Similar corporate bonds selection by clustering algorithms to increase investment efficiency

Portfolio managers (PM) are limited in their ability to purchase bonds based on the inventory currently available on the market. Although a PM may wish to purchase a very specific bond to satisfy a portfolio requirement, the bond may not be available at the appropriate price or in desired quantity.To address this issue, PMs may substitute their ideal bond with similar bonds that are more readily available. Due to the large sizes of the corporate bond inventory it is not possible for a human analyst to come up with all the information and features that will help in a systematically decision-making process. The objective of this research is to develop a systematic methodology based on unsupervised learning clustering algorithms, which can potentially help portfolio managers and traders efficiently identify a list of similar bonds that meet the same criteria as their original selection. Similarity is defined by the features of a bond, with a preference for bonds of comparable issuer, rating, sector and duration to the ideal bond. In addition, the research will provide feature importance for each industry sector and could potentially provide new insights on industry sector investment.

Author

• Wu, Boyu

Language

• en

Keyword

• Machine Learning
• Low liquidity Security
• Unsupervised Learning
• Trading Strategy
• Portfolio Management
• Bond Investment

Date created

• 2020

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Engineering Management

Degree

• D.Engr.

Advisor

• Etemadi, Amir

Committee Member(s)

• Adetunji, Tomi
• Kamp, John

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/xp68kg901