{"paper_id":"2510.10878","title":"Identifying and Quantifying Financial Bubbles with the Hyped Log-Periodic Power Law Model","authors":["Zheng Cao","Xingran Shao","Yuheng Yan","Helyette Geman"],"paper_date":"2025-10-13","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2510.10878v1","page_url":"https://thequant.space/flowcharts/identifying-and-quantifying-financial-bubbles-with-the-hyped/","scores":{"empirical_rigor":7,"math_complexity":6.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes the Hyped Log-Periodic Power Law Model (HLPPL), a unified framework that detects both positive and negative financial bubbles by combining LPPL dynamics with sentiment and hype indicators via a dual‑stream transformer.","quote":"We propose a novel model, the Hyped Log-Periodic Power Law Model (HLPPL), to the problem of quantifying and detecting financial bubbles, an ever-fascinating one for academics and practitioners alike.","summary":"We propose a novel model, the Hyped Log-Periodic Power Law Model (HLPPL), to the problem of quantifying and detecting financial bubbles, an ever-fascinating one for academics and practitioners alike. Bubble labels are generated using a Log-Periodic Power Law (LPPL) model, sentiment scores, and a…","strategy":"Bubble Score trading strategy"},"assumptions":{"status":"available","items":[{"assumption":"Bubble price dynamics follow a power‑law acceleration toward a finite‑time singularity.","type":"model","stated":"explicit","quote":"Suppose that the expected growth of prices follows a nonlinear feedback mechanism such that d ln p(t) / dt ∝(tc −t)^{m−1}, 0 < m < 1"},{"assumption":"Log‑periodic oscillations arise from investor herding and increase in frequency as the critical time approaches.","type":"model","stated":"explicit","quote":"By assuming that investor herding generates log-periodic fluctuations, the growth term is decorated with an oscillatory modulation leading to (1)."},{"assumption":"Trading signals are generated using fixed entry/exit thresholds on the normalized Bubble Score.","type":"execution","stated":"explicit","quote":"5. Risk Management: The strategy enforces a 15% stop-loss, caps maximum position size at 50% of capital, and incorporates transaction costs of 0.1%."},{"assumption":"The backtest uses daily price and return data from CRSP and sentiment signals from WSJ, aligned at a daily frequency.","type":"data","stated":"explicit","quote":"The stock-level daily price and return data are sourced from the CRSP database via WRDS, while the news sentiment signals are derived from the Wall Street Journal (WSJ). Both datasets span the period from January 1, 2018 to December 31, 2024, and are aligned at a daily frequency."},{"assumption":"Only firms with at least 100 daily observations are included to ensure reliable LPPL fitting.","type":"data","stated":"explicit","quote":"Each stock requires a minimum of 100 daily observations to ensure statistical reliability of the LPPL fitting process."},{"assumption":"All market‑level features (except Hype Index and FinBERT sentiment scores) are sourced from WRDS.","type":"data","stated":"explicit","quote":"All market-level features (except Hype Index and FinBERT sentiment scores) are sourced from WRDS (Wharton Research Data Services)."}]},"datasets":{"status":"available","items":[{"name":"CRSP daily stock file","provider":"CRSP via WRDS","kind":"market-prices","access":"commercial","period":"2018-01-01 to 2024-12-31","universe":"real estate sector stocks (SIC 6500‑6552, 6798)","quote":"The stock-level daily price and return data are sourced from the CRSP database via WRDS, while the news sentiment signals are derived from the Wall Street Journal (WSJ). Both datasets span the period from January 1, 2018 to December 31, 2024, and are aligned at a daily frequency.","known_id":"crsp"},{"name":"Hype Index","provider":null,"kind":"alternative","access":"unknown","period":null,"universe":null,"quote":"Sector exposure indicator from Hype Index"},{"name":"FinBERT sentiment scores","provider":null,"kind":"text-news","access":"unknown","period":null,"universe":null,"quote":"One-hot: negative sentiment (FinBERT) Aggregated by date"},{"name":"Macro and valuation features","provider":"WRDS","kind":"macro","access":"commercial","period":null,"universe":null,"quote":"Normalized CBOE Volatility Index; market risk proxy","known_id":"wrds"}],"known":[{"id":"wrds","name":"WRDS","provider":"Wharton Research Data Services","kind":"other","access":"commercial","quote":"The stock-level daily price and return data are sourced from the CRSP database via WRDS, while the news sentiment signals are derived from the Wall Street Journal (WSJ).","mentions":3,"used_in":3},{"id":"crsp","name":"CRSP","provider":"Center for Research in Security Prices","kind":"market-prices","access":"commercial","quote":"The stock-level daily price and return data are sourced from the CRSP database via WRDS, while the news sentiment signals are derived from the Wall Street Journal (WSJ).","mentions":2,"used_in":2},{"id":"cboe","name":"Cboe","provider":"Cboe Global Markets","kind":"options","access":"commercial","quote":"n Yes Daily log return: log(Pt/Pt−1) Calculated from price Yes Book-to-market ratio Drop if missing No P/E ratio (extraordinary in-come included) Drop if missing No Price-to-sales ratio Drop if missing No Return on equity Drop if missing No Month extracted from date Integer encoded No Day of month Integer encoded No Table 2: Market-level feature set used in the Transformer input Description Preprocessing Normalized C","mentions":1,"used_in":1}],"sample":{"start":"2018","end":"2024","market":"U.S. equities","asset_class":"equities","frequency":"daily"},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"A flat transaction cost of 0.1% per trade is applied in the backtest.","quote":"The strategy enforces a 15% stop-loss, caps maximum position size at 50% of capital, and incorporates transaction costs of 0.1%.","source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"A flat transaction cost of 0.1% per trade is applied in the backtest.","card_quote":"The strategy enforces a 15% stop-loss, caps maximum position size at 50% of capital, and incorporates transaction costs of 0.1%.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The machine‑learning‑enhanced HLPPL trading strategy yields an average annualized return of 34.13% with a Sharpe ratio of 1.19.","kind":"performance","quote":"When implemented as a trading strategy, the machine-learning–enhanced approach generates an average annualized return of 34.13% with an average Sharpe ratio of 1.19, significantly outperforming the traditional Bubble Score rules strategy, which delivers 16.64% annualized return and a Sharpe ratio of 0.72."},{"finding":"The Transformer model achieves a mean Pearson correlation of 0.625 between predicted and actual Bubble Scores, with MSE = 0.087 and RMSE = 0.295.","kind":"predictive-accuracy","quote":"the empirical analysis shows that the prediction module achieves a mean Pearson correlation of 0.625 with the Bubble Score (MSE = 0.087; RMSE = 0.295)"},{"finding":"Across 32 real‑estate stocks, 30 (93.75%) produced positive returns, with an average annualized return of 16.64% and average Sharpe ratio of 0.72.","kind":"performance","quote":"Total Stocks Analyzed 32 … Stocks with Positive Returns 30 (93.75%)"},{"finding":"The model’s predictive performance remains stable across horizons, with correlation ranging from 0.628 on Day 1 to 0.609 on Day 5.","kind":"predictive-accuracy","quote":"The observed correlations are relatively stable, ranging from 0.628 on Day 1 to 0.609 on Day 5."}],"reported_performance":{"strategy":"Bubble Score trading strategy","sharpe":2.49,"sharpe_benchmark":-0.54,"annual_return_pct":34.13,"total_return_pct":null,"max_drawdown_pct":11.35},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":53,"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":"2026-11-14T17:24:47Z"},"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/"}