Search the portal

Please enter a term

Full record

« Back to home page
TitleThe impact of news narrative on the economy and financial markets
AuthorTilly, Sonja
AbstractThis thesis investigates the impact of news narrative on socio-economic systems across four experiments. Recent years have witnessed a rise in the use of so-called alternative data sources to model and predict dynamics in socio-economic systems. Notably, sources such as newspaper text allow researchers to quantify the elusive concept of narrative, to incorporate text-based features into forecasting frameworks and thus to evaluate the impact of narrative on economic events. The first experiment proposes a new method of incorporating a wide array of sentiment scores from global newspaper articles into macroeconomic forecasts, attempting to forecast industrial production and consumer prices leveraging narrative and sentiment from global newspapers. I model industrial production and consumer prices across a diverse range of economies using an autoregressive framework. The second experiment uses narrative from global newspapers to construct themebased knowledge graphs about world events, demonstrating that features extracted from such graphs improve forecasts of industrial production in three large economies. The third experiment proposes a novel method of including news themes and their associated sentiment into predictions of changes in breakeven inflation rates (BEIR) for eight diverse economies with mature fixed income markets. I utilise five types of machine learning algorithms incorporating narrative-based features for each economy. In the above experiments, models incorporating narrative-based features generally outperform their benchmarks that do not contain such variables, demonstrating the predictive power of features derived from news narrative. The fourth experiment utilises GDELT data and the filtering methodology introduced in the first experiment to create a profitable systematic trading strategy based on the average tone scores for 15 diverse economies.
Date2022-05-28
TypeThesis; Doctoral
PublisherUCL (University College London)
Languageeng
Source Doctoral thesis, UCL (University College London).
Formattext
Identifier
Identifier
Rightsopen