Paper: arXiv 2510.23183

Authors: E. Benhamou, JJ. Ohana, B. Guez, E. Setrouk, T. Jacquot

Abstract

In this work, we introduce PEARL (Private Equity Accessibility Reimagined with Liquidity), an AI-powered framework designed to replicate and decode private equity funds using liquid, cost-effective assets. Relying on previous research methods such as Erik Stafford’s single stock selection (Stafford) and Thomson Reuters - Refinitiv’s sector approach (TR), our approach incorporates an additional asymmetry to capture the reduced volatility and better performance of private equity funds resulting from sale timing, leverage, and stock improvements through management changes. As a result, our model exhibits a strong correlation with well-established liquid benchmarks such as Stafford and TR, as well as listed private equity firms (Listed PE), while enhancing performance to better align with renowned quarterly private equity benchmarks like Cambridge Associates, Preqin, and Bloomberg Private Equity Fund indices. Empirical findings validate that our two-step approachdecoding liquid daily private equity proxies with a degree of negative return asymmetry outperforms the initial daily proxies and yields performance more consistent with quarterly private equity benchmarks.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 6.0/10 (capped from 8.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 8.0)

Why this score: The paper employs advanced machine learning and graphical models for replication (high math), while its validation involves correlation with established benchmarks and backtesting against private equity indices (high empirical rigor).

Research Flowchart

  flowchart TD
  A["Research Goal: Develop AI-powered framework<br>to replicate private equity funds<br>with liquid, cost-effective assets"] --> B["Core Methodology: Two-step approach<br>incorporating negative return asymmetry"]
  B --> C["Inputs & Data<br>TR Refinitiv Sector Approach<br>Erik Stafford Single Stock Selection"]
  C --> D{"Computation: Decode Daily<br>Liquid Private Equity Proxies"}
  D --> E["Apply Asymmetry to capture<br>leverage, timing, & management effects"]
  E --> F["Generate Enhanced Daily<br>Private Equity Proxy"]
  F --> G["Key Outcomes & Findings"]
  G --> H["Correlation: Strong alignment<br>with liquid benchmarks (Stafford, TR, Listed PE)"]
  G --> I["Performance: Outperforms initial proxies<br>Aligns with quarterly PE benchmarks<br>(Cambridge, Preqin, Bloomberg)"]

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 PEARL, an AI‑powered framework that replicates and decodes private‑equity fund performance using liquid, cost‑effective assets.

Source sentence
In this work, we introduce PEARL (Private Equity Accessibility Reimagined with Liquidity), an AI-powered framework designed to replicate and decode private equity funds using liquid, cost-effective assets.

Assumes 3

  • Negative daily returns are scaled down to 9% to mimic private‑equity drawdown mitigation. model
    Source sentence
    Specifically, for any negative returns, only 9% of the return is considered. Additionally, we introduce new asymmetrical factors, denoted as AF, to represent this adjustment.
  • Daily liquid proxies (Erik Stafford and Thomson Reuters Refinitiv) can serve as bases for daily replication of private‑equity benchmarks. data
    Source sentence
    daily liquid proxies such as the Erik Stafford (Stafford) single‑stock selection method implemented via the SummerHaven Private Equity Strategy Index (SHPEISM Index), the Thomson Reuters Refinitiv (TR) Private Equity Benchmark provide more frequent quotations.
  • Graphical models can capture the relationship between asset weights and observed NAVs for daily updates. model
    Source sentence
    The methodology begins with constructing a probabilistic graphical model, as illustrated in Figure 1. The overall idea is to capture the relationships between the weights in the various assets and the observed NAVs.

