Pricing American options with exogenous and endogenous transaction costs

Dong Yan, Xin-Jie Huang, Guiyuan Ma, Xin-Jiang HeComput. Math. Appl 2025 · DOI

TL;DR

This paper develops a novel model for pricing American options that accounts for both liquidity risks (endogenous transaction costs) and explicit transaction fees (exogenous transaction costs). By incorporating these realistic market frictions, the model offers more accurate option pricing, particularly for at-the-money options, and provides insights into how these costs influence optimal exercise strategies.

What Problem Does This Paper Address?

The valuation of American options is notoriously complex due to their early exercise feature. Traditional option pricing models often assume frictionless markets, ignoring transaction costs and liquidity issues. However, in real-world trading, both explicit fees for transactions and the impact of market liquidity (how easily an asset can be bought or sold without affecting its price) significantly influence option prices and trading decisions. This paper aims to bridge this gap by developing a more realistic pricing model that explicitly incorporates both types of transaction costs, addressing the limitations of existing models that often focus on only one or the other, or ignore them entirely.

Methods

The authors propose a new dynamic model for the underlying asset that includes a mean-reverting Ornstein-Uhlenbeck process to capture stochastic liquidity. They then formulate a system of nonlinear partial differential equations (PDEs) to describe the option values for both the holder and the writer. To solve these complex PDEs, they employ an alternating direction implicit (ADI) finite difference scheme, which is known for its efficiency and stability in multi-dimensional problems. The model is calibrated using market data via maximum likelihood estimation (MLE), and its performance is compared against a benchmark model (Leland model) using root mean square error (RMSE).

Datasets

The study utilizes two datasets from the Wind database:

  • Daily closing prices of soybean meal futures contracts traded on the Dalian Commodity Exchange (January 2022 to January 2024).
  • American put options on soybean meal futures, also traded on the Dalian Commodity Exchange.

A filter was applied to exclude option contracts with daily trading volumes below 1,200 contracts to ensure data quality and liquidity.

Key Insights and Ideas

  • Interplay of Costs: The model highlights how exogenous transaction fees and endogenous liquidity risks interact to affect option pricing and optimal exercise decisions.
  • Holder vs. Writer: The option holding price is consistently lower than the writing price, and this "fair price range" widens with increasing transaction costs.
  • Liquidity's Impact: Rising market illiquidity generally leads to higher option prices, as holders are incentivized to wait for better trading conditions, and the optimal exercise boundary shifts.
  • Model Superiority: The proposed model, by incorporating liquidity dynamics, significantly outperforms the Leland model in pricing accuracy, especially for at-the-money options where precision is critical.

Key Results

  • The proposed model demonstrates significantly lower pricing errors (RMSE) compared to the benchmark Leland model across all moneyness categories. For instance, the overall RMSE for the new model is 4.82, compared to 16.60 for the benchmark.
  • The numerical scheme is shown to be approximately second-order convergent in time and fourth-order convergent in space.
  • Empirical analysis using soybean meal futures and options data confirms that the model's incorporation of liquidity risks leads to improved pricing accuracy.

Limitations and Future Work

  • Model Complexity: The resulting pricing PDEs are highly nonlinear and involve mixed derivative terms, necessitating sophisticated numerical methods like the ADI scheme.
  • Parameter Estimation: While the EKF-MLE approach is used for parameter estimation, the nonlinear nature of the model can still pose challenges.
  • Future Extensions: The authors suggest that the framework can be extended to incorporate other market factors such as stochastic volatility, stochastic interest rates, and jump diffusions, and applied to price various other option types.

Summary updated

BibTeX
@article{yan2025pricing,
  author = {Dong Yan and Xin-Jie Huang and Guiyuan Ma and Xin-Jiang He},
  title = {{Pricing American options with exogenous and endogenous transaction costs}},
  journal = {Comput. Math. Appl.},
  year = {2025},
  doi = {10.1016/j.camwa.2025.09.008},
}
Figure