Improving stock price prediction using supply chain and director networks

Are you in to disrupt the global investment community the same way as e-Commerce changed the shopping and Uber the transportation experience? Then read on. The EIT Digital Doctoral is offering you an industrial doctorate position at Actelligent and the University of Edinburgh to set a new investment process paradigm in adopting a Man + Machine strategy.

When making investment decisions, investors now need to look into large amounts of unstructured data. They go through hard data as stock prices, trade volumes, company financials, and they ready research reports, news articles and review commentaries from various sources.  With machine learning and big data analytics, the opportunity exists to come with sentiment analysis to save time for investors to take informed decisions.

Challenge

Data is the lifeblood of the “third wave of investing” in the finance world. Actelligent’s view is that a new investment process paradigm can change asset management the same way e-Commerce changed the shopping and Uber changed the transportation experience.

Actelligent started already to work on generating signals based on company-specific reports, using the so-called BERT model. This project will extend the previous approach by incorporating each company’s related network of companies.

The ultimate aim of this industrial doctorate would be an application that helps investors to look at sentiment indicators in one instance, thus strongly enhancing the efficiency of investors. You will improve the accuracy of stock price forecasts by using the supply chain network between companies, as well as that of the boards of directors using machine learning and big data analytics.

Actelligent believes the project will support the global investment community in adopting a “Man + Machine” strategy in investing. The outcomes of this project are relevant for all the companies in the financial sector that are interested in an accurate stock price prediction. If this project succeeds, it will  the first in the Finance Industry to apply Artificial Intelligence to the supply chain and in turn predict stock prices.

Approach

This PhD project includes mapping the network between corporations based on common directors in companies’ boards using a unique dataset. A second network will be created based on competitor-partner and customer-supplier relationships using the same data and analyse the interplay between the two. The candidate then will create a new dynamic econometric model using a special econometrics approach using time varying over fix matrix.

Machine learning and big data analytics will be used to apply a network analysis to estimate the matrices and then to estimate an econometric model. Estimation techniques will be used to validate the best parameters. Different types of algorithms will be used in this PhD work such as neural network, spatial econometric (conditional autoregressive models), exponential random graph models and other machine learning techniques.

Expected outcome

The expected outcome is a novel dynamic econometric model to predict stock price using the director and supply chain networks, implemented and tested on Actelligent database. The industrial doctorate candidate will also come with a proof of concept for commercial use by Actelligent of product and services that predict stock prices using sentiment signals using machine learning and big data analytics.

From a modelling point of view, this project will propose a new econometric approach that includes the two networks to predict the stock price. In this way the dependence between stock prices of different companies can be captured. As the proposal removes the unrealistic assumption of independence between stock prices from different companies at a given time period, the stock price prediction can be improved taking into account contagion effects.

The research will, furthermore, lead to three papers on stock price prediction and a high-quality PhD thesis.

Location

The doctoral student involved in this programme will share its time between the Co-Location Centre of the EIT Digital Edinburgh satellite Node, the premises Actelligent and a mobility programme involving with the Wharton School of the University of Pennsylvania the Bocconi University in Milano, Italy.  In addition to the industrial doctorate research, the PhD student will also be following the EIT Digital Doctoral School leadership seminars.

Facts

Apply

If you are interested in applying to this position, please follow this two-step process: 

  1. Send an e-mail to dsl.office@eitdigital.eu, including a CV, a motivation letter, and documents showing their academic track records, expressing your interest in this PhD position
  2. Apply on the relevant University system at this link.

© 2010-2020 EIT Digital IVZW. All rights reserved. Legal notice