Data

Backtest sample: 2012-01 – 2025-01 · equities · equities · daily

  • Cambridge Associates benchmark benchmark unknown
    31/03/2011 to 29/12/2023
    Source sentence
    Table 1 presents the historical performance of two widely recognized private equity benchmarks: the Cambridge Associates (CA) and Preqin indexes.
  • Preqin benchmark benchmark unknown
    31/03/2011 to 29/12/2023
    Source sentence
    Table 1 presents the historical performance of two widely recognized private equity benchmarks: the Cambridge Associates (CA) and Preqin indexes.
  • Bloomberg PEALL index · Bloomberg benchmark unknown
    31/03/2011 to 29/12/2023
    Source sentence
    return (11.4%) compared to the Cambridge Associates and Preqin indexes but benefits from the lowest volatility (5.9%) and the highest Sharpe ratio (1.95).
  • Bloomberg PEBUY index · Bloomberg benchmark unknown
    31/03/2011 to 29/12/2023
    Source sentence
    ... The Bloomberg PEBUY index, on the other hand, delivers a stronger annual return of 13.2% with a slightly higher volatility of 7.6%.
  • Erik Stafford (Stafford) single‑stock selection index · SummerHaven benchmark unknown
    Source sentence
    daily liquid proxies such as the Erik Stafford (Stafford) single‑stock selection method implemented via the SummerHaven Private Equity Strategy Index (SHPEISM Index)
  • Thomson Reuters Refinitiv (TR) Private Equity Benchmark · Thomson Reuters Refinitiv benchmark unknown
    Source sentence
    ... the Thomson Reuters Refinitiv (TR) Private Equity Benchmark provide more frequent quotations.

Recognised sources: Refinitiv / LSEG (Datastream, Eikon, Tick History) Bloomberg

Costs

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

Finds 3

Reported performance (PEARL): Sharpe 1.27 · annual return 17.7% · max drawdown 17.6%

  • PEARL’s model shows strong correlation with existing liquid benchmarks (Stafford and TR) and listed PE, while improving alignment with quarterly private‑equity benchmarks such as Cambridge Associates, Preqin, and Bloomberg indices. performance
    Source sentence
    our model exhibits a strong correlation with well-established liquid benchmarks such as Stafford and TR, as well as listed private equity firms (Listed PE), while enhancing performance to better align with renowned quarterly private equity benchmarks like Cambridge Associates, Preqin, and Bloomberg Private Equity Fund
  • Applying the asymmetrical transformation (scaling negative returns to 9%) yields performance that outperforms the initial daily proxies and matches quarterly private‑equity benchmark characteristics. performance
    Source sentence
    The transformation ensures that positive returns remain unchanged, while negative returns are scaled down to 9%, reflecting the drawdown adjustment characteristic of private equity valuations.
  • Quarterly private‑equity benchmarks exhibit high Sharpe ratios (1.5–2.0) and low volatility compared to daily liquid proxies, underscoring the challenge of daily replication. statistical
    Source sentence
    These benchmarks exhibit significantly higher Sharpe ratios, typically around 1.5, far above the traditional Sharpe ratio of approximately 0.5 observed in major large equity indexes.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 15 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive

