(Over)performance in 13F filings

: extracting investment managers' alpha generating ideas

Abstract

The dissertation assesses the performance of a portfolio constructed using a piggyback strategy, leveraging disclosures from a subset of investment companies selected based on their holdings and assets under management. To address the research question, a portfolio comprising twenty securities was assembled by selecting "winning" stocks identified within the overall dataset. These selections were made according to three primary criteria: consensus, conviction, and momentum. The findings indicate that the strategy outlined in the dissertation failed to yield outperformance between May 2018 and November 2023 despite achieving a similar performance of the S&P500 index (before expenses). This conclusion is drawn from the analysis of both the CAPM alpha (which is found to be significant but close to zero) and the Fama and French 4-factor model (which shows a significant but small negative alpha).

Date of Award	17 Jul 2024
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Paulo Alves (Supervisor)