Narrative Digital Finance
A tale of structural breaks, bubbles & market narratives | SNSF Research Project
This project develops a comprehensive framework using advanced ML and NLP techniques to predict market outcomes, detect asset price bubbles, and identify structural breaks in financial markets.
GabinTB/PhD-Narrative-Finance
Gabin Taibi's PhD Research on Narrative Dynamics in Financial Markets
Central research hub for the PhD thesis "Modeling Narrative Dynamics for Volatility Regime Detection in Financial Markets" - combining transformer-based NLP with high-frequency financial data analysis.
Featured: TOPol: Semantic Polarity Fields (2025)
Taibi, Gomez - Semi-unsupervised framework for multidimensional narrative polarity using transformer embeddings. OSF Repository
SLR on Financial Narratives (2025)
Taibi et al. - AI-enhanced systematic literature review on financial narrative modeling. Under revision at Financial Innovation.
Nanosecond Microstructure (2025)
Taibi, Osterrieder, Schlamp - UFT/HFT classification using Deutsche Borse nanosecond-level data.
HFT Reaction Times (2025)
Osterrieder, Schlamp (Deutsche Borse) - Reaction times to macro-news announcements. SSRN
Multimodal Influence (2024)
Bolesta, Taibi et al. - NLP and Generative AI for financial instrument pricing. SSRN
Macro Narratives & Fed Communication (2025)
Taibi - PCA-based macro index with BERTopic narrative shift detection around 2008 crisis.
A Primer on Narrative Finance (2023)
Osterrieder - Foundation paper on narratives shaping financial markets. SSRN
Background
Large fluctuations, instabilities, trends and uncertainty of financial markets constitute a substantial challenge for asset management companies, pension funds and regulators. Nowadays, most asset management companies and financial institutions follow a systematic trading approach in their investment decisions. However, automated or rules-based trading activities bring certain risks for market participants and the whole financial market. In times of increased market volatility or market sell-offs, investors applying similar trading rules might undertake the same actions, escalating and increasing systemic market risk.
Rationale
Despite advancements in econometric methods, the detection of asset price bubbles and structural breaks remains uncertain and underexplored, particularly in the context of classic financial assets. Recent research underscores the powerful role of narratives in financial decision-making, suggesting that integrating textual analysis could enhance market understanding and prediction accuracy.
Aim 1: Econometric Validation
Validate and refine existing econometric models using real-world financial data
Aim 2: Narrative Analysis
Integrate narrative analysis to understand and predict market behaviors and asset price dynamics
Aim 3: AI/ML Framework
Create a multidimensional AI and ML framework that enhances detection of market anomalies
Methods
The approach involves collecting and processing a wide range of data, including stock prices, macroeconomic indicators, and textual content from the web. We employ text mining and NLP techniques to analyze sentiment, narrative structures, and their impact on market movements. Developing and testing new AI models that combine traditional financial analysis with narrative insights to predict market changes and detect structural breaks.
Marius Jan Klein
Team Member
BFH
Direct outputs from this research project (WP1-WP4: NLP, narratives, bubbles, HFT)
TOPol: Capturing and Explaining Multidimensional Semantic Polarity Fields and Vectors
Taibi, G., Gomez, L. (2025) | arXiv:2510.25069 | OSF
AI-Enhanced Systematic Literature Review on Financial Narratives
Taibi, G. et al. (2025) | Under revision at Financial Innovation
Nanosecond Microstructure: High-Frequency Traders Participation Stylized Facts
Taibi, G., Osterrieder, J., Schlamp, S. (2025) | Working Paper
Reaction Times to Economic News in High-Frequency Trading
Osterrieder, J., Schlamp, S. (2025) | SSRN 5112295
Hypothesizing Multimodal Influence: Assessing the Impact of Textual and Non-Textual Data on Financial Instrument Pricing Using NLP and Generative AI
Bolesta, K., Taibi, G., Mare, C., Osterrieder, J. et al. (2024) | SSRN 4698153
Macro Narrative Index Construction
BERTopic + PCA-based sentiment analysis of central bank communications | In Progress
Complete publication record of all team members via OpenAlex
| Year | Title | Authors | Venue | Cites | Links |
|---|---|---|---|---|---|
| Loading... | |||||
Gabin Taibi - University of Twente / BFH (Dec 2023 - Nov 2027)
Thesis: Modeling Narrative Dynamics for Volatility Regime Detection in Financial Markets
Stage 1
Data Collection & SLR
Completed
Stage 2
Narrative Detection (TOPol)
In Progress
Stage 3
Volatility Estimation
2026
Stage 4
Narrative-Volatility Integration
2026-27
Stage 5
Thesis Synthesis
2027
Research Question
How do forms of narratives influence volatility and regime changes in financial markets?
