Following the collapse of the Bretton Woods system in 1971, currency values began to fluctuate freely, driven by supply and demand dynamics. In recent years, the forex market has emerged as a fertile ground for algorithmic and machine learning-based trading, also due to its high liquidity. Additionally, a growing body of literature highlights how certain exchange rates exhibit similar movements and, individually, tend to follow mean-reverting dynamics. This study explores the exploitation of temporary disruptions in the comovements of exchange rate pairs, as well as the oscillatory behavior of individual rates, through the application of machine learning and statistical techniques. A key aspect of this approach is the careful selection of both exchange rate pairs and individual rates, which is achieved through innovative methodologies tailored to the different trading strategies.
In seguito al crollo degli accordi di Bretton Woods nel 1971, i valori dei tassi di cambio tra le valute hanno iniziato a fluttuare liberamente, guidati dalle dinamiche di domanda e offerta. Negli ultimi anni, il mercato dei tassi di cambio si è affermato come un terreno fertile per il trading algoritmico e per l'utilizzo di tecniche di machine learning anche grazie alla sua elevata liquidità. Inoltre, un crescente corpus di letteratura evidenzia come alcuni tassi di cambio tendano a muoversi similarmente e, individualmente, tendano a seguire dinamiche di mean reversion. Il presente studio si propone di sfruttare temporanee deviazioni nei comovimenti delle coppie di tassi di cambio, nonché il comportamento oscillatorio dei singoli tassi, attraverso l'applicazione di tecniche statistiche e di machine learning. Un aspetto chiave di questo approccio è un'accurata selezione sia delle coppie di tassi di cambio sia dei tassi individuali, ottenuta mediante metodologie innovative adattate alle diverse strategie di trading.
Statistical arbitrage in forex: a study on pairs trading and mean-reverting rates
PICCININI, LUCA
2023/2024
Abstract
Following the collapse of the Bretton Woods system in 1971, currency values began to fluctuate freely, driven by supply and demand dynamics. In recent years, the forex market has emerged as a fertile ground for algorithmic and machine learning-based trading, also due to its high liquidity. Additionally, a growing body of literature highlights how certain exchange rates exhibit similar movements and, individually, tend to follow mean-reverting dynamics. This study explores the exploitation of temporary disruptions in the comovements of exchange rate pairs, as well as the oscillatory behavior of individual rates, through the application of machine learning and statistical techniques. A key aspect of this approach is the careful selection of both exchange rate pairs and individual rates, which is achieved through innovative methodologies tailored to the different trading strategies.File | Dimensione | Formato | |
---|---|---|---|
Article_Format_Thesis_LucaPiccinini.pdf
accessibile in internet per tutti a partire dal 12/03/2026
Dimensione
3.82 MB
Formato
Adobe PDF
|
3.82 MB | Adobe PDF | Visualizza/Apri |
Executive_Summary_LucaPiccinini.pdf
accessibile in internet per tutti
Dimensione
1.15 MB
Formato
Adobe PDF
|
1.15 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/236340