{"paper":{"title":"Decision-Induced Ranking Explains Prediction Inflation and Excessive Turnover in SPO-Based Portfolio Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"SPO-based portfolio optimization inflates return predictions and drives excessive turnover because decisions reduce to ranking over risk- and transaction-cost-adjusted marginal scores.","cross_cats":["q-fin.CP"],"primary_cat":"q-fin.PM","authors_text":"Takashi Hasuike, Yi Wang","submitted_at":"2026-05-02T00:48:16Z","abstract_excerpt":"Decision-focused learning (DFL) is attractive for portfolio optimization because it trains predictors according to downstream decision quality rather than prediction accuracy alone. However, SPO(Smart, Predict then Optimize surrogate)-based DFL may produce inflated return signals and unstable portfolio reallocations. This study provides a KKT-based interpretation showing that portfolio decisions can be viewed as ranking over risk- and transaction-cost-adjusted marginal scores. Empirically, we examine prediction inflation and excessive turnover in SPO-trained portfolios, and evaluate clipping, "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This study provides a KKT-based interpretation showing that portfolio decisions can be viewed as ranking over risk- and transaction-cost-adjusted marginal scores.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the observed prediction inflation and excessive turnover are primarily caused by the decision-induced ranking mechanism rather than other modeling choices, data characteristics, or optimization details.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SPO-based decision-focused learning inflates predictions and causes excessive portfolio turnover because decisions rank assets on risk- and cost-adjusted scores, which can be mitigated by output constraints and partial adjustments.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"SPO-based portfolio optimization inflates return predictions and drives excessive turnover because decisions reduce to ranking over risk- and transaction-cost-adjusted marginal scores.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7f023ef1c27487976c44b6aa8298823d4d65629afb2b421b06c4208644c7235a"},"source":{"id":"2605.01176","kind":"arxiv","version":2},"verdict":{"id":"0ead1695-c24e-467e-aedc-219c02431dfb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T15:54:39.302655Z","strongest_claim":"This study provides a KKT-based interpretation showing that portfolio decisions can be viewed as ranking over risk- and transaction-cost-adjusted marginal scores.","one_line_summary":"SPO-based decision-focused learning inflates predictions and causes excessive portfolio turnover because decisions rank assets on risk- and cost-adjusted scores, which can be mitigated by output constraints and partial adjustments.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the observed prediction inflation and excessive turnover are primarily caused by the decision-induced ranking mechanism rather than other modeling choices, data characteristics, or optimization details.","pith_extraction_headline":"SPO-based portfolio optimization inflates return predictions and drives excessive turnover because decisions reduce to ranking over risk- and transaction-cost-adjusted marginal scores."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.01176/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T18:37:10.085013Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:31:12.207732Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"7d065ba4673d6099e560e093c60c0c2b881eb3820f423e9f0986b118e780e4d8"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}