Skip to content
Curriculum / Phase 3 · Career Pathways

Text as Data and NLP for Finance

Study financial text analysis with lexicons, embeddings, classification and validation of NLP measurements.

Why study Text as Data and NLP for Finance?

Filings and earnings calls contain information that a price table cannot capture, but turning language into a measurement introduces ambiguity. This programme examines whether a text feature means what you think it means and whether its financial use survives validation.

Who this is for

For researchers comfortable with Python, statistics and financial research design; familiarity with classification helps.

What you should be able to do

  • Represent financial disclosures and earnings-call text.
  • Compare lexicon, embedding and classification methods.
  • Evaluate annotation agreement, measurement validity and financial signals.

An example of the work this prepares you for

A small financial-text study with documented labels, an interpretable baseline and an analysis of measurement errors.

Prerequisites and suggested preparation

You should be comfortable with the background covered in these programmes. Equivalent knowledge from university study or independent learning is also suitable.

The goal is to use the foundations in context: explain the assumptions, carry out a calculation and interpret the result. If one of those steps feels unfamiliar, review the linked preparation first.

A preview of the learning journey

The lessons build towards these kinds of reasoning and practical work.

Represent financial language

Choose a representation suited to the question and document collection.

Evaluate a text measurement

Check labels, agreement and domain-specific mistakes.

Connect text to financial evidence

Assess a signal with correct timestamps and a suitable benchmark.

Members work through the detailed lessons, derivations, worked examples and solutions inside the learning dashboard.

Explore lesson previews by topic

Open a topic preview to see its introduction, learning objectives and prerequisites. Previous and next links let you explore the lessons in study order. Full lessons are available in the member dashboard; selected samples are free to read in full.

Topic previewText representation2 lessonsTopic previewFinance lexicons and disclosures2 lessonsTopic previewEmbeddings and classification3 lessonsTopic previewLLM annotation and measurement validity3 lessons

Free resources to start learning

Use these alongside the programme or to prepare independently. Start with the introductory material, work through an example, then return to the research papers when the underlying ideas are familiar.

  1. Textbook · Free

    Speech and Language Processing

    Dan Jurafsky and James H. Martin · online draft

    Read the free draft chapters on text classification, representations and language models.

  2. Research paper · Free

    FinBERT: Financial Sentiment Analysis with Pre-trained Language Models

    Dogu Araci

    Study a primary example of adapting language models to financial sentiment.

  3. Textbook · Free

    An Introduction to Statistical Learning

    Gareth James, Daniela Witten, Trevor Hastie, Robert Tibshirani and Jonathan Taylor · Python edition

    Review classification and evaluation before comparing NLP methods.

  4. Video lectures · Free

    Statistical Learning on YouTube

    Stanford · companion lectures

    Use the classification lectures as preparation for text-model evaluation.

Links reviewed 14 October 2026. Free refers to the linked reading, notes or videos; printed books and certificates may cost extra.

Find more papers and learning materials in the SelfMadeQuant Directory.

Where this programme fits

  • Systematic Trading and Quantitative Research · elective programme

Ready to study Text as Data and NLP for Finance? Explore membership for access to the complete lessons.

Explore membership
Phase 3Advanced Derivatives and Exotic OptionsStudy volatility surfaces, local and stochastic volatility, and exotic derivatives beyond foundational option pricing.Phase 3Counterparty Credit Risk and CVAStudy counterparty credit exposure, CVA, DVA, collateral and Monte Carlo exposure simulation.Phase 3Convertible ArbitrageStudy convertible bond valuation, delta hedging, credit overlays and the financing risks of convertible arbitrage.Phase 3Statistical ArbitrageStudy pairs trading, cointegration, factor residuals and the research and execution of relative-value strategies.Phase 3Financial Machine LearningStudy financial machine learning with event labels, finance-aware validation, feature importance and model monitoring.Phase 3Event Studies and Causal Measurement in FinanceStudy event-study design, abnormal returns and robust causal measurement in financial data.Phase 3Empirical Asset Pricing and Factor InvestingStudy empirical asset pricing, factor construction and the costs, capacity and risks of factor investing.PlannedModern C++ for Quantitative FinanceExplore the planned modern C++ programme for quantitative finance, its intended outcomes and free preparation resources.PlannedQuantitative Data EngineeringExplore the planned quantitative data engineering programme, from reliable financial datasets to reproducible research pipelines.PlannedTrading Systems and Low-Latency InfrastructureExplore the planned trading systems programme, including systems preparation, latency measurement and reliable event processing.