Paper: arXiv 2512.02036

Authors: Juan C. King, Jose M. Amigo

Abstract

The aim of this paper is the analysis and selection of stock trading systems that combine different models with data of different nature, such as financial and microeconomic information. Specifically, based on previous work by the authors and applying advanced techniques of Machine Learning and Deep Learning, our objective is to formulate trading algorithms for the stock market with empirically tested statistical advantages, thus improving results published in the literature. Our approach integrates Long Short-Term Memory (LSTM) networks with algorithms based on decision trees, such as Random Forest and Gradient Boosting. While the former analyze price patterns of financial assets, the latter are fed with economic data of companies. Numerical simulations of algorithmic trading with data from international companies and 10-weekday predictions confirm that an approach based on both fundamental and technical variables can outperform the usual approaches, which do not combine those two types of variables. In doing so, Random Forest turned out to be the best performer among the decision trees. We also discuss how the prediction performance of such a hybrid approach can be boosted by selecting the technical variables.

Complexity vs Empirical Score

  • Math Complexity: 4.0/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Street Traders — empirical and applied, lower math complexity

Why this score: The paper focuses on applied machine learning (LSTM, Random Forest) without heavy theoretical derivations, but reports detailed backtests with specific metrics, cross-validation, and trading simulations on financial data.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Formulate trading algorithms with<br>statistical advantages using hybrid ML"] --> B["Data Preparation:<br>Financial + Microeconomic Data"]
  B --> C["Model Architecture"]
  C --> D["LSTM Network<br>(Technical Analysis<br>Price Patterns)"]
  C --> E["Decision Tree Ensemble<br>(Fundamental Analysis<br>Economic Data)"]
  E --> F{"Algorithmic Trading Simulation"}
  D --> F
  F --> G["Comparative Analysis<br>(10-Week Prediction Window)"]
  G --> H{"Outcomes & Selection"}
  H --> I["Random Forest<br>Selected as best performer"]
  H --> J["Hybrid Approach<br>Outperforms single-type models"]
  H --> K["Feature Selection<br>Boosts prediction accuracy"]

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 hybrid trading model that feeds LSTM technical predictions into a Random Forest fundamental model.

Source sentence
a hybrid model whose inputs include outputs of technical models through the integration of predictions obtained from LSTM networks into the fundamental models.

Assumes 3

  • The prediction target is a binary variable indicating whether the price will rise or fall in 10 business days. model
    Source sentence
    The prediction that these models are going to make is whether the price of a given asset will rise or fall in 10 business days (two weeks).
  • The investment strategy diversifies capital equally across the 30 assets with the highest AUC. execution
    Source sentence
    diversifying the entire capital across the 30 assets with the highest area under…
  • Fundamental variables capture more than 90% of explanatory variance in equity returns. data
    Source sentence
    This selection captures more than 90% of the explanatory variance in equity returns according to fundamental factor studies

Data

Backtest sample: equities · weekly

  • Fundamental micro‑economic data · anonymous multinational financial data company fundamentals commercial
    2023 (bimonthly) · 482 companies
    Source sentence
    The data consists of bimonthly microeconomic data from 482 companies, including tech giants like Apple, Microsoft and Nvidia; visit [15] for a list of all these companies.
  • Historical price data market-prices unknown
    1971 onward · same 482 companies
    Source sentence
    Price data, on the other hand, were collected from 1971 onwards, providing a long-term historical series calculation of indicators and technical models.

Costs

Not modeled A trading or execution result is reported without costs.

Finds 4

  • The hybrid model achieved a Test AUC of 0.566, a slight increase over the best fundamental model (0.5628). predictive-accuracy
    Source sentence
    Test AUC of the hybrid model is 0.566, an increase of 0.003 with respect to the best fundamental model
  • After filtering to LSTM AUC > 0.6, the hybrid model’s Test AUC rose to 0.73. performance
    Source sentence
    in the last iteration, where only 318 records with AUC greater than 0.6 remain, the Test AUC of the hybrid model is 0.73.
  • The hybrid model outperformed benchmark indices in weekly returns during the evaluated period. performance
    Source sentence
    the system consistently outperforms the benchmark indices during the evaluated dates.
  • Technical models alone yielded an AUC of 0.493 with a wide confidence interval, indicating no statistical significance. negative
    Source sentence
    Technical model yielded an AUC of 0.493 with a wide 95% confidence interval of [-0.296, 1.282], and therefore, no conclusion regarding its statistical significance can be drawn.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 49 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.

