{"paper_id":"2310.09578","title":"Sparse Index Tracking via Topological Learning","authors":["Anubha Goel","Puneet Pasricha","Juho Kanniainen"],"paper_date":"2023-10-14","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2310.09578v1","page_url":"https://thequant.space/flowcharts/sparse-index-tracking-via-topological-learning/","scores":{"empirical_rigor":7,"math_complexity":8.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a novel sparse index tracking methodology that leverages topological data analysis to learn regularization parameters and accommodates both ℓ1 and ℓ2 penalties.","quote":"In this research, we introduce a novel methodology for the index tracking problem with sparse portfolios by leveraging topological data analysis (TDA).","summary":"In this research, we introduce a novel methodology for the index tracking problem with sparse portfolios by leveraging topological data analysis (TDA). Utilizing persistence homology to measure the riskiness of assets, we introduce a topological method for data-driven learning of the parameters for…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The investor holds the portfolio weights fixed until the end of a specific investment horizon.","type":"execution","stated":"explicit","quote":"We assume that the investor holds the portfolio weights fixed until the end of a specific investment horizon."},{"assumption":"TDA norms can effectively track changes in stock return dynamics and are used as risk measures.","type":"model","stated":"explicit","quote":"Overall, the TDA norms can effectively track changes in the state of stock returns dynamics, thus motivating the use of Lp-norms as a measure of risk associated with the stock returns."},{"assumption":"In‑sample period length is 504 trading days and out‑of‑sample period length is 21 trading days, using a rolling‑window scheme.","type":"data","stated":"explicit","quote":"In this paper, the length of the in-sample periods d1 = 504 trading days (25 months) and the length of the out-of-sample periods d2 = 21 trading days (one month)."}]},"datasets":{"status":"available","items":[{"name":"S&P500 daily closing prices and constituents","provider":"Thomson Reuters Datastream","kind":"market-prices","access":"commercial","period":"October 11, 1999 to March 8, 2023","universe":"S&P500 constituents","quote":"To investigate the performance of our proposed index tracking model, we focus on the daily closing prices of the S&P500 and its constituents over a period spanning nearly 23 years from October 11, 1999, to March 8, 2023. The data is collected from Thomson Reuters Datastream.","known_id":"refinitiv"}],"known":[{"id":"refinitiv","name":"Refinitiv / LSEG (Datastream, Eikon, Tick History)","provider":"LSEG","kind":"market-prices","access":"commercial","quote":"The data is collected from Thomson Reuters Datastream.","mentions":2,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"partial","headline_net_of_costs":null,"detail":"Turnover (average absolute value of trades) is used as a proxy for transaction costs; TDAEN12 turnover is less than half of EN.","quote":"Fourthly, the results show that the turnover (defined as the average absolute value of trades) with TDAEN12 is less than half of that with EN, making it the method with the lowest turnover among all examined.","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"partial","evidence_card_quoted":true},"card_detail":"Turnover (average absolute value of trades) is used as a proxy for transaction costs; TDAEN12 turnover is less than half of EN.","card_quote":"Fourthly, the results show that the turnover (defined as the average absolute value of trades) with TDAEN12 is less than half of that with EN, making it the method with the lowest turnover among all examined.","meaning":"costs appear only as a robustness check, a flat fee without spread or impact, or outside the headline numbers"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"TDAEN12 consistently yields the lowest risk measures across volatility, downside deviation, VaR, and CVaR compared to other ℓ1/ℓ2 models.","kind":"performance","quote":"TDAEN12 consistently yields the lowest risk measures when compared to other models using both ℓ1 and ℓ2 penalties"},{"finding":"TDAEN12 achieves substantially lower turnover—less than half of EN—indicating reduced transaction costs.","kind":"performance","quote":"the turnover (defined as the average absolute value of trades) with TDAEN12 is less than half of that with EN"},{"finding":"The proposed TDA‑based approach often outperforms conventional methods in tracking error during turbulent market periods.","kind":"performance","quote":"Our empirical findings demonstrate that the tracking error of our approach is not merely comparable, but often excels in turbulent market conditions relative to the prevailing methods."},{"finding":"Risk‑adjusted performance is enhanced and overall risk is reduced when using the TDA‑based method.","kind":"performance","quote":"the risk associated with the tracking portfolio is reduced, risk-adjusted performance is enhanced"}],"reported_performance":null,"out_of_sample":{"method":"walk-forward","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":39,"citations":{"cited_by_in_corpus":1,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":1},"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":"2026-11-14T18:20:47Z"},"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/"}