{"paper_id":"2305.17881","title":"Integrating Different Informations for Portfolio Selection","authors":["Yi Huang","Wei Zhu","Duan Li","Shushang Zhu","Shikun Wang"],"paper_date":"2023-05-29","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2305.17881v1","page_url":"https://thequant.space/flowcharts/integrating-different-informations-for-portfolio-selection/","scores":{"empirical_rigor":7,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a novel combined approach that integrates backward‑looking historical data and forward‑looking market portfolio information via a Gaussian mixture model to forecast asset returns.","quote":"In this paper, we propose a novel combined approach to estimate the future returns of assets and test the out-of-sample performance of investment decisions.","summary":"Following the idea of Bayesian learning via Gaussian mixture model, we organically combine the backward-looking information contained in the historical data and the forward-looking information implied by the market portfolio, which is affected by heterogeneous expectations and noisy trading…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Investors are categorized into a rational less‑informed investor, a rational informed investor, and a noise trader.","type":"market","stated":"explicit","quote":"... namely a representative rational less‑informed investor, a representative rational informed investor, and a representative noise trader."},{"assumption":"Gaussian mixture components are predetermined, reducing the number of parameters to estimate.","type":"model","stated":"explicit","quote":"To reduce the number of parameters to be estimated, we reasonably suppose the Gaussian components pk (r|µk, Σk) (k = 1, · · · , m) are predetermined."},{"assumption":"Covariance matrix can be exactly estimated by historical data (ΣU = ΣI = Σ).","type":"statistical","stated":"explicit","quote":"assume that the covariance matrix can be exactly estimated by historical data, i.e., ΣU = ΣI = Σ."},{"assumption":"Informed and less‑informed investors share the same risk‑aversion coefficient (δU = δI = δ).","type":"model","stated":"explicit","quote":"We also assume the homogeneous risk aversion for the informed and the less‑informed investor, i.e., δU = δI = δ."},{"assumption":"Risk‑aversion of the informed investor varies over time with a volatility‑index based factor.","type":"model","stated":"explicit","quote":"we introduce a volatility index‑based time‑varying risk aversion coefficient for the informed investor,"},{"assumption":"Markets with higher informed‑investor share have lower noise‑trading intensity.","type":"market","stated":"explicit","quote":"we assume that markets with a higher share of the informed investor exhibit lower noise trading intensity."}]},"datasets":{"status":"available","items":[{"name":"35 constituent markets in the MSCI ACWI Index","provider":"Bloomberg","kind":"market-prices","access":"commercial","period":"more than ten years (weekly returns)","universe":"major stock markets worldwide","quote":"each of which has a complete time series of returns covering more than ten years. Weekly returns, volatility indices, and market capitalization data of primary sector indices are obtained from the Bloomberg database.","known_id":"bloomberg"}],"known":[{"id":"bloomberg","name":"Bloomberg","provider":"Bloomberg L.P.","kind":"market-prices","access":"commercial","quote":"Weekly returns, volatility indices, and market capitalization data of primary sector indices are obtained from the Bloomberg database.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"transaction cost ratio set to c = 0.5% per trade","quote":"We set the transaction cost ratio as c = 0.5% (DeMiguel et al., 2009).","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"transaction cost ratio set to c = 0.5% per trade","card_quote":"We set the transaction cost ratio as c = 0.5% (DeMiguel et al., 2009).","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The combined estimation usually has a lower RMSE than the backward‑looking estimation and adapts to market efficiency.","kind":"predictive-accuracy","quote":"In addition, the combined estimation usually has a lower RMSE than the backward‑looking one."},{"finding":"In partially informative markets, the combined approach provides the highest prediction accuracy, outperforming both backward‑ and forward‑looking methods.","kind":"performance","quote":"for partially informative markets, the combined approach provides the highest accuracy in prediction, highlighting its robustness and self‑adaptivity in integrating different sources of information."},{"finding":"Out‑of‑sample, the combined strategy achieves higher Sharpe ratios and lower turnover and drawdown in many markets compared with the other two strategies.","kind":"performance","quote":"Table 4 reports the average (across the rebalancing window) annualized SR, TRN, and MDD of the three types of strategies with moderate risk‑aversion setting in the 35 markets."},{"finding":"The combined estimation responds more quickly at market turning points than the backward‑looking approach, especially when forward‑looking information is efficient.","kind":"predictive-accuracy","quote":"As expected, the forward‑looking and combined approaches can respond more quickly in markets with efficient forward‑looking information."},{"finding":"The method is applicable across markets with different degrees of efficiency, performing well in both emerging and developed markets.","kind":"performance","quote":"While the backward‑looking (forward‑looking) strategy has better performance in emerging (developed) markets, we show that the combined strategy can be well applied to a wider range of stock markets, not limited to emerging or developed markets."}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":25,"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":"2026-11-14T18:26:22Z"},"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/"}