Optimal Multiple Stopping with Negative Discount Rate and Random Refraction Times under Levy Models

Tim Leung, Kazutoshi Yamazaki, Hongzhong ZhangSSRN 2014 · DOI

TL;DR

This paper tackles the complex problem of deciding when to exercise a series of options, especially when the "cost" of waiting (discount rate) is negative, and there are random delays between opportunities. They develop a rigorous mathematical framework to find the optimal strategy, which is crucial for financial applications like stock loans and real options.

What Problem Does This Paper Address?

Imagine you have a sequence of opportunities to buy something at a fixed price (a call option), but you can only exercise one at a time, and there's a random waiting period between each potential exercise. This is the core of an "optimal multiple stopping problem." The authors extend this by considering two key challenges:

  1. Negative Discount Rate: In finance, a discount rate reflects the time value of money – money today is worth more than money tomorrow. However, in some situations, like when a loan's interest rate is higher than the market's risk-free rate, or when investment costs grow faster than a firm's discount rate, the effective "cost" of waiting can be negative. This means waiting might actually be more expensive than exercising immediately. Standard models often struggle with this.
  2. Random Refraction Times: The time between when one option can be exercised and when the next one becomes available isn't fixed. It's a random variable, adding another layer of uncertainty to the decision-making process.

The paper aims to provide a precise mathematical solution for when to exercise these options under these challenging conditions, which has direct implications for valuing financial instruments like stock loans and real options.

Methods

The authors employ a sophisticated combination of mathematical tools to solve this problem:

  • Lévy Processes: The underlying asset price (e.g., stock price) is modeled using Lévy processes, which are more general than standard Brownian motion as they can incorporate sudden jumps (like unexpected market news).
  • Laplace Transform and Change of Measure: To handle the negative discount rate, they use Laplace transforms and a technique called "change of measure." This effectively transforms the problem into a more manageable one where the discount rate is positive, allowing them to leverage existing mathematical machinery.
  • Martingale Theory: This is a fundamental tool in probability theory for analyzing processes that evolve randomly over time. It's used here to establish properties of the value functions and to prove the optimality of their derived strategies.
  • Recursive Approach: The multiple stopping problem is broken down into a sequence of single stopping problems, solved recursively. This means the solution for k opportunities builds upon the solution for k-1 opportunities.
  • Threshold Strategies: They show that the optimal strategy often involves exercising the option when the asset price crosses a certain "threshold." The paper focuses on characterizing these thresholds.

Datasets

Not discussed in the paper. The paper is a theoretical exploration of a mathematical model and does not involve empirical data analysis or datasets.

Key Insights and Ideas

  • Negative Discount Rate is Manageable: The paper demonstrates that even with a negative discount rate, a rigorous mathematical solution can be found using advanced techniques like change of measure. This expands the applicability of option pricing models.
  • Optimal Strategy is Threshold-Based: For a given number of remaining exercise opportunities, the optimal strategy is characterized by an "up-crossing time" relative to a specific threshold. This provides a clear decision rule.
  • Decreasing Thresholds: As more exercise opportunities are used up, the optimal exercise thresholds tend to decrease. This makes intuitive sense: with fewer chances left, you might be willing to exercise at a lower price.
  • Refraction Time Impact: The distribution of the random refraction times influences the optimal exercise thresholds, though the paper shows how to analyze this impact systematically.

Key Results

  • The paper provides a rigorous mathematical framework for optimal multiple stopping under Lévy processes with negative discount rates and random refraction times.
  • They derive explicit formulas for the optimal exercise thresholds in the single stopping case and recursive formulas for the multiple stopping case.
  • The optimal strategy is shown to be a threshold strategy, meaning the option is exercised when the underlying asset price crosses a specific level.
  • Numerical examples illustrate how optimal thresholds change with the mean refraction time, showing a general decrease in thresholds as more opportunities are exercised.

Limitations and Future Work

  • Computational Complexity: While the paper provides analytical results, the numerical implementation for a large number of exercise opportunities can still be computationally challenging, as noted in the numerical examples section.
  • Generalization to Other Payoffs: The current work focuses on call-like payoffs. The authors suggest that their framework could be extended to analyze other payoff structures, such as put options, which would be a natural direction for future research.
  • Specific Lévy Processes: While the framework is general, specific applications might require detailed analysis for particular Lévy process specifications.

Summary updated

BibTeX
@misc{leung2014optimal,
  author = {Tim Leung and Kazutoshi Yamazaki and Hongzhong Zhang},
  title = {{Optimal Multiple Stopping with Negative Discount Rate and Random Refraction Times under Levy Models}},
  year = {2014},
  doi = {10.2139/ssrn.2395256},
  howpublished = {SSRN},
}
Figure