Privacy-preserving portfolio pricing

Gilad Asharov, Tucker Balch, Antigoni Polychroniadou

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Investment banks offer the service of purchasing an entire portfolio from a client in a single transaction. The transaction enables the client to remove unwanted assets from their books, and provides immediate cash to be used in other investments. The bank usually offers a price that is slightly lower than the market price, therefore potentially gaining in this transaction when selling its content later in the market, or by internalizing the portfolio. Even though liquidating the portfolio is advantageous for the client, they are concerned about information regarding their strategies or holdings "leaking"and they would also like to be able to "shop"the portfolio to multiple banks. The current industry practice calls for the client to provide a summarized description of the portfolio without full details of the constituents, leading to imprecise pricing by the banks. We propose a new way for pricing portfolios by adapting secure computation to this domain. Our approach allows the client to maintain secrecy regarding the constituents of the portfolio, while enabling the bank to provide a fair price. We study several metrics for pricing portfolios and provide a suite of two-party protocols for computing those metrics, which are all provably secure. We test our protocols and show their scalability experimentally. We believe that this privacy-preserving pricing method offers the potential to transform the practice of portfolio pricing.

Original languageEnglish
Title of host publicationICAIF 2021 - 2nd ACM International Conference on AI in Finance
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9781450391481
StatePublished - 3 Nov 2021
Externally publishedYes
Event2nd ACM International Conference on AI in Finance, ICAIF 2021 - Virtual, Online
Duration: 3 Nov 20215 Nov 2021

Publication series

NameICAIF 2021 - 2nd ACM International Conference on AI in Finance


Conference2nd ACM International Conference on AI in Finance, ICAIF 2021
CityVirtual, Online

Bibliographical note

Publisher Copyright:
© 2021 ACM.


  • portfolio pricing
  • secure computation


Dive into the research topics of 'Privacy-preserving portfolio pricing'. Together they form a unique fingerprint.

Cite this