We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Quantitative Finance > Risk Management

Title: Proving prediction prudence

Authors: Dirk Tasche
Abstract: We study how to perform tests on samples of pairs of observations and predictions in order to assess whether or not the predictions are prudent. Prudence requires that that the mean of the difference of the observation-prediction pairs can be shown to be significantly negative. For safe conclusions, we suggest testing both unweighted (or equally weighted) and weighted means and explicitly taking into account the randomness of individual pairs. The test methods presented are mainly specified as bootstrap and normal approximation algorithms. The tests are general but can be applied in particular in the area of credit risk, both for regulatory and accounting purposes.
Comments: 21 pages, 2 appendices
Subjects: Risk Management (q-fin.RM); Applications (stat.AP)
MSC classes: 91G40, 91G70
Journal reference: Data Science in Finance and Economics, 2022, 2(4):359-379
DOI: 10.3934/DSFE.2022017
Cite as: arXiv:2005.03698 [q-fin.RM]
  (or arXiv:2005.03698v3 [q-fin.RM] for this version)

Submission history

From: Dirk Tasche [view email]
[v1] Thu, 7 May 2020 18:47:12 GMT (21kb)
[v2] Wed, 24 Jun 2020 17:19:01 GMT (21kb)
[v3] Fri, 30 Sep 2022 16:28:25 GMT (24kb)

Link back to: arXiv, form interface, contact.