Statistics > Methodology
[Submitted on 6 Jun 2016 (v1), revised 11 Oct 2016 (this version, v2), latest version 13 Sep 2017 (v4)]
Title:Regularizing Bayesian Predictive Regressions
View PDFAbstract:We develop a novel approach for regularizing Bayesian Predictive Regressions. By jointly regularizing both expected values and variance-covariance matrices in vector autoregressions (VAR) and seemingly unrelated regressions (SUR), we provide a framework for out-of-sample forecasting in asset pricing. To illustrate our methodology, we analyze the asset allocation decision in Barberis (2000) and Kandel and Stambaugh (1996) based on the equity premium predictability. By exploiting a duality between regularization penalties and prior distributions, we re-interpret the Bayesian analysis regarding our regularization procedure. To perform a prior sensitivity analysis, we use a regularization path with predictive cross-validation to determine the optimal level of regularization and predictor selection under shrinkage priors. Finally, we conclude there exist plausible prior specifications for return predictability for excess S&P 500 index returns using Book-to-Market Ratio, Dividend-Payout Ratio, and Term Spread predictors.
Submission history
From: Guanhao Feng [view email][v1] Mon, 6 Jun 2016 11:44:09 UTC (36 KB)
[v2] Tue, 11 Oct 2016 15:32:12 UTC (68 KB)
[v3] Mon, 3 Apr 2017 20:49:46 UTC (73 KB)
[v4] Wed, 13 Sep 2017 19:31:11 UTC (88 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.