Portfolio optimization with insider's initial information and counterparty risk

Caroline Hillairet, Ying JiaoFinance Stochastics 2015 · DOI

TL;DR

This paper explores how an insider with private information about a counterparty's default risk can achieve better portfolio returns than a standard investor. By modeling default risk and using advanced filtration techniques, the authors show that while insiders generally gain, extreme scenarios can lead to unexpected outcomes, highlighting the crucial role of short-selling constraints.

What Problem Does This Paper Address?

The paper tackles the problem of portfolio optimization in financial markets where a key counterparty's default risk is a significant concern. Specifically, it investigates how an "insider" who possesses private information about the timing or nature of this default can leverage this advantage to improve their investment strategy compared to a "standard investor" who only has access to public information. The challenge lies in quantifying the precise gain an insider can achieve and understanding the conditions under which this advantage might diminish or even reverse, especially when considering the complexities of credit risk and market regulations like short-selling constraints.

Methods

The authors employ a sophisticated mathematical framework rooted in stochastic calculus and financial mathematics. Key techniques include:

  • Enlargement of Filtrations: This is central to modeling the insider's superior information. The insider's knowledge about the default barrier (L) is incorporated by enlarging the standard filtration (F) to a larger one (GM), which represents the insider's information flow.
  • Stochastic Processes and SDEs: The price dynamics of the risky asset and the wealth process are modeled using stochastic differential equations (SDEs). The default event introduces a regime change in these dynamics.
  • Dynamic Programming and Duality: The optimization problems for both the insider and the standard investor are solved using dynamic programming principles and duality methods, which are standard tools for such problems.
  • Radon-Nikodym Derivative: This is used to handle the change of probability measure induced by the insider's information, playing a crucial role in relating the insider's problem to a more tractable one.
  • Backward Stochastic Differential Equations (BSDEs): These are used to characterize the optimal strategies and value functions, particularly in the after-default and before-default optimization stages.

What's novel is the application of these techniques to the specific context of counterparty default risk with an initial information advantage (knowing the default barrier L from the start) rather than just information revealed at the default time.

Datasets

Not discussed in the paper. The paper focuses on theoretical modeling and numerical illustrations based on assumed parameter values, not real-world datasets.

Key Insights and Ideas

  • Insider Advantage is General but Not Absolute: Insiders generally achieve higher expected utility due to their superior information. However, this advantage is not guaranteed, especially in extreme scenarios of default and loss risks.
  • Short-Selling Constraints are Crucial: Without regulation, an insider could achieve unbounded wealth. The paper demonstrates the necessity of imposing short-selling constraints to ensure well-posed optimization problems and prevent unrealistic outcomes.
  • Information Structure Matters: The paper meticulously models how the insider's knowledge of the default barrier (L) from the initial time, as opposed to only observing it at default, impacts their strategy and outcomes.
  • Extreme Scenarios Can Flip the Advantage: In situations with very high default intensity and loss probabilities, an investor (even without insider information) might outperform the insider if they can exploit aggressive short-selling strategies that are disallowed for the insider. This highlights a trade-off between information and strategic flexibility.
  • The Role of the Default Density: The paper contrasts its approach with previous work that relied on "default density." It shows that its method, using Radon-Nikodym derivatives related to the insider's information, is more general and can handle cases where default density might not exist (e.g., when the law of L has atoms).

Key Results

  • The paper provides a theoretical framework for portfolio optimization for an insider with initial information about counterparty default risk.
  • It proves the necessity of short-selling constraints for the insider to avoid unbounded wealth.
  • The authors derive explicit formulas for the optimal wealth and value functions for both the insider and a standard investor.
  • Numerical illustrations (Figures 1-4) demonstrate that:
  • The insider generally outperforms the standard investor and a Merton strategy.
  • Extreme default and loss risks can lead to the investor outperforming the insider.
  • The short-selling constraint significantly impacts the insider's gain.
  • The value function increases with the loss given default, particularly for the investor in high-risk scenarios.

Limitations and Future Work

  • Model Assumptions: The model relies on specific assumptions about the market structure (e.g., two assets, risk-free bond as numéraire), the utility function (CRRA), and the dynamics of the risky asset and default process.
  • "Insider" Definition: The insider's information is precisely defined as knowing the random barrier L from the start. Real-world insider information might be more complex.
  • Numerical Illustrations: While insightful, the numerical results are based on specific parameter choices and may not generalize to all possible scenarios.

Future work could explore:

  • More complex market structures or information sets.
  • Different types of utility functions.
  • The impact of other market frictions or regulations.
  • Empirical validation using real market data.

Summary updated

BibTeX
@article{hillairet2015portfolio,
  author = {Caroline Hillairet and Ying Jiao},
  title = {{Portfolio optimization with insider's initial information and counterparty risk}},
  journal = {Finance Stochastics},
  year = {2015},
  doi = {10.1007/s00780-014-0246-7},
}
Figure