{"paper_id":"2511.15960","title":"Machine Learning vs. Randomness: Challenges in Predicting Binary Options Movements","authors":["Gabriel M. Arantes","Richard F. Pinto","Bruno L. Dalmazo","Eduardo N. Borges","Giancarlo Lucca","Viviane L. D. de Mattos","Fabian C. Cardoso","Rafael A. Berri"],"paper_date":"2025-11-20","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2511.15960v1","page_url":"https://thequant.space/flowcharts/machine-learning-vs.-randomness-challenges-in-predicting-bi/","scores":{"empirical_rigor":7.5,"math_complexity":2.5,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper evaluates a range of machine learning models against a ZeroR baseline to assess predictability of binary options movements.","quote":"This study demonstrates that machine learning algorithms struggle to outperform a simple baseline in predicting binary options movements.","summary":"Binary options trading is often marketed as a field where predictive models can generate consistent profits. However, the inherent randomness and stochastic nature of binary options make price movements highly unpredictable, posing significant challenges for any forecasting approach. This study…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Binary options price movements are entirely random, providing no predictable patterns.","type":"market","stated":"explicit","quote":"the hypothesis that binary options are entirely random"},{"assumption":"Technical indicators (SMA and RSI) are sufficient features to capture any potential predictive signal.","type":"data","stated":"explicit","quote":"We derived a set of technical indicators: Simple Moving Averages (SMA) with windows of 2, 3, 4, and 5 minutes"},{"assumption":"Training and test sets are split chronologically to avoid information leakage.","type":"data","stated":"explicit","quote":"The dataset was divided into a training set (2021–2022, N = 742,240) and a testing set (2023, N = 322,572) to ensure models learn from historical patterns without information leakage."},{"assumption":"Accuracy is an appropriate metric for evaluating binary classification performance.","type":"statistical","stated":"explicit","quote":"Accuracy was the main performance metric, enabling direct comparison."}]},"datasets":{"status":"available","items":[{"name":"Historical EUR/USD minute data","provider":"HistData","kind":"market-prices","access":"public","period":"2021-2023","universe":"EUR/USD","quote":"We utilized a historical Forex dataset from HistData, consisting of minute-by-minute Euro/Dollar (EUR/USD) data from 2021–2023."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"intraday-bars"},"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":"unclear","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"No model surpassed the ZeroR baseline accuracy of 0.5389; Logistic Regression matched it while others performed worse.","kind":"negative","quote":"only Logistic Regression matched the ZeroR baseline (0.5389), while others performed worse."},{"finding":"After hyper‑parameter tuning, Random Forest, Gradient Boosting, and Logistic Regression each reached 0.5389 but still did not exceed the baseline.","kind":"negative","quote":"... improved the accuracy of Random Forest, Gradient Boosting, and kNN. However, despite these improvements, no model surpassed the ZeroR baseline of 0.5389"},{"finding":"Both the MLP and LSTM neural networks achieved an accuracy of 0.5389, identical to the ZeroR model.","kind":"negative","quote":"the MLP achieved an accuracy of 0.5389, identical to the ZeroR model"},{"finding":"Feature selection retained only RSI features, with SMA indicators having negligible impact on accuracy.","kind":"statistical","quote":"Feature selection using SelectFromModel with a Random Forest classifier retained only the four RSI features"},{"finding":"Models converged to predicting the majority class, showing a failure to learn any true predictive signal.","kind":"negative","quote":"most algorithms, including the LSTM and MLP, converged to a simplistic strategy of predicting the majority class, as evidenced by their confusion matrices."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":82,"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":"2026-11-14T17:21:13Z"},"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/"}