Cited References (64)

  1. Evaluating Company Performance: The Role of EBITDA as a Key Financial Metric (2025) — link
  2. Impact of shareholder yield on financial distress: using competition and firm size as moderators (2025) — link
  3. Greedy algorithms: a review and open problems (2024) — link
  4. Investor clientele and intraday patterns in the cross section of stock returns (2024) — link
  5. Evolution of the Legal Framework for UCITS Funds in the European Union (1985–2023) (2024) — link
  6. A prospective randomized double-blind study comparing the dose-response curves of epidural ropivacaine for labor analgesia initiation between parturients with and without obesity (2024) — link
  7. Blockchain Metrics and Indicators in Cryptocurrency Trading — → 2403.00770
  8. Optimal and Non-Optimal MACD Parameter Values and Their Ranges for Stock-Index Futures: A Comparative Study of Nikkei, Dow Jones, and Nasdaq (2023) — link
and 56 more
  1. Gaussian Process-Based Random Search for Continuous Optimization via Simulation (2023) — link
  2. Linear Combination of Exponential Moving Averages for Wireless Channel Prediction (2023) — link
  3. Panel data nowcasting: The case of price–earnings ratios (2023) — link
  4. Automated Stock Picking using Random Forests (2023) — link
  5. A Combined Model Based on Recurrent Neural Networks and Graph Convolutional Networks for Financial Time Series Forecasting (2023) — link
  6. Ichimoku Cloud Forecasting Returns in the U.S. (2022) — link
  7. Stock Index Prediction Based on Time Series Decomposition and Hybrid Model (2022) — link
  8. Algorithmic Trading using Technical Indicators (2021) — link
  9. Surplus Free Cash Flow, Stock Market Segmentations and Earnings Management: The Moderating Role of Independent Audit Committee (2020) — link
  10. Deep Learning for Algorithmic Trading: Enhancing MACD Strategy (2020) — link
  11. Combining Principal Component Analysis, Discrete Wavelet Transform and XGBoost to trade in the financial markets (2019) — link
  12. Short-Term Load Forecast for Energy Management Systems Using Time Series Analysis and Neural Network Method with Average True Range (2019) — link
  13. Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) Network (2018) — link
  14. Integrating Fundamental and Technical Analysis of Stock Market through Multi-layer Perceptron (2018) — link
  15. An innovative neural network approach for stock market prediction (2018) — link
  16. Buy/sell signal detection in stock trading with bollinger bands and parabolic SAR: With web application for proofing trading strategy (2017) — link
  17. Directional prediction of stock prices using breaking news on Twitter (2015) — link
  18. Technical analysis of three stock oscillators testing MACD, RSI and KDJ rules in SH & SZ stock markets (2015) — link
  19. Gradient boosting machines, a tutorial (2013) — link
  20. Algorithmic Trading and the Market for Liquidity (2013) — link
  21. Quantitative Trading: How to Build Your Own Algorithmic Trading Business (2008)
  22. Investment Valuation: Tools and Techniques for Determining the Value of Any Asset (1995)
  23. Integrating artificial neural networks with rule-based expert systems (1994) — link
  24. CORPORATE FORECASTS OF EARNINGS PER SHARE AND STOCK-PRICE BEHAVIOR - EMPIRICAL TESTS (1976) — link
  25. The influence of book value per share (BVS) on Islamic and conventional financial institutions: A bibliometric study using VOSviewer and literature review (2023)
  26. A Data Analytics Architecture for the Exploratory Analysis of High-Frequency Market Data (2022) — link
  27. Equity Risk Premiums: Determinants, Estimation and Implications - The 2020 Edition (2020) — link
  28. The Prevalence and Validity of EBITDA as a Performance Measure (2019) — link
  29. Evidence of technical analysis profitability in the foreign exchange market (2019)
  30. Price-to-earnings ratio: A state-of-art review (2017) — link
  31. The basics of ebitda (2017)
  32. Deep Learning (2016) — link
  33. Technical Indicators as Predictors of Position Outcome for Technical trading (2015)
  34. Optimization of MACD and RSI indicators: An Empirical Study of Indian Equity Market for Profitable Investment Decisions (2015) — link
  35. The relationship between portfolio returns and market multiples: A case study of Pakistan (2015)
  36. Data, computer codes, and list of 482 companies
  37. (l) % Capital contracted daily: Percentage of capital traded daily
  38. h) 3y Price Volatility
  39. c) Models that combine fundamental and technical variables will be referred to as hybrid models
  40. Analysis of the Bollinger Band Mean Regression Trading Strategy
  41. j) Long term growth %: Long-term growth percentage
  42. (k) 3y BETA Rel to Loc Idx: Three-year Beta relative to the local index
  43. i) Issue Common Shares Outstdg
  44. i) Target Price 3Months, var % PO 3Months: 3-month price target and percentage change in the price target
  45. view of the market and can identify significant patterns and trends influencing stock prices
  46. (e) 12 months %
  47. c) Free-float EUR millions: The market capitalization of a company adjusted for the proportion of shares available for public trading, expressed in millions of euros
  48. ) Recommendation and numerical recommendation for the stock
  49. Volume/shares %: Volume per share percentage
  50. (b) Float Pct Total Outstdg: The percentage of total outstanding shares that are available for trading in the open market
  51. Fundamental models,which are machine learning algorithms based on decision trees and fed only with fundamental variables
  52. (g) Sales +1E 3Months, Sales +1E Current, var % Sales +1E 3Months: Next quarter estimated sales, actual sales, and percentage change in estimated sales
  53. (g) % From lows 1 year, % from highs 1 year
  54. Last 52 Weeks Low Price and Last 52 Wks High Price: Lowest and highest prices over the last 52 weeks
  55. (m) Diff % Mean 200, diff % Mean 50, diff % Mean 25, Mean 50/200: the percentage difference from the 200 moving average daily timeframe
  56. M2) Technical models (one per asset), which are LSTM networks trained only with technical variables

1 of 64 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.