This thesis analyzes the use of matrix cointegrated models in the context of quantitative pairs trading strategies, with particular attention to the comparison between the MECM (Matrix Error Correction Model) and the ECCMAR (Error Correction Cointegrated Matrix Autoregressive) model. After a brief introduction to the main concepts of cointegration and error correction models, the thesis presents an identification procedure based on Johansen’s trace test, applied to the rows and columns of the data matrix in order to determine the most appropriate model between MECM and ECCMAR. The empirical analysis focuses on the application of these models to Fama–French portfolios observed from the 1990s onward, with the aim of identifying long-run equilibrium relationships and evaluating their use in the implementation of pairs trading strategies.
Questa tesi analizza l’utilizzo di modelli matriciali cointegrati nell’ambito delle strategie quantitative di pairs trading, con particolare attenzione al confronto tra il modello MECM (Matrix Error Correction Model) e il modello ECCMAR (Error Correction Cointegrated Matrix Autoregressive). Dopo una breve introduzione ai principali concetti di cointegrazione e modelli a correzione dell’errore, il lavoro presenta una procedura di identificazione basata sul test della traccia di Johansen, applicato alle righe e alle colonne della matrice dei dati, al fine di determinare il modello più appropriato tra MECM ed ECCMAR. La parte empirica si concentra sull’applicazione di tali modelli a portafogli Fama–French osservati a partire dagli anni Novanta, con l’obiettivo di individuare relazioni di equilibrio di lungo periodo e valutarne l’utilizzo nell’implementazione di strategie di pairs trading.
Modelli autoregressivi matriciali cointegrati per serie temporali: un’applicazione al pair trading
BENVENUTI, ENRICO
2025/2026
Abstract
This thesis analyzes the use of matrix cointegrated models in the context of quantitative pairs trading strategies, with particular attention to the comparison between the MECM (Matrix Error Correction Model) and the ECCMAR (Error Correction Cointegrated Matrix Autoregressive) model. After a brief introduction to the main concepts of cointegration and error correction models, the thesis presents an identification procedure based on Johansen’s trace test, applied to the rows and columns of the data matrix in order to determine the most appropriate model between MECM and ECCMAR. The empirical analysis focuses on the application of these models to Fama–French portfolios observed from the 1990s onward, with the aim of identifying long-run equilibrium relationships and evaluating their use in the implementation of pairs trading strategies.| File | Dimensione | Formato | |
|---|---|---|---|
|
Benvenuti_Enrico.pdf
Accesso riservato
Dimensione
858.9 kB
Formato
Adobe PDF
|
858.9 kB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/112237