IDEAS home Printed from https://ideas.repec.org/p/arx/papers/2512.11913.html

Return-Decay Residuals and Tail-Risk Forecasting: Timing Artifacts, Conditional Inference, and Cross-Market Evidence

Author

Listed:
  • Chorok Lee

Abstract

Can residuals from fitted return-decay curves predict future factor tail losses, and do they measure crowding? We distinguish economic edge, overlapping performance statistics, and predictive evidence. A Gaussian counterexample produces a same-month crash-risk ratio of 1.61 without predictable future returns. An eight-factor US reconstruction finds no robust advantage under the original specification. A geographic extension tests six factors in Europe, Japan, and Asia Pacific excluding Japan, with matched baselines and separate 2015-2024 and 2025-August 2026 evaluations. For 2015-2024, pooled regional log loss increases by 0.00704 with frozen augmentation and 0.00233 with annual refitting, although Brier rankings differ and a matched-history US comparison favors the residual. Controlled diagnostics link ranking reversals to feature histories and validation-selected penalties. Holding reference penalties fixed largely removes the reversal, while attribution to history and reselection depends on decomposition order. Sparse validation events do not necessarily imply sensitivity to individual validation months. An equivalent-basis experiment isolates regularization effects: large gaps between representations often reflect poor component forecasts rather than substantial gains over a simple Sharpe baseline. Established sequential bounds allow dependent, nonstationary forecast comparisons under a disclosed research filtration; all 48 intervals include zero and permit nontrivial benefits. These findings establish sensitivity to representation, target definition, and scoring rule, without identifying crowding or establishing prospective forecasting value. Forecasts, unfavorable findings, configurations, and provenance are retained.

Suggested Citation

  • Chorok Lee, 2025. "Return-Decay Residuals and Tail-Risk Forecasting: Timing Artifacts, Conditional Inference, and Cross-Market Evidence," Papers 2512.11913, arXiv.org, revised Sep 2026.
  • Handle: RePEc:arx:papers:2512.11913
    as

    Download full text from publisher

    File URL: https://arxiv.org/pdf/2512.11913
    File Function: Latest version
    Download Restriction: no
    ---><---

    More about this item

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:arx:papers:2512.11913. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    We have no bibliographic references for this item. You can help adding them by using this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: arXiv administrators (email available below). General contact details of provider: https://arxiv.org/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.