• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Textual Analysis of News as Applied to Corporate Default Prediction

Student: Ishutin Sergey

Supervisor: Elena Fedorova

Faculty: Faculty of Economic Sciences

Educational Programme: Economics (Bachelor)

Final Grade: 9

Year of Graduation: 2019

This study encompasses four models of corporate default prediction: logit regression, multilayer perceptron, support-vector machine, and random forest. Its main purpose is to determine whether textual analysis of news can be successfully applied to bankruptcy prediction models so as to improve their accuracy. The predictor set consisted of 17 financial, 4 market, and 3 macroeconomic variables. In addition, a number of sets of sentiment variables derived from news headliners database provided by Thomson Reuters was introduced. Word lists of Harvard GI, NRC, and Loughran & McDonald’s were used. All four models are tested against each other, with and without textual factors, on different prediction horizons. The sample of bankrupt firms was provided by UCLA-LoPucki BRD. It consists of 137 large-sized public US firms that filed bankruptcy between 2011 and 2018 inclusively, to which financially stable counterparts randomly drawn from NYSE and NASDAQ listings were added. Study results confirmed the raised hypotheses only partially. L&M dictionary was found to be the most effective, while the most accurate model of prediction was the MLP. The accuracy scores of 90-95% achieved on one-year horizon were only slightly and occasionally improved after introduction of sentiment variables. Keywords: bankruptcy prediction, textual analysis, sentiment analysis, logit, neural network, support-vector machine, random forest

Full text (added May 7, 2019)

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses