1{"text": "{\"genre\":\"mixed_theory_empirical\",\"subfield\":\"asset_pricing\",\"subfield_secondary\":\"corporate_finance (capital structure / structural credit risk models)\",\"primary_method\":\"reduced_form_regression\",\"identification\":{\"estimand\":\"Cross-sectional pricing of unlevered market beta, size, BTM and volatility in leverage-adjusted (unlevered) equity returns, compared with levered stock returns\",\"strategy\":\"Unlever stock returns in two ways: data-driven scaling by (1 - leverage), and a parametric Merton or Leland-Toft structural model. Then run portfolio sorts and Fama-MacBeth regressions on the unlevered returns.\",\"key_assumptions\":[\"Linear asset pricing models allow levered risk to be decomposed into asset risk plus leverage-related risk\",\"The structural credit model (Merton/Leland-Toft) correctly maps equity returns to asset returns\",\"Book liabilities and assumed debt maturity proxy adequately for the debt structure\"],\"threats_addressed\":[\"Model dependence of the unlevering (checked with Leland-Toft and the scaling approach)\",\"Debt maturity and face-value measurement choices\",\"Portfolio weighting choices\",\"Nonlinearity of leverage that regression controls fail to capture (heteroskedasticity)\"],\"causal_language_strength\":2,\"uses_causal_claim\":false},\"abstract_moves\":{\"question\":{\"present\":true,\"sentence_index\":0},\"gap\":{\"present\":false,\"sentence_index\":-1},\"method\":{\"present\":true,\"sentence_index\":0},\"results\":{\"present\":true,\"sentence_index\":1},\"magnitude\":{\"present\":false,\"sentence_index\":-1},\"contribution\":{\"present\":true,\"sentence_index\":4}},\"introduction\":{\"moves\":[{\"paragraph_index\":0,\"move\":\"motivation\"},{\"paragraph_index\":0,\"move\":\"tension_or_puzzle\"},{\"paragraph_index\":0,\"move\":\"literature_gap\"},{\"paragraph_index\":0,\"move\":\"contribution_statement\"},{\"paragraph_index\":0,\"move\":\"mechanism_or_theory_sketch\"},{\"paragraph_index\":1,\"move\":\"preview_of_results\"},{\"paragraph_index\":1,\"move\":\"robustness_mention\"}],\"contribution_paragraph_index\":0,\"contribution_types\":[\"new_evidence\",\"new_theory_or_model\",\"resolution_of_puzzle\"],\"roadmap_present\":false,\"literature_positioning_location\":\"inside_introduction\"},\"section_roles\":[{\"idx\":0,\"role\":\"introduction\"},{\"idx\":1,\"role\":\"theory\"},{\"idx\":2,\"role\":\"data\"},{\"idx\":3,\"role\":\"results\"},{\"idx\":4,\"role\":\"results\"},{\"idx\":5,\"role\":\"robustness\"},{\"idx\":6,\"role\":\"results\"},{\"idx\":7,\"role\":\"conclusion\"},{\"idx\":8,\"role\":\"appendix\"}],\"results_presentation\":{\"economic_significance_discussed\":true,\"statistical_significance_language\":\"stars_and_tstats\",\"magnitude_statements_in_intro\":false},\"robustness\":{\"location\":\"main_text_section\",\"types\":[\"alternative_specification\",\"alternative_measure\",\"sensitivity_to_parameters\",\"other\"]},\"tables\":{\"convention_notes\":\"Panels report value-weighted returns for 25 double-sorted or 5 single-sorted portfolios, followed by Fama-MacBeth regression panels with t-statistics in parentheses. Notes define the leverage measure and its timing, and give the simulation parameters.\",\"units_stated_in_notes\":true,\"standard_errors_description\":\"unclear\"},\"figures\":{\"uncertainty_shown\":\"unclear\"},\"conclusion\":{\"bounded_claims\":true,\"policy_implications\":false,\"open_questions\":false,\"restates_contribution\":true},\"hedging\":{\"level\":\"moderate\",\"typical_devices\":[\"'in part' qualifiers\",\"'suggest'\",\"'virtually disappear' / 'weakened'\",\"'consistent with theoretical predictions'\"]},\"claim_evidence\":[{\"claim\":\"Unlevered market beta is more strongly priced than levered beta, even after controlling for size and BTM\",\"evidence_type\":\"table\",\"evidence_ref\":\"Tables I, III, IV; Figure 2\"},{\"claim\":\"The value premium virtually disappears and the size effect weakens for unlevered