Key Outputs
TOPol framework, SLR on financial narratives, HFT microstructure analysis, realized-library (Python/C++)
RavenPack
News headlines for sentiment analysis
LSEG (Refinitiv)
Earnings call transcripts for corporate narratives
BIS Gigando
Worldwide central bank speeches
SEC EDGAR
10-K and 10-Q filings for regulatory text
Deutsche Borse
Nanosecond-level Xetra/Eurex trading data
St. Louis FED FRED
Macroeconomic indicators (CPI, GDP, Unemployment)
COST Action CA19130 - Fintech and AI in Finance (2020-2024)
Chair: J. Osterrieder | 49 countries | 270+ researchers | EU H2020 & Horizon Europe funded | Successfully concluded October 2024
Deutsche Borse AG
Dr. Stefan Schlamp, Head of Quantitative Analytics - HFT, Market Microstructure. Joint SSRN publication 2025.
MSCA Digital Finance Network
Coordinator: J. Osterrieder | 4.5M EUR | 13 institutions | 100+ researchers
Humboldt-University Berlin
Financial econometrics, joint publications
American University of Sharjah
Keynote presentation May 2024, blockchain & NFT research
Babes-Bolyai University
Romania - PhD training, joint research
Masaryk University
Czech Republic - Quantitative finance research collaboration
SGH Warsaw
Poland - PhD training, fintech research
Quoniam Asset Management
Germany - PhD internship, quantitative strategies research
SNSF - Narrative Digital Finance (Main)
Grant: IZCOZ0_213370 | CHF 236,118 | August 2023 - August 2026
EU Horizon Europe MSCA
Digital Finance Network | Grant 101119635 | 4.5M EUR
EU COST Action CA19130
Fintech and AI in Finance | Chair: Osterrieder | 49 countries | 2020-2024 (concluded)
BFH Institute for Applied Data Science
Host institution providing research infrastructure and support
9th European Conference on AI in Finance (Planned)
2025 | Location TBD | Post-COST continuation
Final Project Dissemination
August 2026 | Project concludes with final report
8th European COST Conference on AI in Finance
Sep 2024 | Bern, Switzerland | BFH
COST FinAI PhD School 2024
2024 | Treviso, Italy | GenAI, Chatbots, LLMs in Finance
COST FinAI Meets Istanbul
May 2024 | Turkey
COST FinAI Meets Brussels
2024 | Belgium
AUS Keynote
May 2024 | UAE | Data Science in Finance
7th European COST Conference
Sep 2023 | Bern, Switzerland
16th ERCIM WG Conference
Dec 2023 | Berlin, Germany
Website Ultra-Deep Review
Comprehensive content audit and improvements to GitHub Pages site
TOPol Framework Released
Taibi & Gomez publish semi-unsupervised framework for narrative polarity fields on OSF
SLR Under Revision
AI-enhanced systematic literature review paper under revision at Financial Innovation
HFT Stylized Facts Paper
Taibi, Osterrieder, Schlamp complete HFT market microstructure analysis with Deutsche Borse data
HFT Reaction Times Paper
Osterrieder & Schlamp publish on SSRN: Reaction Times to Economic News
Final Year Milestone
Entering final year with 6 project publications; project concludes August 2026
8th COST Conference Bern
Annual AI in Finance conference hosted at BFH
Multimodal Influence Paper
Bolesta, Taibi et al. on NLP and GenAI for financial pricing (SSRN)
Deutsche Borse Collaboration
Partnership with Dr. Stefan Schlamp begins for HFT microstructure research
Project Launch
SNSF funding started - CHF 236,118
Principal Investigator
Prof. Dr. Joerg Osterrieder
Bern University of Applied Sciences
University of Twente
joerg.osterrieder@bfh.ch
Project Website
SNSF Grant
IZCOZ0_213370
SNSF Data Portal
August 2023 - August 2026