{"paper_id":"2402.05364","title":"Coarse graining correlation matrices according to macrostructures: Financial markets as a paradigm","authors":["M. Mijaíl Martínez-Ramos","Parisa Majari","Andres R. Cruz-Hernández","Hirdesh K. Pharasi","Manan Vyas"],"paper_date":"2024-02-08","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2402.05364v2","page_url":"https://thequant.space/flowcharts/coarse-graining-correlation-matrices-according-to-macrostruc/","scores":{"empirical_rigor":4,"math_complexity":6.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes to coarse‑grain Pearson correlation matrices into low‑dimensional sectorial Guhr matrices and to use these matrices for clustering market states and analyzing their evolution.","quote":"This matrix, which we shall call the ‘sectorial Guhr matrix’, provides the appropriate tool for our endeavor. We will thus repeat the analysis of Ref [8] in some detail and compare the results of noise reduction by a power map (hereinafter “PM”) [27–29], and KM [30, 31] with the cor-responding results for the sectorial Guhr matrices.","summary":"We analyze correlation structures in financial markets by coarse graining the Pearson correlation matrices according to market sectors to obtain Guhr matrices using Guhr’s correlation method according to Ref. [“P. Rinn {\"\\it et. al.”}, Europhysics Letters 110, 68003 (2015)\"]. We compare the results…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Closing prices are used to compute returns, assuming they reflect investor decisions better than intra‑day prices.","type":"data","stated":"explicit","quote":"Note that the closing prices data are used as some random walk properties are arXiv:2402.05364v2 [q-fin.ST] 25 Jun 2024 2 expected for intra-day trading and opening data seem to be more volatile than closing data; also they don’t reflect the same degree of the decisions of investors/investment groups that are based on fundamental data.","support":"full"},{"assumption":"Correlation matrices are built over epochs of 20 trading days with a one‑day shift.","type":"model","stated":"explicit","quote":"We shall construct the correlation matrices as a function of the trading days with epoch lengths of 20 and one day shift of the epoch for the daily closing price returns.","support":"full"},{"assumption":"Stocks that have not been traded for more than two consecutive days are excluded from the analysis.","type":"data","stated":"explicit","quote":"Note that N is usu-ally smaller than the number of listed stocks as we use the criterion to exclude stocks that have not been traded for more than two consecutive days during the chosen time horizon.","support":"full"},{"assumption":"The sectorial Guhr matrix is obtained by averaging intra‑ and inter‑sector correlation elements, assuming this coarse‑graining preserves the relevant market information.","type":"model","stated":"explicit","quote":"We then obtain the Ns × Ns Guhr matrices by coarse graining Pearson correlation ma-trices taking the average over different intra- and inter-sectorial correlation matrix elements.","support":"full"},{"assumption":"The power‑map noise‑suppression parameter ϵ is chosen in (0,1) and applied to Pearson matrices before coarse‑graining.","type":"statistical","stated":"explicit","quote":"We apply the PM method as a noise suppression tech-nique [27–29] on Pearson correlation matrices using the following definition: C′ ij(τ) = sign(Cij)|Cij|1+ϵ (2) where 0 < ϵ < 1 is the noise-suppression parameter.","support":"full"},{"assumption":"KM clustering on Pearson and Guhr matrices yields meaningful market states.","type":"model","stated":"explicit","quote":"Using these Pearson and Guhr matrices, we use KM clustering to obtain the market states.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"S&P 500 stocks","provider":null,"kind":"market-prices","access":"unknown","period":"3523 trading days","universe":"350 stocks after purification","quote":"The S&P 500 market consists of N = 350 number of stocks over T = 3523 trad-ing days and the Nikkei 225 consists of N = 150 stocks over T = 3458 trading days.","support":"full"},{"name":"Nikkei 225 stocks","provider":null,"kind":"market-prices","access":"unknown","period":"3458 trading days","universe":"150 stocks after purification","quote":"The S&P 500 market consists of N = 350 number of stocks over T = 3523 trad-ing days and the Nikkei 225 consists of N = 150 stocks over T = 3458 trading days.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"n/a","evidence_card":"no","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"unclear","findings":[{"finding":"The behavior of market states obtained from sectorial Guhr matrices is rather similar to that obtained from Pearson correlation matrices.","kind":"statistical","quote":"We find that the behavior of the market states for the CG matrices is rather similar to the Pearson correlation matrices.","support":"full"},{"finding":"The number of relevant variables is reduced from tens of thousands to 55 for S&P 500 and 21 for Nikkei 225 by using sectorial Guhr matrices.","kind":"statistical","quote":"Note that the number of relevant variables is re-duced to 55 and 21 respectively for S&P 500 and Nikkei 225 markets thus the making the problem more manage-able.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":10,"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":2,"extracted_at":"2026-11-14T19:58:59Z"},"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/"}