MiMIC: Multi-Modal Indian Earnings Calls Dataset to Predict Stock Prices
Paper introduces MiMIC, a multi-modal dataset and cascaded framework combining text, image, and numeric data for predicting Indian stock prices.
What it examines
The paper introduces a multi-modal approach using earnings call transcripts, presentation images, and numerical financial data to predict next-day stock prices. It presents the new MiMIC dataset and a cascaded machine learning framework that integrates various data sources to better capture market reactions in the Indian financial landscape.
What it concludes
The study shows that integrating text and visuals improves stock price predictions. These findings can help develop better trading strategies and financial forecasting tools. Future work may include using audio data and predicting intra-call price movements, enhancing decision support for investors and automated trading systems.
Evidence objects
Researchers unveil a novel multi-modal predictive framework combining textual earnings calls, visual presentation slides, and numerical indicators that significantly improve stock price forecasting accuracy over traditional financial analysis methods remarkably.
key_findings bullet 1 · key_findings · validation V0
The study introduces the pioneering MiMIC dataset for the Indian market, enabling integrated cascaded models using text and image classifiers within a regression framework that substantially reduces prediction errors effectively.
key_findings bullet 2 · key_findings · validation V0
Despite innovative techniques, research highlights limitations including reliance on smaller language models and omission of audio analysis, the balanced methodology and extensive AutoML experiments provide surprising insights into market behavior.
key_findings bullet 3 · key_findings · validation V0
Presenting a pioneering multi-modal dataset, this paper focuses on Indian earnings calls, diverging from US-centric literature. By integrating textual, visual, and tabular data to predict stock prices, it introduces innovative fusion techniques that advance AI-based forecasting, making the work both original and compelling, with significant implications for global finance research.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
Abstract: Predicting stock market prices following corporate earnings calls remains a significant challenge for investors and researchers alike, requiring innovative approaches that can process diverse information sources. This study investigates the impact of corporate earnings calls on stock prices by introducing a multi-modal predictive model. We leverage textual data from earnings call transcripts, alon… ▽ More Predicting stock market prices following corporate earnings calls remains a significant challenge for investors and researchers alike, requiring innovative approaches that can process diverse information sources. This study investigates the impact of corporate earnings calls on stock prices by introducing a multi-modal predictive model. We leverage textual data from earnings call transcripts, along with images and tables from accompanying presentations, to forecast stock price movements on the trading day immediately following these calls. To facilitate this research, we developed the MiMIC (Multi-Modal Indian Earnings Calls) dataset, encompassing companies representing the Nifty 50, Nifty MidCap 50, and Nifty Small 50 indices. The dataset includes earnings call transcripts, presentations, fundamentals, technical indicators, and subsequent stock prices. We present a multimodal analytical framework that integrates quantitative variables with predictive signals derived from textual and visual modalities, thereby enabling a holistic approach to feature representation and analysis. This multi-modal approach demonstrates the potential for integrating diverse information sources to enhance financial forecasting accuracy. To promote further research in computational economics, we have made the MiMIC dataset publicly available under the CC-NC-SA-4.0 licence. Our work contributes to the growing body of literature on market reactions to corporate communications and highlights the efficacy of multi-modal machine learning techniques in financial analysis. △ Less
Source row: 1339 · abstract type: unknown