returns\",\"evidence_type\":\"table\",\"evidence_ref\":\"Tables III, IV\"},{\"claim\":\"Leverage induces heteroskedasticity in stock returns that unlevering removes\",\"evidence_type\":\"figure\",\"evidence_ref\":\"Figure 1, Figure 3\"},{\"claim\":\"Controlling for leverage in regressions does not adequately address its nonlinear effect\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table V\"},{\"claim\":\"The volatility puzzle holds for levered returns but not for unlevered returns\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table VIII\"},{\"claim\":\"Results are robust to debt maturity, face value of debt, the Leland-Toft model and weighting choices\",\"evidence_type\":\"table\",\"evidence_ref\":\"Tables VI, VII\"},{\"claim\":\"Derivative of expected levered returns with respect to levered volatility is negative\",\"evidence_type\":\"equation_or_proof\",\"evidence_ref\":\"Appendix, Equation (3)\"}],\"structural_summary\":\"The paper opens with a question and motivates it with well-known anomalies, then positions itself as the first to test hypotheses on unlevered returns. A theory section derives predictions for asset beta, size, BTM and volatility. After the data section, the empirical work unlevers returns twice: first by data-driven leverage scaling, then with the Merton model, and shows why leverage controls in regressions are inadequate. A dedicated robustness section varies the model and measurement choices, and a separate section treats volatility. The conclusion restates the contribution, and an appendix contains a proof.\",\"writing_profile_keywords\":[\"unlevered returns\",\"anomalies\",\"structural credit models\",\"Fama-MacBeth\",\"portfolio sorts\",\"theory-derived hypotheses\",\"parsimonious factor model\"],\"confidence\":0.8,\"notes\":\"The introduction excerpt is truncated, so a roadmap may exist but is not visible. The standard-error type is not stated in the visible excerpts.\"}", "json": {"genre": "mixed_theory_empirical", "subfield": "asset_pricing", "subfield_secondary": "corporate_finance (capital structure / structural credit risk models)", "primary_method": "reduced_form_regression", "identification": {"estimand": "Cross-sectional pricing of unlevered market beta, size, BTM and volatility in leverage-adjusted (unlevered) equity returns, compared with levered stock returns", "strategy": "Unlever stock returns in two ways: data-driven scaling by (1 - leverage), and a parametric Merton or Leland-Toft structural model. Then run portfolio sorts and Fama-MacBeth regressions on the unlevered returns.", "key_assumptions": ["Linear asset pricing models allow levered risk to be decomposed into asset risk plus leverage-related risk", "The structural credit model (Merton/Leland-Toft) correctly maps equity returns to asset returns", "Book liabilities and assumed debt maturity proxy adequately for the debt structure"], "threats_addressed": ["Model dependence of the unlevering (checked with Leland-Toft and the scaling approach)", "Debt maturity and face-value measurement choices", "Portfolio weighting choices", "Nonlinearity of leverage that regression controls fail to capture (heteroskedasticity)"], "causal_language_strength": 2, "uses_causal_claim": false}, "abstract_moves": {"question": {"present": true, "sentence_index": 0}, "gap": {"present": false, "sentence_index": -1}, "method": {"present": true, "sentence_index": 0}, "results": {"present": true, "sentence_index": 1}, "magnitude": {"present": false, "sentence_index": -1}, "contribution": {"present": true, "sentence_index": 4}}, "introduction": {"moves": [{"paragraph_index": 0, "move": "motivation"}, {"paragraph_index": 0, "move": "tension_or_puzzle"}, {"paragraph_index": 0, "move": "literature_gap"}, {"paragraph_index": 0, "move": "contribution_statement"}, {"paragraph_index": 0, "move": "mechanism_or_theory_sketch"}, {"paragraph_index": 1, "move": "preview_of_results"}, {"paragraph_index": 1, "move": "robustness_mention"}], "contribution_paragraph_index": 