Paper: arXiv 2506.19856

Authors: Ryan Samson, Adrian Banner, Luca Candelori, Sebastien Cottrell, Tiziana Di Matteo, Paul Duchnowski, Vahagn Kirakosyan, Jose Marques, Kharen Musaelian, Stefano Pasquali, Ryan Stever, Dario Villani

Abstract

We introduce a novel proxy for firm linkages, Characteristic Vector Linkages (CVLs). We use this concept to estimate firm linkages, first through Euclidean similarity, and then by applying Quantum Cognition Machine Learning (QCML) to similarity learning. We demonstrate that both methods can be used to construct profitable momentum spillover trading strategies, but QCML similarity outperforms the simpler Euclidean similarity.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10 (capped from 7.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 7.0)

Why this score: The paper employs advanced mathematics, including quantum cognition machine learning with Hilbert spaces and Hermitian operators, scoring high on complexity. It demonstrates backtest-ready empirical results on constructing profitable momentum spillover strategies, with detailed methodology and performance comparison, indicating strong empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Estimate Firm Linkages"] --> B["Data: Firm Characteristics"]
  B --> C["Methodology 1:<br>Euclidean Similarity"]
  B --> D["Methodology 2:<br>QCML Similarity Learning"]
  C --> E["Trading Strategy:<br>Momentum Spillover"]
  D --> E
  E --> F["Outcome:<br>Profitable Strategies"]
  F --> G["Key Finding:<br>QCML Outperforms Euclidean"]

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 new proxy for firm linkages called Characteristic Vector Linkages (CVLs) and applies Quantum Cognition Machine Learning (QCML) to improve similarity-based momentum spillover trading.

Source sentence
We introduce a novel proxy for firm linkages, Characteristic Vector Linkages (CVLs).

Assumes 5

  • All input characteristics are demeaned by GICS Industry Code to avoid industry bias. data
    Source sentence
    All input characteristics are demeaned by GICS Industry Code to avoid characteristics linking firms based on industry biases.
  • Accounting values are averaged over the most recent 4 quarters. data
    Source sentence
    Accounting values are averaged over the most recent 4 quarters, unless otherwise stated.
  • Transaction costs are not removed; results are not meant as a stand‑alone strategy. execution
    Source sentence
    We do not remove transaction costs as these forecasts are not intended to represent a stand-alone investment strategy.
  • QCML model parameters are kept static after initial training for simplicity. model
    Source sentence
    for simplicity we keep the parameters static after initial training.
Show 1 more
  • Portfolio returns are smoothed over 21 days and evaluated over Jan 2014–Jun 2024. market
    Source sentence
    We compute daily returns to the smoothed portfolios, and analyze portfolio performance from January 2014 through June 2024.

Data

Backtest sample: 2014-01 – 2024-06 · US equity markets · equities · daily

  • Bloomberg market data · Bloomberg market-prices commercial
    Source sentence
    We use market data sourced from Bloomberg and accounting data sourced from S&P Capital IQ:
  • S&P Capital IQ accounting data · S&P Capital IQ fundamentals commercial
    Source sentence
    We use market data sourced from Bloomberg and accounting data sourced from S&P Capital IQ:

Recognised sources: Bloomberg S&P Capital IQ

Costs

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

Source sentence
We do not remove transaction costs as these forecasts are not intended to represent a stand-alone investment strategy.

Finds 4

Reported performance (QCML similarity CVL momentum spillover strategy): Sharpe 1.42

  • QCML 252‑day input signal achieves a Sharpe of 1.10 versus 0.73 for Euclidean, a >50% improvement. performance
    Source sentence
    the 252-day input return QCML signal shows a Sharpe of 1.10 compared to 0.73 for the Euclidean signal (a greater than 50% improvement in Sharpe)
  • QCML combined input signal Sharpe 1.42 exceeds Euclidean combined Sharpe 1.24. performance
    Source sentence
    The QCML combined input return signal gives a Sharpe of 1.42, an improvement over the Euclidean combined input return signal, which has a Sharpe of 1.24
  • QCML half‑life (90.9 days) is 2.5× longer than Euclidean half‑life (36.3 days). performance
    Source sentence
    QCML signal has a half-life (90.9 days) 2.5 times longer than the Euclidean signal (36.3 days)
  • Euclidean similarity yields Sharpe ratios from 0.73 to 1.37 depending on horizon. performance
    Source sentence
    The Euclidean similarity approach, despite its simplicity, generates risk controlled portfolio returns with Sharpe ratios ranging from 0.73 to 1.37 depending on the input horizon

