PhD Course – Machine Learning in Finance

Jul 13 2019

This course aims at providing an introductory and broad overview of the field of Machine Learning (ML) with the focus on applications on Finance.

Detailed Program:

1.Introduction to Financial problems and their classical solutions

2.Introduction to Machine Learning

Supervised Learning

  – Overview of regression and classification techniques

  – Financial applications: price prediction, modeling bank failures

Unsupervised Learning

  – Overview of clustering and dimensionality reduction techniques

  – Financial applications: stock returns, estimation of equity correlation matrix

Reinforcement Learning

  – Overview of value-based and policy-based techniques

  – Financial applications: option pricing, stock trading

Venue: Department of Mathematics, Politecnico di Milano

Time Table:

Introduction to Financial applications (Baviera, Marazzina, Rroji):

June 13: 9:30-12:00, 14:30-17:00 (Prof. Marazzina)

June 17: 9:30-12:00 (Prof. Baviera), 14:30-17:00 (Prof. Rroji)

June 18: 9:30-12:00 (Prof. Baviera), 14:30-17:00 (Prof. Rroji)

Machine Learning (Restelli, Baviera):

June 20, 21, 25, 27, 28: 10:00-13:00 (Prof. Restelli)    

July 1: 15:00-17:00 (Prof. Baviera)

For information: daniele.marazzina@polimi.it


Seminar: Prof. Enrico Scalas April 16th, 2019

Apr 16 2019

Si avvisa che in data 16/4/2019, alle ore 14:30 , presso Aula seminari del sesto piano, nell’ambito delle iniziative della sezione di Finanza Quantitativa, si svolgerà il seguente seminario:

Titolo: Limit Theorems for the Fractional Non-homogeneous Poisson Process
Relatore: Enrico Scalas, University of Sussex
Abstract:
The fractional non-homogeneous Poisson process was introduced by a time-change of the non-homogeneous Poisson process with the inverse α-stable subordinator. We propose a similar definition for the (non-homogeneous) fractional compound Poisson process. We give both finite-dimensional and functional limit theorems for the fractional non-homogeneous Poisson process and the fractional compound Poisson process. The results are derived by using martingale methods, regular variation properties and Anscombe\’s theorem. Eventually, some of the limit results are verified in a Monte Carlo simulation.

This is a joint work with Nikolai Leonenko and Mailan Trinh.


Seminar: Carlo Sala April 2, 2019

Apr 02 2019

Si avvisa che in data 02/04/2019, alle ore 12:30 precise, presso l’Aula Seminari al Terzo Piano Dipartimento di Matematica del Politecnico di Milano, si svolgerà il seguente seminario:

Titolo: Information content of implicit quantile and implicit expectile curves
Relatore: Carlo Sala, ESADE Business School.

Abstract:

We compare option implied quantiles and option implied expectiles on a 5-year dataset of prices of weekly S&P500 options. We compute these quantities by means of a fully non-parametric methodology, following Barone-Adesi (2016) and Bellini et al. (2018). We study the relative position of inverse quantile and expectile curves, and compute implicit Interquantile and Interexpectile Differences, that are compared with a weekly VIX-like index. Finally, we investigate the forecasting power of these quantities either on future logreturns or on future realized variances. 


Seminar: Alessandro Calvia – 19 March 2019

Mar 19 2019

Si avvisa che in data 19/3/2019, alle ore 12:30 precise, presso Aula Seminari al Sesto Piano, si svolgerà il seguente seminario:

Titolo: Risk measures and progressive enlargement of filtrations: a BSDE approach
Relatore: Alessandro Calvia, Università degli Studi di Milano-Bicocca

Abstract:

Risk measures are nowadays well-established tools in mathematical finance to evaluate the riskiness of a future financial position, both in a static and in a dynamic (i.e., time dependent) setting. Also, Backward Stochastic Differential Equations (or BSDEs, for short) are widely adopted tools in mathematical finance. From the beginning of the 21st century, connections between dynamic risk measures and BSDEs have been studied in the literature. Thanks to the theory of g-expectations, introduced by S. Peng, one can induce a dynamic risk measure from a BSDE. This can be done mapping the terminal condition of the BSDE (modeling a future financial position) into the first component of the corresponding solution. Clearly, this kind of risk measures depend on the noise and on the map g (called driver) appearing in the BSDE. The case of noise given by either a Brownian motion or a Brownian motion and an independent Poisson random measure is studied in the literature. Properties of the driver g, such as monotonicity, convexity, homogeneity, etc., are reflected in the properties of the dynamic risk measure and vice versa. The aim of this talk is to show that it is possible to induce a dynamic risk measure from a BSDE whose driving noise is given by a Brownian motion and a marked point process. In terms of the underlying information flow, this corresponds to the progressive enlargement of a Brownian reference filtration with the information brought by the occurrence of random events at some random times. We will prove that the original risk measure can be split into different risk measures, evaluating future financial position between each of these random events, that are induced, in turn, by a family of indexed Brownian BSDEs. We also show connections between properties of the driver of the BSDE and the induced risk measure and its time-consistency. This is joint work with Emanuela Rosazza Gianin.


Fintech Round Table – March 14, 2019

Mar 14 2019

Il tempo del FINTECH è adesso : la Digital Transformation ha rivoluzionando in modo radicale tutti i livelli degli attuali modelli di business nel settore finanziario; per poter rispondere in modo proattivo alla Fintech Revolution , il MIP Politecnico di Milano ha sviluppato il nuovo Master internazionale in FINTECH – Finance and Digital Innovation .

In occasione del lancio del nuovo master, siamo lieti di invitarvi alla Roundtable che si terrà giovedì 14 marzo alle ore 18.00 presso il Campus MIP.

Durante l’evento, il Direttore del Master,Prof. Emilio Barucci, le aziende sponsor e un Head Hunter di Aegis Human Consulting Group, presenteranno le nuove frontiere del Fintech e le opportunità di carriera, le opportunità e lesfide per giovani professionisti e per le imprese di questo settore.

Agenda

• 18.00: Roundtable

Modera Prof. Emilio Barucci, Direttore Master FINTECH con la partecipazione di Andrea Marchesini, Partner & Director, Aegis UK

– Roberto Villa, Manager of Research ecosystem, IBM Italy

– Savino Damico, Head of Fintech Ecosystem Management and Monitoring Innovation Dept., Intesa Sanpaolo

– Paolo Gianturco, Head of Fintech&FS Tech-EMEA Blockchain Lab co-leader, Deloitte

– Vittorio Giusti, Chief Operating Officer, Aviva Italia

– Andrea Prampolini, Head of Financial markets technology, Banca IMI

– Marco Scappa, Head of Fabrick Corporate Fintech, Fabrick S.p.A

• 19.15: Q&A

• 20.00: Aperitivo

La partecipazione all’evento è gratuita previa registrazione. Al termine dell’evento è previsto un aperitivo di networking.