{"paper_id":"2511.08306","title":"An extreme Gradient Boosting (XGBoost) Trees approach to Detect and Identify Unlawful Insider Trading (UIT) Transactions","authors":["Krishna Neupane","Igor Griva"],"paper_date":"2025-11-11","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2511.08306v1","page_url":"https://thequant.space/flowcharts/an-extreme-gradient-boosting-xgboost-trees-approach-to-det/","scores":{"empirical_rigor":4,"math_complexity":2.5,"quadrant":"Philosophers"},"proposes":{"status":"available","statement":"The paper proposes applying XGBoost to detect unlawful insider trading transactions with high accuracy.","quote":"We, hence, resort to and apply XGBoost to alleviate challenges of identification and detection of unlawful activities.","summary":"Corporate insiders have control of material non-public preferential information (MNPI). Occasionally, the insiders strategically bypass legal and regulatory safeguards to exploit MNPI in their execution of securities trading. Due to a large volume of transactions a detection of unlawful insider…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The dataset is balanced with a 0.5:0.5 ratio of lawful to unlawful transactions.","type":"data","stated":"explicit","quote":"Each dataset subset, balanced (0.5:0.5 ratio) and sub-divided by feature sets (original vs. PCA-integrated), was then split deterministically 80 percent (training): 20 percent (test) for analysis."},{"assumption":"Features are normalized using z‑score and categorical variables are one‑hot encoded.","type":"model","stated":"explicit","quote":"Numerical features Xi (i = 1,...,n) were normalized using the z-score method1, while categorical features were one-hot-encoded."},{"assumption":"Hyperparameters are tuned via random search over 5 iterations within 5‑fold cross‑validation.","type":"model","stated":"explicit","quote":"Hyperparameters such as η, γ, max depth, and sample rate were initialized in a random search space, with tuning conducted over 5 iterations within 5-fold cross-validation."}]},"datasets":{"status":"available","items":[{"name":"SEC Form 4 filings linked with CRSP and Compustat‑CapitalIQ","provider":"SEC / CRSP / Compustat","kind":"fundamentals","access":"public","period":null,"universe":"3984 fully labeled transactions (1992 unlawful)","quote":"Data originates from SEC Form 4 filings, linked with CRSP and Compustat-CapitalIQ trade and finance data via personid, cik, and companyid. Comprising 3984 fully labeled transactions (1992 unlawful) with 110 dimensions per row, the merged dataset was used...","known_id":"crsp"}],"known":[{"id":"compustat","name":"Compustat","provider":"S&P Global","kind":"fundamentals","access":"commercial","quote":"Data originates from SEC Form 4 filings, linked with CRSP and Compustat-CapitalIQ trade and finance data via personid, cik, and compa-nyid.","mentions":1,"used_in":1},{"id":"crsp","name":"CRSP","provider":"Center for Research in Security Prices","kind":"market-prices","access":"commercial","quote":"Data originates from SEC Form 4 filings, linked with CRSP and Compustat-CapitalIQ trade and finance data via personid, cik, and compa-nyid.","mentions":1,"used_in":1}],"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":"XGBoost identifies unlawful transactions with a high accuracy of 97 percent.","kind":"predictive-accuracy","quote":"The results demonstrate that XGBoost can identify unlawful transactions with a high accuracy of 97 percent and can provide ranking of the features that play the most important role in detecting fraudulent activities."},{"finding":"In optimal configurations XGBoost achieves accuracy exceeding 99 percent.","kind":"performance","quote":"XGBoost achieves accuracy exceeding 99 percent in optimal configurations (Table 4) and demonstrates remarkably strong control over false negative rates, reaching a minimum FNR of 1.02 percent."},{"finding":"The minimum false negative rate achieved by XGBoost is 1.02 percent, marginally lower than the best Random Forest result.","kind":"performance","quote":"reaching a minimum FNR of 1.02 percent, which is marginally lower than the best Random Forest FNR (1.07 percent) reported in Table 5 of [Neupane and Griva, 2024]."},{"finding":"False positive rate drops from ~17 percent with 320 transactions to ~2 percent with the full 3984‑transaction dataset.","kind":"performance","quote":"the implemented method’s FPR averages 17.13 percent with 320 transactions but drops significantly to approximately 2.11 percent with 3984 transactions."},{"finding":"Feature importance rankings differ between mean decrease impurity and permutation importance, highlighting the need for decorrelation.","kind":"statistical","quote":"Permutation Importance before decorrelation ranks features such as Total Debt to Equity, Excess Returns, and Price Operating Earnings (Basic) as most important."}],"reported_performance":null,"out_of_sample":{"method":"cross-validation","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":27,"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":"2026-11-14T17:21:54Z"},"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/"}