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 28 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 (57)

  1. Shared analyst coverage: Unifying momentum spillover effects (2020) — link
  2. Machine Learning for Active Portfolio Management (2021) — link
  3. The relationship between earnings’ yield, market value and return for NYSE common stocks: Further evidence (1983) — link
  4. Geometry of Quantum States: An Introduction to Quantum Entanglement (2006) — link
  5. Debt/Equity Ratio and Expected Common Stock Returns: Empirical Evidence (1988) — link
  6. Robust estimation of the intrinsic dimension of data sets with quantum cognition machine learning (2025) — link
  7. Economic Links and Predictable Returns (2008) — link
  8. Implied Equity Duration: A New Measure of Equity Risk (2004) — link
and 49 more
  1. Value-Glamour and Accruals Mispricing: One Anomaly or Two? (2004)
  2. A meta-analysis of supervised and unsupervised machine learning algorithms and their application to active portfolio management (2025) — link
  3. Do investors overvalue firms with bloated balance sheets? (2004) — link
  4. Replicating Anomalies (2020) — link
  5. Matrix geometry and coherent states (2015) — link
  6. Is There a Replication Crisis in Finance? (2023) — link
  7. Supervised similarity learning for corporate bonds using Random Forest proximities (2022) — link
  8. Taxable Income, Future Earnings, and Equity Values (2004) — link
  9. Do Industries Explain Momentum? (1999) — link
  10. Quantum Cognition Machine Learning: \uppercase AI Needs Quantum (2024)
  11. On Spectral Clustering: Analysis and an algorithm (2001) — link
  12. Quantum Computation and Quantum Information (2000)
  13. Real Asset Illiquidity and the Cost of Capital (2014)
  14. Cash holdings, risk, and expected returns (2012) — link
  15. Geographic Lead-Lag Effects (2020) — link
  16. The Book-to-Price Effect in Stock Returns: Accounting for Leverage (2007)
  17. Does Corporate Headquarters Location Matter for Stock Returns? (2006) — link
  18. Quantum Cognition (2022) — link
  19. Accrual reliability, earnings persistence and stock prices (2005) — link
  20. Supervised Similarity for High-Yield Bonds (2025) — link
  21. Persuasive evidence of market inefficiency (1985) — link
  22. Quantum Cognition Machine Learning: financial forecasting (2024) — link
  23. Do Stock Prices Fully Reflect Information in Accruals and Cash Flows about Future Earnings? (1996) — link
  24. Geometric Measures of Quantum Correlations with Bures and Hellinger Distances (2017) — link
  25. Corporate Real Estate Holdings and the Cross-Section of Stock Returns (2010) — link
  26. Company Similarity using Large Language Models — → 2308.08031
  27. Introduction to Latent Variable Energy-Based Models: A Path Towards Autonomous Machine Intelligence (2022)
  28. LinkedIn posts (2023-2024)
  29. Quantum Models of Cognition and Decision (2012) — link
  30. Applications of Quantum Statistics in Psychological Studies of Decision Processes (1995) — link
  31. Quantum-like brain: “Interference of minds” (2006) — link
  32. Quantamentals: Refining Short Interest (March 2010)
  33. Signal Processing: The long and the short of it (January 2011)
  34. The Interaction of Value and Momentum Strategies (1997)
  35. Empirical Asset Pricing via Machine Learning (2020) — link
  36. The relationship between return and market value of common stocks (1981) — link
  37. Common risk factors in the returns on stocks and bonds (1993) — link
  38. CAPITAL ASSET PRICES: A THEORY OF MARKET EQUILIBRIUM UNDER CONDITIONS OF RISK (1964) — link
  39. Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency (1993) — link
  40. On Persistence in Mutual Fund Performance (1997) — link
  41. Risk, Return, and Equilibrium: Empirical Tests (1973) — link
  42. New Evidence on the Relation between the Enterprise Multiple and Average Stock Returns (2011) — link
  43. Do Sales–Price and Debt–Equity Explain Stock Returns Better than Book–Market and Firm Size? (1996) — link
  44. The Use of Dupont Analysis by Market Participants (2007) — link
  45. Predicting Stock Returns Using Industry-Relative Firm Characteristics (2000) — link
  46. Low-Risk Investing without Industry Bets (2014) — link
  47. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift (2015) — link
  48. What is the best multi-stage architecture for object recognition? (2009) — link
  49. Rectified Linear Units Improve Restricted Boltzmann Machines (2010) — link

1 of 57 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.