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. Private equity vs. PLC boards in the UK: A comparison of practices and effectiveness (2009)
  2. Portfolio Choice with Illiquid Assets (2014) — link
  3. Estimating Private Equity Returns from Limited Partner Cash Flows (2014)
  4. Quality Minus Junk (2014)
  5. Borrow Cheap, Buy High? The Determinants of Leverage and Pricing in Buyouts (2013)
  6. Global Asset Management 2020 (2020) — link
  7. Five Concerns with the Five-Factor Model (2018) — link
  8. Valuescope, Inc. (2020)
and 64 more
  1. The Systematic Risk of Private Equity (2014) — link
  2. Equilibrium liquidity premia of private equity funds (2016) — link
  3. On Persistence in Mutual Fund Performance (1997)
  4. Private Equity Demystified (2020) — link
  5. The Growing Private Equity Market (2020) — link
  6. Evaluating investments in unlisted equity for the norwegian government pension fund global (gpfg) (2018)
  7. Capital and Liquidity Solutions for Private Equity Funds (2020) — link
  8. IRR vs. TVPI (2020) — link
  9. Common Risk Factors in the Returns on Stocks and Bonds (1993)
  10. A Five-Factor Asset Pricing Model (2015)
  11. International Tests of a Five-Factor Asset Pricing Model (2017)
  12. Private Equity Performance and Liquidity Risk (2012) — link
  13. A Test of the Efficiency of a Given Portfolio (1989)
  14. Benchmarking Private Equity: The Direct Alpha Method (2014) — link
  15. 5q (2018)
  16. Which Factors? (2019)
  17. Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers (1986)
  18. Private Equity Performance: Returns, Persistence, and Capital Flows (2005) — link
  19. Leveraged Buyouts and Private Equity (2009)
  20. How Do Private Equity Investments Perform Compared to Public Equity? (2015)
  21. Why Incentive Plans Cannot Work (1993) — link
  22. Risk and Expected Returns of Private Equity Investments: Evidence Based on Market Prices (2014) — link
  23. Organizational Forms and Risk of Listed Private Equity (2009)
  24. A Private Investment Benchmark (1996) — link
  25. Lessons from Private Equity Any Company Can Use (2008) — link
  26. Portfolio Selection (1952)
  27. A Theory of Liquidity in Private Equity (2023) — link
  28. McKinsey’s Private Markets Annual Review (2020) — link
  29. McKinsey’s Private Markets Annual Review 2023 (2023) — link
  30. The liquidity cost of private equity investments: Evidence from secondary market transactions (2019) — link
  31. Liquidity Risk and Expected Stock Returns (2003)
  32. What is Private Debt? (2021) — link
  33. Private Equity Performance Data Guide (2017) — link
  34. The Definition of Private Capital (2021) — link
  35. Data Solutions You Can Trust (2021) — link
  36. Private Equity & Venture Capital (2021) — link
  37. Leveraged Small Value Equities (2015) — link
  38. Thomson Reuters Private Equity Buyout Index (2021) — link
  39. Funding Liquidity, Political Geography, and Private Equity Performance: Evidence from China (2019) — link
  40. The Voice of Experience: Public Versus Private Equity (2008) — link
  41. Cyclicality, performance measurement, and cash flow liquidity in private equity (2016) — link
  42. Cyclicality, Performance Measurement, and Cash Flow Liquidity in Private Equity (2016)
  43. Replicating Private Equity with Value Investing, Homemade Leverage, and Hold-to-Maturity Accounting (2021) — link
  44. Private Equity Fund Fees (2014) — link
  45. 2021 Global Private Equity Outlook (2021) — link
  46. Global Asset Manager AuM Tops US$100 Trillion for the First Time (2020) — link
  47. Thomson Reuters Private Equity Buyout Index Methodology (2014) — link
  48. Valuing private equity investments strip by strip (2021)
  49. GRIP : Graphical Models Revealing Insights for Portfolio Replication - A Learning Approach (2024) — link
  50. Deep Decoding of Strategies (2022) — link
  51. An alternative approach to alternative beta (2007)
  52. Machine learning: a probabilistic perspective (2012)
  53. ‘Finance Wizard’at the FinLLM Challenge Task: Financial Text Summarization (2024) — link
  54. Report by the Commission on the Measurement of Economic Performance and Social Progress (2010)
  55. The Economics of Biodiversity: The Dasgupta Review (2021)
  56. The Age of Sustainable Development (2015)
  57. AI, Finance, and Sustainability: Building Responsible and Resilient Systems (2023)
  58. Private Equity Performance: What Do We Know? (2014)
  59. Do Private Equity Funds Manipulate Reported Returns? (2019) — link
  60. Machine Learning Private Equity Returns (2024) — link
  61. Sustainable Investing: Long-Term Evidence and Machine Learning Insights (2022)
  62. Probabilistic Graphical Models: Principles and Techniques (2009)
  63. Machine Learning: A Probabilistic Perspective (2012)
  64. PEARL: Private Equity Accessibility Reimagined with Liquidity (2024)

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