[Paper Review] Demystifying the trend of the healthcare index: Is historical price a key driver?

Payel Sadhukhan, Samrat Gupta+2 more|arXiv (Cornell University)|Jan 20, 2026

Global Health Care Issues0 citations

TL;DR

The paper shows that a feature-rich ML model using OHLC data, volatility indicators, and novel nowcasting features can predict the next day’s healthcare index opening with accuracy above 0.8 and MCC above 0.6 for US and Indian markets, with nowcasting features dominating model explanations via Shapley values.

ABSTRACT

Healthcare sector indices consolidate the economic health of pharmaceutical, biotechnology, and healthcare service firms. The short-term movements in these indices are closely intertwined with capital allocation decisions affecting research and development investment, drug availability, and long-term health outcomes. This research investigates whether historical open-high-low-close (OHLC) index data contain sufficient information for predicting the directional movement of the opening index on the subsequent trading day. The problem is formulated as a supervised classification task involving a one-step-ahead rolling window. A diverse feature set is constructed, comprising original prices, volatility-based technical indicators, and a novel class of nowcasting features derived from mutual OHLC ratios. The framework is evaluated on data from healthcare indices in the U.S. and Indian markets over a five-year period spanning multiple economic phases, including the COVID-19 pandemic. The results demonstrate robust predictive performance, with accuracy exceeding 0.8 and Matthews correlation coefficients above 0.6. Notably, the proposed nowcasting features have emerged as a key determinant of the market movement. We have employed the Shapley-based explainability paradigm to further elucidate the contribution of the features: outcomes reveal the dominant role of the nowcasting features, followed by a more moderate contribution of original prices. This research offers a societal utility: the proposed features and model for short-term forecasting of healthcare indices can reduce information asymmetry and support a more stable and equitable health economy.

Motivation & Objective

• Understand whether short-term healthcare index movements can be predicted from historical OHLC data.
• Identify which feature types (intrinsic, volatility-based, nowcasting) most strongly relate to next-day opening movements.
• Develop an interpretable ML framework with transparency for feature contributions to predictions.
• Assess cross-market generalizability using US and Indian healthcare indices over diverse economic phases.

Proposed method

• Construct a labeled dataset framing a one-step-ahead rolling classification task for rise/fall of the next-day open index.
• Engineer a diverse feature set: intrinsic OHLC features, volatility-based indicators (Donchian Channel, Bollinger Bands, Keltner Channel), and nowcasting features based on mutual OHLC ratios.
• Form a final feature vector by concatenating intrinsic, volatility, and nowcasting features for model input.
• Train multiple classifiers on US and Indian healthcare index data from 2019–2024 with 80/20 train/test split.
• Evaluate performance using accuracy and Matthews Correlation Coefficient (MCC).
• Apply Shapley values to quantify feature contributions and interpret model decisions.

(a) S $\&$ P 500

Experimental results

Research questions

• RQ1 Can we predict the rise or fall of the next day’s healthcare index opening using historical OHLC data?
• RQ2 Which feature families (intrinsic prices, volatility indicators, nowcasting ratios) most strongly drive predictive performance?
• RQ3 Do the models generalize across two distinct markets (US and India) with similar predictive accuracy and explanation patterns?
• RQ4 How does Shapley-based explainability characterize the relative importance of nowcasting features versus original price features?

Key findings

• Prediction accuracy exceeds 0.8 on held-out test data.
• Matthews correlation coefficient (MCC) exceeds 0.6 on held-out test data.
• Nowcasting features derived from mutual OHLC ratios emerge as a key determinant of market movement.
• Shapley-based explanations reveal the dominant role of nowcasting features, followed by a more moderate contribution from original prices.
• The framework uses publicly available OHLC data and offers interpretable insights into drivers of healthcare index movements.

(b) BSE