{"paper_id":"2501.15828","title":"Hybrid Quantum Neural Networks with Amplitude Encoding: Advancing Recovery Rate Predictions","authors":["Ying Chen","Paul Griffin","Paolo Recchia","Lei Zhou","Hongrui Zhang"],"paper_date":"2025-01-27","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2501.15828v5","page_url":"https://thequant.space/flowcharts/hybrid-quantum-neural-networks-with-amplitude-encoding-adva/","scores":{"empirical_rigor":8,"math_complexity":7.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a hybrid quantum‑machine‑learning model that uses amplitude encoding for recovery rate prediction.","quote":"We propose a hybrid Quantum Machine Learning (QML) model with Amplitude Encoding, leveraging the unitarity constraint of Parametrized Quantum Circuits (PQC) and the exponential data compression capability of qubits.","summary":"Recovery rate prediction plays a pivotal role in bond investment strategies by enhancing risk assessment, optimizing portfolio allocation, improving pricing accuracy, and supporting effective credit risk management. However, accurate forecasting remains challenging due to complex nonlinear…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Recovery amount of each bond is defined as the bond’s price 30 days after the default date.","type":"data","stated":"explicit","quote":"The recovery amount of each bond in this study is defined as the bond’s price 30 days after the default date."},{"assumption":"All QML models are trained and evaluated on quantum simulators, which provide noiseless training and evaluation.","type":"model","stated":"explicit","quote":"Simulation allows for noiseless training and evaluation, which does not fully reflect the limitations of existing quantum devices."},{"assumption":"The dataset is high‑dimensional (256 features) but limited in sample size (1,725 observations), making overfitting a concern for classical models.","type":"data","stated":"explicit","quote":"complex nonlinear dependencies, high-dimensional feature spaces, and limited sample sizes—conditions under which classical machine learning models are prone to overfitting."}]},"datasets":{"status":"available","items":[{"name":"global recovery rate dataset","provider":"NRF Research Project UP5 of the National University of Singa","kind":"fundamentals","access":"proprietary","period":"1996 to 2023","universe":"576 firms","quote":"We evaluate the model on a global recovery rate dataset comprising 1,725 observations and 256 features from 1996 to 2023."}],"known":[{"id":"bloomberg","name":"Bloomberg","provider":"Bloomberg L.P.","kind":"market-prices","access":"commercial","quote":"The UP5 data contains Macroeco-nomic and market-related features obtained from FRED and Refinitiv; financial statement features of firms sourced from Bloomberg; and bond-level features, as well as firm-level or market-level credit product features provided by the Credit Research Initiative (CRI) of the National University of Singapore.","mentions":4,"used_in":4},{"id":"fred","name":"FRED","provider":"Federal Reserve Bank of St. Louis","kind":"macro","access":"public","quote":"The UP5 data contains Macroeco-nomic and market-related features obtained from FRED and Refinitiv; financial statement features of firms sourced from Bloomberg; and bond-level features, as well as firm-level or market-level credit product features provided by the Credit Research Initiative (CRI) of the National University of Singapore.","mentions":2,"used_in":2},{"id":"refinitiv","name":"Refinitiv / LSEG (Datastream, Eikon, Tick History)","provider":"LSEG","kind":"market-prices","access":"commercial","quote":"The UP5 data contains Macroeco-nomic and market-related features obtained from FRED and Refinitiv; financial statement features of firms sourced from Bloomberg; and bond-level features, as well as firm-level or market-level credit product features provided by the Credit Research Initiative (CRI) of the National University of Singapore.","mentions":2,"used_in":2}],"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 amplitude‑encoding QML model achieves the lowest RMSE of 0.228, outperforming classical neural networks (0.246) and angle‑encoding QML (0.242).","kind":"predictive-accuracy","quote":"Our hybrid method significantly outperforms both classical neural networks and QML models using Angle Encoding, achieving a lower Root Mean Squared Error (RMSE) of 0.228, compared to 0.246 and 0.242, respectively."},{"finding":"Amplitude‑encoding QML trains faster (0.73 s per epoch) than angle‑encoding QML (0.81 s per epoch).","kind":"performance","quote":"the QML model with Amplitude Encoding has fewer trainable parameters, leading to a faster training time of 0.73 seconds per epoch compared to the 0.81 seconds necessary for the QML with Angle Encoding."},{"finding":"Compared to XGBoost, amplitude‑encoding QML does not have a clearly superior average RMSE, but its variance is smaller and it outperforms XGBoost in leave‑one‑out cross‑validation.","kind":"performance","quote":"When compared to XGBoost, the proposed QML model with Amplitude Encoding does not demonstrate clearly superior performance. In fact, it yields a higher average RMSE in cross-validation settings, albeit with a smaller standard deviation."},{"finding":"The model’s practical applicability is limited by the need for deep circuits on real quantum hardware, which are more prone to noise and decoherence.","kind":"negative","quote":"its implementation on actual quantum devices remains challenging due to the deep circuits required, which are more prone to noise and decoherence"}],"reported_performance":null,"out_of_sample":{"method":"cross-validation","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[{"paper":{"paper_id":"2306.12965","title":"Improved Financial Forecasting via Quantum Machine Learning","paper_date":"2023-05-31","authors":["Sohum Thakkar","Skander Kazdaghli","Natansh Mathur","Iordanis Kerenidis","André J. Ferreira-Martins","Samurai Brito"],"empirical_rigor":4,"page_url":"https://thequant.space/flowcharts/improved-financial-forecasting-via-quantum-machine-learning/"},"question":"Can quantum-inspired techniques improve financial forecasting accuracy?","this_paper_finding":"The model’s practical applicability is limited by the need for deep circuits on real quantum hardware, which are more prone to noise and decoherence.","other_paper_finding":"Current quantum hardware is not powerful enough to provide real improvements; performance gains can be achieved today by turning quantum ideas into classical ML solutions.","other_paper_quote":"In the present day, quantum hardware is not powerful enough to provide real improvements or conclusive large‑scale benchmarks. Performance enhancements can be achieved today by turning these quantum ideas into classical ML solutions run on GPUs.","note":"Both findings address this, but A notes current hardware limits and suggests classical translations, while B warns deep quantum circuits suffer noise and decoherence.","confidence":0.76}],"contradicts":[],"qualifies":[],"compared_with":19,"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":"2026-11-14T17:46:32Z"},"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/"}