logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-02122024-152135


Tipo di tesi
Tesi di laurea magistrale
Autore
MORELLI, GIANLUCA
URN
etd-02122024-152135
Titolo
Can ChatGPT predict stock market trends and overperform the stock market?
Dipartimento
ECONOMIA E MANAGEMENT
Corso di studi
BANCA, FINANZA AZIENDALE E MERCATI FINANZIARI
Relatori
relatore Prof. Cambini, Riccardo
correlatore Schittenhelm, Frank-Andreas
Parole chiave
  • chatgpt
  • stock market
Data inizio appello
26/02/2024
Consultabilità
Non consultabile
Data di rilascio
26/02/2094
Riassunto
In the modern era of information technology, Artificial Intelligence (AI) techniques
are increasingly taking their place in the field of financial analysis.
The aim of this thesis is to explore the potential of artificial intelligence, using a
specific type of AI in the field of equity investing. We will use ChatGPT as a specific
artificial intelligence.
We will focus on one question, "Can ChatGPT predict stock price trends and
overperform the stock market?"
To answer this question, we must first understand what it means to "overperform the
stock market".
In financial terms, "overperforming the market" means having an investment
portfolio that generates returns above those of a benchmark index, in our specific
case the S&P 500 US benchmark index.
Achieving this goal is the ambition of many investors, but it is a rather difficult goal.
In this thesis we are going to analyse the market according to the indications that
ChatGPT will provide, and thus we are going to explore whether ChatGPT's
capabilities can be extended to financial analysis.
ChatGPT is an artificial intelligence language model developed by Open AI that, like all
artificial intelligences, uses learning techniques to understand and generate text in
response to user input.
A very important aspect of our study will be psychic analysis, a technique that AIs use
to determine the emotional orientation of a text.
In a financial context, psychic analysis can be used to measure or analyse news about
companies and give feedback.
What is analysed in this thesis is that ChatGPT, when properly trained and
implemented, can use psychic analysis to make predictions about

market trends and may even be able to outperform the market index.
During the development of the thesis, all these dynamics will be explored with the
aim of understanding whether an AI language model can effectively "overperform the
market"
File