{"paper_id":"2510.16066","title":"Cash Flow Underwriting with Bank Transaction Data: Advancing MSME Financial Inclusion in Malaysia","authors":["Chun Chet Ng","Wei Zeng Low","Jia Yu Lim","Yin Yin Boon"],"paper_date":"2025-10-17","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2510.16066v3","page_url":"https://thequant.space/flowcharts/cash-flow-underwriting-with-bank-transaction-data-advancing/","scores":{"empirical_rigor":6.5,"math_complexity":3.5,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes a cash‑flow‑based underwriting pipeline and introduces a novel dataset of Malaysian MSME loan applicants.","quote":"First, we propose a cash flow-based underwriting pipeline where we utilize bank statement data for end-to-end data extraction and machine learning credit scoring.","summary":"Despite accounting for 96.1% of all businesses in Malaysia, access to financing remains one of the most persistent challenges faced by Micro, Small, and Medium Enterprises (MSMEs). Newly established businesses are often excluded from formal credit markets as traditional underwriting approaches rely…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The training/validation split (60%/40%) using stratified random sampling preserves the class distribution.","type":"data","stated":"implied","quote":"The dataset is split into training (60%) and validation (40%) sets using stratified random sampling to preserve class distribution."},{"assumption":"Class imbalance (518 non‑default vs 93 default) reflects the natural distribution of real‑world lending portfolios.","type":"market","stated":"implied","quote":"the class imbalance in our dataset (518 non-default vs. 93 default cases) reflects the natural distribution observed in real-world lending portfolios"},{"assumption":"AUROC is a robust evaluation metric for imbalanced credit‑scoring data.","type":"statistical","stated":"implied","quote":"we addressed this limitation through appropriate model selection, WOE-based feature engineering, and careful evaluation using AUROC, which is robust to class imbalance."}]},"datasets":{"status":"available","items":[{"name":"Malaysian MSME bank‑statement transaction dataset","provider":"Malaysian consulting firm","kind":"alternative","access":"proprietary","period":null,"universe":"MSME loan applicants","quote":"Second, we introduce a novel dataset of 611 loan applicants from a Malaysian consulting firm."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"The best model achieves an AUROC of 0.806 on the validation set, a 24.6% improvement over models using only application information.","kind":"predictive-accuracy","quote":"our best model achieving an AUROC of 0.806 on validation set, representing a 24.6% improvement over models using application information only."},{"finding":"Logistic regression consistently outperforms tree‑based ensembles across all feature configurations and data splits.","kind":"performance","quote":"LR consistently outperforms tree-based ensembles."},{"finding":"Models that incorporate bank‑transaction features consistently outperform those that rely solely on application information.","kind":"predictive-accuracy","quote":"models incorporating bank transaction data consistently outperform those relying on application information alone."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":21,"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":"2026-11-14T17:24:17Z"},"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/"}