Umeå University's logo

umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Exploring Bayesian Vector Autoregression Models In Portfolio Allocation
Umeå University, Faculty of Science and Technology, Department of Mathematics and Mathematical Statistics.
Umeå University, Faculty of Science and Technology, Department of Mathematics and Mathematical Statistics.
2024 (English)Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

This thesis investigate the application of Bayesian Vector Autoregression (BVAR) models in portfolio allocation strategies within a subset of the US financial market. Utilizing three models with different prior distributions, this study examined the predictive accuracy of financial returns, aiming to enhance strategic decision-making in portfolio allocation strategies. Using several model based allocation strategies the three BVAR models with Diffuse, Minnesota, and Horseshoe priors were tested across multiple portfolios to evaluate their performance against the equally weighted portfolio strategy acting as a benchmark. The study includes financial market data and macroeconomic indicators, analyzed over a distinct training, validation, and test period. A rolling training window with a weekly model based forecast was used to evaluate the models and the strategies performance. Performance metrics such as the Sharpe ratio, Sortino ratio, and Maximum Drawdown were employed to assess the strategies. Despite varied results, the thesis explores the potential of incorporating prior distributions and uncertainty into financial forecasts to improve asset allocation outcomes. 

Abstract [sv]

Denna avhandling utforskar användningen av Bayesiansk Vektor Autoregressions (BVAR) modeller inom portföljallokering, med aktier från den Amerikanska finansmarknaden. I studien används tre olika BVAR-modeller, var och en baserad på olika prior fördelningar: Diffuse, Minnesota och Horseshoe. Syftet med analysen är att bedöma hur väl dessa modeller kan förutsäga finansiella avkastningar och därigenom förbättra beslutsfattandet inom portföljhantering. Genom tillämplingen av flera strategier baserade på modellerna testas dessa strategier över flera portföljer och jämförs mot en likaviktad portfölj. Studien omfattar finansiell marknadsdata och makroekonomiska indikatorer som analyseras under en tränings, validering och test period. Ett rullande träningsfönster med veckovisa prediktioner från modellen används för att kontinuerligt utvärdera både modellernas och strategiernas prestanda. För att bedöma strategierna används prestandamått som Sharpe-ratio, Sortino-ratio och maximal nedgång. Trots blandade resultat belyser avhandlingen potentialen i att integrera fördelningar och osäkerhet inom finansiella prognoser för att förbättra resultatet av portföljallokering.

Place, publisher, year, edition, pages
2024. , p. 37
National Category
Mathematics
Identifiers
URN: urn:nbn:se:umu:diva-226191OAI: oai:DiVA.org:umu-226191DiVA, id: diva2:1869951
External cooperation
Nepa Sweden AB
Educational program
Master of Science in Engineering and Management
Supervisors
Examiners
Available from: 2024-06-14 Created: 2024-06-13 Last updated: 2024-06-14Bibliographically approved

Open Access in DiVA

Exploring Bayesian Vector Autoregression Models In Portfolio Allocation(7645 kB)392 downloads
File information
File name FULLTEXT01.pdfFile size 7645 kBChecksum SHA-512
d516ed7f35d5a3c47ff42fe553dfd29d259784d472e6d28bad6337a1645fb65c4e2f7425cb88d7bc29df36ebd771da3a0171b159c2bce0a717b421bab44cc035
Type fulltextMimetype application/pdf

By organisation
Department of Mathematics and Mathematical Statistics
Mathematics

Search outside of DiVA

GoogleGoogle Scholar
Total: 392 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 627 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf