Paper: arXiv 2601.14062

Authors: Payel Sadhukhan, Samrat Gupta, Subhasis Ghosh, Tanujit Chakraborty

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.

Complexity vs Empirical Score

  • Math Complexity: 5.0/10
  • Empirical Rigor: 5.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs standard ML techniques and Shapley explainability (moderate math), and conducts empirical evaluation on multi-year datasets across two markets with backtest-ready classification metrics.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Can historical OHLC data predict<br>healthcare index movement?"] --> B["Methodology:<br>Supervised Classification<br>Rolling Window Approach"]
  B --> C["Input Data:<br>US & India Healthcare Indices<br>5 Years (2018-2023)"]
  C --> D["Feature Engineering:<br>Original Prices, Volatility Indicators<br>Nowcasting OHLC Ratios"]
  D --> E["Modeling & Explainability:<br>Machine Learning Model<br>Shapley Value Analysis"]
  E --> F["Outcome 1:<br>High Predictive Power<br>Accuracy > 0.8, MCC > 0.6"]
  E --> G["Outcome 2:<br>Feature Dominance<br>Nowcasting ratios are top drivers"]

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 supervised classification model that uses original OHLC prices, volatility indicators, and novel nowcasting features derived from mutual OHLC ratios to predict the next day's opening index.

Source sentence
... 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.

Assumes 5

  • Class labels are defined by comparing the current day's index values with the open index of the next trading day. data
    Source sentence
    Day t’s class label is derived by comparing its index (open/ high/ low/ close) with the open index of the next trading day (t + 1).
  • The first 19 days are excluded because historical feature calculation requires a 20‑day window. data
    Source sentence
    The calculation of historical features requires a window of 20 days, out of which 19 days are before the day of calculation. Hence, we have to exclude the initial 19 days from our calculation.
  • A rolling one‑step‑ahead prediction framework is used rather than multi‑step forecasting. model
    Source sentence
    We have incorporated one-step-ahead rolling window approach for prediction instead of multi-step-ahead forecasting
  • The first 80 % of chronological observations are used for training and the remaining 20 % for testing. data
    Source sentence
    The first 80% days are considered for training the model, and the remaining 20% are used to test the model.
Show 1 more
  • The feature set remains the same across all four sub‑problems; only the target class changes. model
    Source sentence
    Note that, depending on ∗, the dataset changes. The change happens due to the change in the target class only; the feature set remains the same across all cases.

Data

  • Healthcare index – S&P 500 (U.S.A.) · public OHLC data market-prices public
    Apr 1, 2019 – Mar 31, 2024 · Healthcare sector index
    Source sentence
    Period Apr 1, 2019 – Mar 31, 2024 ... No. of trading days 1256

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The models consistently achieved MCC scores greater than 0.6 and accuracy scores greater than 0.8 on held‑out test data. predictive-accuracy
    Source sentence
    The models consistently achieved MCC scores greater than 0.6 and accuracy scores greater than 0.8 on the held-out test data
  • Nowcasting features substantially improve prediction performance, delivering the highest accuracy and MCC across three of four sub‑problems. performance
    Source sentence
    The results show that the addition of nowcasting features has elevated the correctness of predictions. For three of the four sub-problems, it has achieved substantial improvements; both accuracy and MCC scores have
  • Volatility‑based technical indicators (Donchian Channel, Bollinger Bands, Keltner Channels) show inconsistent or negligible correlation with the rise/fall of the open index. negative
    Source sentence
    ... have failed to show consistent correlation with the rise/ fall of indices.
  • Raw original price features provide moderate but inconsistent predictive utility compared to nowcasting features. performance
    Source sentence
    In contrast, historical price movement exhibits negligible predictive association with the directional movement of the opening index, while raw index values offer moderate but inconsistent predictive utility.
  • Shapley‑based explainability confirms that nowcasting features are the dominant contributors, followed by original prices. statistical
    Source sentence
    ... outcomes reveal the dominant role of the nowcasting features, followed by a more moderate contribution of original prices.

Checked by others

Qualified by

  • Risk-Adjusted Performance of Random Forest Models in High-Frequency Trading (2024)
    On the question: Do price‑based features provide predictive utility for the target variable?
    Compare the findings

    This paper: Raw original price features provide moderate but inconsistent predictive utility compared to nowcasting features.

    That paper: Primary price‑based features account for over 60% of feature importance, while technical indicators contribute only 14%–15%.

    A finds price features account for >60% importance, B finds raw price offers moderate but inconsistent utility versus nowcasting features.

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.

Cited References (86)

  1. Financial market reaction to R&D volatility in the pharmaceutical industry. A multi-country study (2022)
  2. Learning scikit-learn: machine learning in python (2013)
  3. Modelling, forecasting and trading with a new sliding window approach: the crack spread example (2016)
  4. The economic value of predicting stock index returns and volatility (2004)
  5. The relationship between trading volume, volatility and returns of non-fungible tokens: evidence from a quantile approach (2022)
  6. Stimulating pharmaceutical research and development for neglected diseases (2003)
  7. How a nation’s well-being influences its health profile: an analysis of critical indicators (2025)
  8. Dividend policy issues in the European pharmaceutical industry: new empirical evidence (2023)
and 78 more
  1. Impact of COVID-19 on the neglected tropical diseases: a scoping review (2024)
  2. Multivariate volatility forecasts for stock market indices (2021)
  3. When Buffett meets Bollinger: An integrated approach to fundamental and technical analysis (2024)
  4. The impact of AI-driven algorithmic trading on market efficiency and volatility: Evidence from global financial markets (2024)
  5. Pharmaceuticals and society: Power, promises and prospects (2015)
  6. Emerging stock market volatility and economic fundamentals: The importance of US uncertainty spillovers, financial and health crises (2022)
  7. The relationship between stock returns and volatility in international stock markets (2005)
  8. The behavior of stock-market prices (1965)
  9. Herd behaviour experimental testing in laboratory artificial stock market settings. Behavioural foundations of stylised facts of financial returns (2013)
  10. Investor sentiment and stock returns: Some international evidence (2009)
  11. Investor sentiment in the stock market (2007)
  12. A decision support system for stock investment recommendations using collective wisdom (2014)
  13. Systematic analysis and review of stock market prediction techniques (2019)
  14. Stock returns and real activity: A century of evidence (1990)
  15. Corporate financial distress prediction in a transition economy (2024)
  16. Forecasting CPI inflation under economic policy and geopolitical uncertainties (2025)
  17. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation (2020)
  18. Relationships between stock returns and corporate financial ratios based on a statistical analysis of corporate data from the Hong Kong stock market (2016)
  19. Stochastic dominance statistics for risk averters and risk seekers: An analysis of stock preferences for USA and China (2015)
  20. Statistical properties of stock order books: empirical results and models (2002)
  21. Investor sentiment and the cross-section of stock returns (2006)
  22. Stock market efficiency and economic efficiency: is there a connection? (1997)
  23. DETERMINANTS OF BANK PROFITABILITY IN A DEVELOPING ECONOMY: EMPIRICAL EVIDENCE FROM THE PHILIPPINES. (2008)
  24. Stock market prediction using artificial intelligence: A systematic review of systematic reviews (2024)
  25. A decision support system for stock investment recommendations using collective wisdom (2014)
  26. Investor psychology and security market under-and overreactions (1998)
  27. Strategic insider trading: Disguising order flows to escape trading competition (2021)
  28. Extremely randomized trees (2006)
  29. CatBoost: unbiased boosting with categorical features (2018)
  30. A unified approach to interpreting model predictions (2017)
  31. COVID-19 pandemic impact on the national health care surveys (2021)
  32. Health as a cornerstone of good business and sustainable development (2016)
  33. Macro-economic factors influencing the architectural business model shift in the pharmaceutical industry (2016)
  34. Economics of the pharmaceutical industry (2018)
  35. Multimodal market information fusion for stock price trend prediction in the pharmaceutical sector (2025)
  36. Forecasting multistep daily stock prices for long-term investment decisions: A study of deep learning models on global indices (2024)
  37. Activity of informed traders and stock returns (2024)
  38. Informed trading and portfolio returns (2013)
  39. Stock market one-day ahead movement prediction using disparate data sources (2017)
  40. Winners and losers from Pfizer and Biontech’s vaccine announcement: Evidence from S&P 500 (Sub) sector indices (2022)
  41. Analysis of cyclical behavior in time series of stock market returns (2018)
  42. Comparative analysis of volatility forecasting for healthcare stock indices amid public health crises: a study based on the Bayes-CNN model (2024)
  43. Cyclicality and sectoral linkages: Aggregate fluctuations from independent sectoral shocks (1998)
  44. High-growth firms in changing competitive environments: the US pharmaceutical industry (1963 to 2002) (2015)
  45. US biopharmaceutical companies’ stock market reaction to the COVID-19 pandemic. Understanding the concept of the ‘paradoxical spiral’from a sustainability perspective (2022)
  46. Stock markets, banks, and economic growth: Evidence from more homogeneous panels (2018)
  47. A meta-analysis of overfitting in machine learning (2019)
  48. Addressing overfitting and underfitting in Gaussian model-based clustering (2018)
  49. Aggregate implications of changing sectoral trends (2022)
  50. COVID-19 pandemic and stability of stock market—A sectoral approach (2021)
  51. Systematic analysis and review of stock market prediction techniques (2019)
  52. Analysis of financial time series (2005)
  53. Fusion in stock market prediction: A decade survey on the necessity, recent developments, and potential future directions (2021)
  54. Using bollinger bands (1992)
  55. How to make money in commodities (1960)
  56. Trend-Following Methods in Commodity Price Analysis (1957)
  57. Intraday periodicity and volatility persistence in financial markets (1997)
  58. Sector analysis for a FTSE portfolio of stocks (2007)
  59. Asymmetric information and liquidity constraints: a new test (2007)
  60. Are the Biotechnology and Pharmaceutical Sectors Defensive Relative to the S&P 500? (2013)
  61. Machine learning (2021)
  62. A graph theoretic approach to assess quality of data for classification task (2025)
  63. Are the Biotechnology and Pharmaceutical Sectors Defensive Relative to the S&P 500? (2013)
  64. Application of neural networks to an emerging financial market: forecasting and trading the Taiwan Stock Index (2003)
  65. Asymmetric volatility connectedness among US stock sectors (2021)
  66. Investing in the healthcare sector: mutual funds or ETFs (2018)
  67. Volatility spillovers across sectors and their magnitude: A sector-based analysis for Australia (2023)
  68. Shock and volatility spillovers among equity sectors of the Gulf Arab stock markets (2009)
  69. Predicting Stock Prices Based on Machine Learning to Build Self-adaptive Trading Strategy (2025)
  70. Deep-learning-based stock market prediction incorporating ESG sentiment and technical indicators (2024)
  71. Unlocking the power of voice for financial risk prediction: A theory-driven deep learning design approach (2023)
  72. A tensor-based information framework for predicting the stock market (2016)
  73. Automated trading systems statistical and machine learning methods and hardware implementation: a survey (2019)
  74. Forecasting benchmarks of long-term stock returns via machine learning (2021)
  75. The two-stage machine learning ensemble models for stock price prediction by combining mode decomposition, extreme learning machine and improved harmony search algorithm (2022)
  76. Decision support from financial disclosures with deep neural networks and transfer learning (2017)
  77. Outperformance of the pharmaceutical sector during the COVID-19 pandemic: Global time-varying screening rule development (2022)
  78. Stocks opening price gaps and adjustments to new information (2024)

0 of 86 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.