Portfolio Optimisation: a Bayesian Model Averaging approach

This paper adopts a Bayesian Model Averaging procedure to forecast excess returns. With a dataset compiling of 78 companies from the FTSE 100, we use in-sample performance to compare BMA with the Historical Expectation, and out- of-sample performance for the comparison of BMA with realized returns a...

Full description

Bibliographic Details
Main Author: Hart, Adao Dante
Format: Dissertation (University of Nottingham only)
Language:English
Published: 2016
Online Access:https://eprints.nottingham.ac.uk/36649/
_version_ 1848795322196688896
author Hart, Adao Dante
author_facet Hart, Adao Dante
author_sort Hart, Adao Dante
building Nottingham Research Data Repository
collection Online Access
description This paper adopts a Bayesian Model Averaging procedure to forecast excess returns. With a dataset compiling of 78 companies from the FTSE 100, we use in-sample performance to compare BMA with the Historical Expectation, and out- of-sample performance for the comparison of BMA with realized returns and the comparison with the market strategy. Our in-sample results are compared with the Historical Expectation strategy, where the Historical Expectation produces a smaller overall error and a greater overall Sharpe ratio in October 2015. For our out-of-sample results, in January 2016, our findings show that the Consumer Goods and Services industries perform better in real time with all three different portfolio choices. We show that over the period November 2015 to July 2016, the BMA portfolio choices that outperform the market strategy are the Global Minimum Variance portfolio without short sales and the Tangency portfolio.
first_indexed 2025-11-14T19:30:15Z
format Dissertation (University of Nottingham only)
id nottingham-36649
institution University of Nottingham Malaysia Campus
institution_category Local University
language English
last_indexed 2025-11-14T19:30:15Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling nottingham-366492017-10-19T17:03:50Z https://eprints.nottingham.ac.uk/36649/ Portfolio Optimisation: a Bayesian Model Averaging approach Hart, Adao Dante This paper adopts a Bayesian Model Averaging procedure to forecast excess returns. With a dataset compiling of 78 companies from the FTSE 100, we use in-sample performance to compare BMA with the Historical Expectation, and out- of-sample performance for the comparison of BMA with realized returns and the comparison with the market strategy. Our in-sample results are compared with the Historical Expectation strategy, where the Historical Expectation produces a smaller overall error and a greater overall Sharpe ratio in October 2015. For our out-of-sample results, in January 2016, our findings show that the Consumer Goods and Services industries perform better in real time with all three different portfolio choices. We show that over the period November 2015 to July 2016, the BMA portfolio choices that outperform the market strategy are the Global Minimum Variance portfolio without short sales and the Tangency portfolio. 2016-09-14 Dissertation (University of Nottingham only) NonPeerReviewed application/pdf en https://eprints.nottingham.ac.uk/36649/2/Diss.pdf Hart, Adao Dante (2016) Portfolio Optimisation: a Bayesian Model Averaging approach. [Dissertation (University of Nottingham only)]
spellingShingle Hart, Adao Dante
Portfolio Optimisation: a Bayesian Model Averaging approach
title Portfolio Optimisation: a Bayesian Model Averaging approach
title_full Portfolio Optimisation: a Bayesian Model Averaging approach
title_fullStr Portfolio Optimisation: a Bayesian Model Averaging approach
title_full_unstemmed Portfolio Optimisation: a Bayesian Model Averaging approach
title_short Portfolio Optimisation: a Bayesian Model Averaging approach
title_sort portfolio optimisation: a bayesian model averaging approach
url https://eprints.nottingham.ac.uk/36649/