Paper: arXiv 2511.05030

Authors: Panagiotis G. Papaioannou, Athanassios N. Yannacopoulos

Abstract

We introduce a Geometry Informed Model for financial forecasting by embedding high dimensional market data onto constant curvature 2manifolds. Guided by the uniformization theorem, we model market dynamics as Brownian motion on spherical S2, Euclidean R2, and hyperbolic H2 geometries. We further include the torus T, a compact, flat manifold admissible as a quotient space of the Euclidean plane anticipating its relevance for capturing cyclical dynamics. Manifold learning techniques infer the latent curvature from financial data, revealing the torus as the best performing geometry. We interpret this result through a macroeconomic lens, the torus circular dimensions align with endogenous cycles in output, interest rates, and inflation described by IS LM theory. Our findings demonstrate the value of integrating differential geometry with data-driven inference for financial modeling.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 4.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper presents advanced differential geometry, including explicit SDE derivations for multiple manifolds, demonstrating high mathematical complexity. However, it lacks concrete empirical evidence like backtests, performance metrics, or implementation details, focusing more on theoretical methodology than data-driven validation.

Research Flowchart

  flowchart TD
  A["Research Goal: Market Prediction via Differential Geometry"] --> B["Data Input: High-Dimensional Financial Data"]
  B --> C["Methodology: Embedding onto 2-Manifolds<br>S2 Euclidean R2 Hyperbolic H2 Torus T"]
  C --> D["Computational Process<br>Manifold Learning & Uniformization Theorem"]
  D --> E["Outcome: Curvature Inference<br>Torus T Identified as Optimal Geometry"]
  E --> F["Key Finding: Torus Dynamics<br>Aligns with Cyclical IS-LM Macro Theory"]

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 Geometry-Informed Model that embeds high‑dimensional market data onto constant‑curvature 2‑manifolds for financial forecasting.

Source sentence
We introduce a Geometry-Informed Model for financial forecasting by embedding high-dimensional market data onto constant-curvature 2-manifolds.

Assumes 6

  • Market dynamics follow Brownian motion on the chosen manifolds. model
    Source sentence
    we model market dynamics as Brownian motion on spherical (S2), Euclidean (R2), and hyperbolic (H2) geometries.
  • The torus is a compact, flat manifold admissible as a quotient space of the Euclidean plane for capturing cyclical dynamics. model
    Source sentence
    We further include the torus (T), a compact, flat manifold admissible as a quotient space of the Euclidean plane‑anticipating its relevance for capturing cyclical dynamics.
  • Raw daily prices are taken from Yahoo Finance and transformed into log‑returns. data
    Source sentence
    Raw daily prices (Yahoo Finance) ; log-returns are formed
  • Long‑only (LO) and risk‑parity (RP) benchmark series are constructed for reference. execution
    Source sentence
    basic long-only (LO) and risk-parity (RP) benchmark series are computed for reference.
Show 2 more
  • Eigenportfolios are built using expanding‑window PCA with lagged loadings to avoid look‑ahead bias. data
    Source sentence
    so pi,t is the (out-of-sample) return of the ith eigenportfolio.
  • Simulated regimes are generated as correlated Brownian motions projected to 3D. other
    Source sentence
    Simulated Regimes: CrossManifold Performance

Data

Backtest sample: 2005 – 2025 · multi-asset · daily

  • Yahoo Finance daily price data · Yahoo Finance market-prices public
    broad multi‑asset universe (equities, sectors, rates, credit, commodities, volatility indi
    Source sentence
    Raw daily prices (Yahoo Finance) ; log-returns are formed and basic long-only (LO) and risk-parity (RP) benchmark series are computed for reference.
  • Simulated correlated Brownian motions (CBMs) synthetic unknown
    Source sentence
    To check that geometry performance is not a PCA artifact, we evaluate on equicorrelated CBMs projected to 3D .
  • Simulated regime data for cross‑manifold performance synthetic unknown
    Source sentence
    Table 1 Crossmanifold performance on simulated data

Costs

Unclear - sources disagree The platform classifier and the paper-text extraction give different answers, so no value is asserted.

Platform classifier: Not modeled

Paper text extraction: Not applicable

Finds 5

Reported performance (Geometry-Informed Model (GIM) with toroidal embedding): Sharpe 0.72 (benchmark 0.39)

  • The toroidal geometry yields the highest Sharpe ratio (0.721) on the z‑coordinate, outperforming Euclidean and spherical benchmarks. performance
    Source sentence
    The standout is the toroidal z coordinate (Sharpe 0.721), consistent with a latent cyclic component amplified by a T2 embedding.
  • Integrated Geometry‑Informed Model (GIM) outperforms all baselines, delivering a Sharpe of 0.64 with lower drawdown and turnover. performance
    Source sentence
    Integrated GIM (eigenvalue + curvature) outperforms all baselines.
  • On simulated data, the torus (T2 GIM) achieves the lowest MAE (0.02) and RMSE (0.04) among tested manifolds. predictive-accuracy
    Source sentence
    T2 GIM 0.02 0.04 49.59 0.407, 0.022, 0.020
  • For flat correlated Brownian motions with high correlation (ρ=0.9), Euclidean geometry gives the highest Sharpe (0.211), while toroidal performs worse. performance
    Source sentence
    High correlation ρ = 0.9: – Euclidean: Total Sharpe 0.211 – Spherical: −0.119 Toroidal: −0.172 Hyperbolic: −0.109
  • With moderate correlation (ρ=0.6), toroidal geometry achieves the highest Sharpe (0.273) among the tested manifolds. performance
    Source sentence
    Moderate correlation ρ = 0.6: – Toroidal: Total Sharpe 0.273 – Euclidean: 0.097 Spherical: −0.337 Hyperbolic: −0.310