0, "contribution_types": ["new_evidence", "new_theory_or_model", "resolution_of_puzzle"], "roadmap_present": false, "literature_positioning_location": "inside_introduction"}, "section_roles": [{"idx": 0, "role": "introduction"}, {"idx": 1, "role": "theory"}, {"idx": 2, "role": "data"}, {"idx": 3, "role": "results"}, {"idx": 4, "role": "results"}, {"idx": 5, "role": "robustness"}, {"idx": 6, "role": "results"}, {"idx": 7, "role": "conclusion"}, {"idx": 8, "role": "appendix"}], "results_presentation": {"economic_significance_discussed": true, "statistical_significance_language": "stars_and_tstats", "magnitude_statements_in_intro": false}, "robustness": {"location": "main_text_section", "types": ["alternative_specification", "alternative_measure", "sensitivity_to_parameters", "other"]}, "tables": {"convention_notes": "Panels report value-weighted returns for 25 double-sorted or 5 single-sorted portfolios, followed by Fama-MacBeth regression panels with t-statistics in parentheses. Notes define the leverage measure and its timing, and give the simulation parameters.", "units_stated_in_notes": true, "standard_errors_description": "unclear"}, "figures": {"uncertainty_shown": "unclear"}, "conclusion": {"bounded_claims": true, "policy_implications": false, "open_questions": false, "restates_contribution": true}, "hedging": {"level": "moderate", "typical_devices": ["'in part' qualifiers", "'suggest'", "'virtually disappear' / 'weakened'", "'consistent with theoretical predictions'"]}, "claim_evidence": [{"claim": "Unlevered market beta is more strongly priced than levered beta, even after controlling for size and BTM", "evidence_type": "table", "evidence_ref": "Tables I, III, IV; Figure 2"}, {"claim": "The value premium virtually disappears and the size effect weakens for unlevered returns", "evidence_type": "table", "evidence_ref": "Tables III, IV"}, {"claim": "Leverage induces heteroskedasticity in stock returns that unlevering removes", "evidence_type": "figure", "evidence_ref": "Figure 1, Figure 3"}, {"claim": "Controlling for leverage in regressions does not adequately address its nonlinear effect", "evidence_type": "table", "evidence_ref": "Table V"}, {"claim": "The volatility puzzle holds for levered returns but not for unlevered returns", "evidence_type": "table", "evidence_ref": "Table VIII"}, {"claim": "Results are robust to debt maturity, face value of debt, the Leland-Toft model and weighting choices", "evidence_type": "table", "evidence_ref": "Tables VI, VII"}, {"claim": "Derivative of expected levered returns with respect to levered volatility is negative", "evidence_type": "equation_or_proof", "evidence_ref": "Appendix, Equation (3)"}], "structural_summary": "The paper opens with a question and motivates it with well-known anomalies, then positions itself as the first to test hypotheses on unlevered returns. A theory section derives predictions for asset beta, size, BTM and volatility. After the data section, the empirical work unlevers returns twice: first by data-driven leverage scaling, then with the Merton model, and shows why leverage controls in regressions are inadequate. A dedicated robustness section varies the model and measurement choices, and a separate section treats volatility. The conclusion restates the contribution, and an appendix contains a proof.", "writing_profile_keywords": ["unlevered returns", "anomalies", "structural credit models", "Fama-MacBeth", "portfolio sorts", "theory-derived hypotheses", "parsimonious factor model"], "confidence": 0.8, "notes": "The introduction excerpt is truncated, so a roadmap may exist but is not visible. The standard-error type is not stated in the visible excerpts."}, "model": "claude-opus-5-5", "stop_reason": "end_turn", "latency_s": 17.34, "usage": {"input": 6351, "output": 2125, "cache_read": 3968, "cache_write": 0}, "cost_usd": 0.068698, "request_id": "req_011CggqMveUkwtMKV3FMVLW2", "cached": false, "ts": "2026-11-03T23:07:31+00:00", "effort": "medium"}