper_finding":"Weekly rebalancing with the copula‑based method yields higher gains than daily rebalancing, especially in volatile periods.","other_paper_finding":"Strategies with less frequent rebalancing (e.g., 30‑, 60‑, 90‑day) outperform those with very short rebalancing periods.","other_paper_quote":"Strategies with less frequent rebalancing typically perform better than these with the shortest rebalancing windows across all considered methods","note":"Both findings show that longer intervals (weekly or 30‑, 60‑, 90‑day) outperform daily or very short rebalancing.","confidence":0.735},{"paper":{"paper_id":"2409.10859","title":"Macroscopic properties of equity markets: stylized facts and portfolio performance","paper_date":"2024-09-17","authors":["Steven Campbell","Qien Song","Ting-Kam Leonard Wong"],"empirical_rigor":8,"page_url":"https://thequant.space/flowcharts/macroscopic-properties-of-equity-markets-stylized-facts-and/"},"question":"Does increasing rebalancing frequency improve portfolio returns after transaction costs?","this_paper_finding":"Weekly rebalancing with the copula‑based method yields higher gains than daily rebalancing, especially in volatile periods.","other_paper_finding":"Frequent rebalancing (1–2 day intervals) underperforms in most years once transaction costs of 0.25% are included.","other_paper_quote":"frequent rebalancing (say, every 1–2 days) underperforms in most years under 0.25% transaction cost","note":"A shows 1–2 day rebalancing underperforms with 0.25% costs; B finds weekly rebalancing outperforms daily, confirming less frequent trades yield higher gains.","confidence":0.598}],"contradicts":[],"qualifies":[],"compared_with":16,"citations":{"cited_by_in_corpus":1,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":1,"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":"2026-11-14T17:40:40Z"},"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/"}