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 24 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 (72)

  1. Machine Learning for Algorithmic Trading (2023)
  2. Advances in Financial Machine Learning (2018)
  3. Statistical arbitrage in the US equities market (2010) — link
  4. Empirical properties of asset returns: stylized facts and statistical issues (2001) — link
  5. The science of monetary policy: a New Keynesian perspective (1999)
  6. Interest and Prices: Foundations of a Theory of Monetary Policy (2003)
  7. The Uniformization Theorem (1981) — link
  8. Sur l’uniformisation des fonctions analytiques (1907) — link
and 64 more
  1. Machine Learning for Asset Managers (2020)
  2. The Deflated Sharpe Ratio: Correcting for Selection Bias, Backtest Overfitting, and Non-Normality (2014) — link
  3. The Probability of Backtest Overfitting (2014)
  4. Machine Learning and Data Sciences for Financial Markets: A Guide to Contemporary Practices (2023)
  5. Mathematical Methods for Foreign Exchange: A Financial Engineer’s Approach (2001)
  6. Financial Engineering: Selected Works of Alexander Lipton (2018)
  7. Triumph of the Empiricists: The Birth of Financial Data Science (2019)
  8. Enhancing Markowitz’s Portfolio Selection Paradigm with Machine Learning (2024)
  9. FinEAS: Financial Embedding Analysis of Sentiment (2021) — link
  10. Artificial Intelligence in Finance, Volume 2: Reinforcement Learning — Theory and Practice (2025)
  11. Three-Dimensional Geometry and Topology, Vol. 1 (1997)
  12. Differential Geometry of Curves and Surfaces (2016)
  13. Mr. Keynes and the Classics: A Suggested Interpretation (1937)
  14. The entropy formula for the Ricci flow and its geometric applications (2002)
  15. Nonlinear manifold learning for early warnings in financial markets (2017) — link
  16. Nonlinear Manifold Learning for Financial Markets Integration (2018)
  17. Manifold Learning for Financial Market Visualization (2020)
  18. Manifold Learning-Based Phase Space Reconstruction for Financial Time Series (2014)
  19. A kernel entropy manifold learning approach for financial data analysis (2014)
  20. Multivariate Time-Series Analysis via Manifold Learning (2018)
  21. Time Series Forecasting Using Manifold Learning (2022) — link
  22. Stochastic Analysis on Manifolds (2002) — link
  23. Stochastic Analysis on Manifolds (2002) — link
  24. Stochastic Differential Equations and Diffusion Processes (1989)
  25. A geometrical method for the Smoluchowski equation on the sphere (2021) — link
  26. Brownian motion on a sphere: distribution of solid angles (2000)
  27. Hyperbolic and fractional hyperbolic Brownian motion (2007) — link
  28. Spherical and hyperbolic fractional Brownian motion (2005) — link
  29. Intrinsic and extrinsic measurement for Brownian motion (2014)
  30. Brownian motion meets Riemann curvature (2010) — link
  31. Brownian motion parameterized with metric space of constant curvature (1981)
  32. Hydrodynamics in curved membranes: The effect of geometry on particulate mobility (2010) — link
  33. Geometry of financial markets—Towards information theory model of markets (2006)
  34. Gauge geometry of financial markets (1999) — link
  35. Physics of Finance: Gauge Modelling in Non-Equilibrium Pricing (2001)
  36. A differential geometric approach to time series forecasting (2021) — link
  37. The hyperbolic geometry of financial networks (2021) — link
  38. The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin, and Reward (2004)
  39. Geometric Learning and Filtering in Finance (2017)
  40. Network geometry and market instability (2021) — link
  41. A possible interpretation of financial markets affected by dark volatility (2023)
  42. Financial Risk Forecasting with Non-Stationarity (2011)
  43. Riemannian Geometry (1992) — link
  44. Riemannian Geometric Statistics in Medical Image Analysis (2019) — link
  45. Riemannian Center of Mass and Mollifier Smoothing (1977)
  46. Estimating differential quantities using polynomial fitting of osculating jets (2005) — link
  47. Second fundamental measure of geometric sets and local approximation of curvatures (2006)
  48. Random Forests — link
  49. Detecting Strange Attractors in Turbulence (1981) — link
  50. Computational Topology: An Introduction (2010)
  51. Barcodes: The persistent topology of data (2008) — link
  52. Topological data analysis of financial time series: Landscapes of crashes (2018) — link
  53. Early Warning Signals of Financial Crises Using Persistent Homology and Critical Slowing Down (2022) — link
  54. Mild explosivity, persistent homology and cryptocurrencies’ bubbles: An empirical exercise (2024) — link
  55. Principal Component Analysis (2002) — link
  56. Continuous Martingales and Brownian Motion (1999) — link
  57. Principal eigenportfolios for U.S. equities (2022) — link
  58. The Properties of Equally Weighted Risk Parity Portfolios (2010) — link
  59. Introduction to Risk Parity and Budgeting (2013)
  60. Elements of Applied Bifurcation Theory (2013)
  61. The Hopf Bifurcation and the Existence and Stability of Closed Orbits in Multisector Models of Optimal Economic Growth (1979) — link
  62. Non-robust dynamic inferences from macroeconometric models: Bifurcation stratification of confidence regions (2008) — link
  63. Hopf Bifurcations in a Mathematical Model for Economic Growth, Corruption, and Unemployment: Computation of Economic Limit Cycles (2025) — link
  64. Coupled Oscillator Model of the Business Cycle with Fluctuating Goods Markets (2012) — link

0 of 72 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.