Paper: arXiv 2512.20216

Authors: Linuk Perera

Abstract

This research introduces a novel quantitative methodology tailored for quantitative finance applications, enabling banks, stockbrokers, and investors to predict economic regimes and market signals in emerging markets, specifically Sri Lankan stock indices (S&P SL20 and ASPI) by integrating Environmental, Social, and Governance (ESG) sentiment analysis with macroeconomic indicators and advanced time-series forecasting. Designed to leverage quantitative techniques for enhanced risk assessment, portfolio optimization, and trading strategies in volatile environments, the architecture employs FinBERT, a transformer-based NLP model, to extract sentiment from ESG texts, followed by unsupervised clustering (UMAP/HDBSCAN) to identify 5 latent ESG regimes, validated via PCA. These regimes are mapped to economic conditions using a dense neural network and gradient boosting classifier, achieving 84.04% training and 82.0% validation accuracy. Concurrently, time-series models (SRNN, MLP, LSTM, GRU) forecast daily closing prices, with GRU attaining an R-squared of 0.801 and LSTM delivering 52.78% directional accuracy on intraday data. A strong correlation between S&P SL20 and S&P 500, observed through moving average and volatility trend plots, further bolsters forecasting precision. A rule-based fusion logic merges ESG and time-series outputs for final market signals. By addressing literature gaps that overlook emerging markets and holistic integration, this quant-driven framework combines global correlations and local sentiment analysis to offer scalable, accurate tools for quantitative finance professionals navigating complex markets like Sri Lanka.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced mathematics including transformer architectures (FinBERT), manifold learning (UMAP), density-based clustering (HDBSCAN), and validation via PCA, with several mathematical formulations provided, warranting a high complexity score. It presents strong empirical evidence with specific performance metrics (82% validation accuracy, R²=0.801), uses real-world datasets (Bloomberg, CBSL, Yahoo Finance), and includes a pipeline with data preprocessing, model training, and validation, indicating substantial implementation and backtesting readiness.

Research Flowchart

  flowchart TD
  A["Research Goal: Develop Quantitative Model<br>for Sri Lankan Financial Markets"] --> B["Data Collection & Input<br>ESG Texts, S&P SL20/ASPI, S&P 500, Macro Indicators"]
  B --> C1["NLP & Clustering<br>FinBERT + UMAP/HDBSCAN<br>Identifies 5 ESG Regimes"]
  B --> C2["Time-Series Forecasting<br>LSTM/GRU/MLP/SRNN<br>Predicts Prices & Direction"]
  C1 --> D["Data Fusion<br>Rule-based Logic combining<br>ESG Regimes & Forecast Signals"]
  C2 --> D
  D --> E["Final Market Signals<br>For Risk, Portfolio, & Trading"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a novel quantitative methodology that integrates ESG sentiment analysis with macroeconomic indicators and time-series forecasting to predict economic regimes and market signals for Sri Lankan stock indices.

Source sentence
This research introduces a novel quantitative methodology tailored for quantitative finance applications, enabling banks, stockbrokers, and investors to predict economic regimes and market signals in emerging markets, specifically Sri Lankan stock indices (S&P SL20 and ASPI) by integrating Environmental, Social, and Go

Assumes

No explicit assumptions found in the text.

Data

Backtest sample: 2024 – 2025 · Sri Lankan stock indices (S&P SL20, ASPI) · equities · daily

  • ESG-related text data · CBSL and Trading Economics websites text-news public
    Source sentence
    Raw ESG-related text data were sourced from the CBSL the trading economics websites.
  • Macroeconomic ESG indicators · Central Bank of Sri Lanka (CBSL) and Trading Economics websi macro public
    Source sentence
    Macroeconomic ESG indicators included Average Weighted Lending Rate (AWLR), Sri Lanka Exports, Monthly Average Exchange Rates - USD, Broad Money M2b, Currency in Circulation, Tourist Earnings, Unemployment Rate, Industrial Production, and Consumer Price Index (CPI), obtained from the Central Bank of Sri Lanka (CBSL) an
  • Financial stock data for S&P SL20 and ASPI · Bloomberg and Yahoo Finance market-prices commercial
    S&P SL20 and ASPI indices
    Source sentence
    Financial stock data for time-series forecasting were retrieved from Bloomberg and Yahoo Finance, focusing on the S&P SL20 and ASPI indices
  • S&P 500 and select American stocks · Bloomberg and Yahoo Finance market-prices commercial
    S&P 500 constituents
    Source sentence
    with additional inputs from the S&P 500 and select American stocks to account for observed inter-market relationships.

Recognised sources: Bloomberg Yahoo Finance

Costs

Unclear The classifier was not confident.

Finds 4

  • The dense neural network classifier achieved approximately 84.04% training accuracy and 82% validation accuracy in mapping ESG embeddings to five latent ESG regimes. performance
    Source sentence
    The DNN achieved a training accuracy of approximately 84.04% and a validation accuracy of 82%.
  • The LSTM model achieved a directional accuracy of 52.78% on intraday data. predictive-accuracy
    Source sentence
    LSTM delivering 52.78% directional accuracy on intraday data.
  • The fused framework produced approximately 25–30% fewer trade signals compared to baseline models, reducing over‑trading. performance
    Source sentence
    This approach yields a conservative and interpretable trading framework that produces approximately 25–30% fewer trade signals compared to standalone baseline models
  • Preliminary backtests showed a hit rate exceeding 60–70%, indicating consistent directional accuracy with moderate trading activity. performance
    Source sentence
    the model demonstrated a hit rate exceeding 60–70%, indicating consistent directional accuracy while maintaining moderate trading activity.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 30 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.