Paper: arXiv 2306.05750
Authors: Takuji Arai, Yuto Imai
Abstract
The Barndorff-Nielsen and Shephard model is a representative jump-type stochastic volatility model. Still, no method exists to compute option prices numerically for the non-martingale case with infinite active jumps. We develop two simulation methods for such a case under change of measure and conduct some numerical experiments.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 5.0/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper is mathematically dense, focusing on advanced stochastic calculus and measure theory to develop novel simulation methods for a complex financial model. While it includes numerical experiments and algorithmic details, the empirical component is more theoretical validation than heavy backtesting with real-world data.
Research Flowchart
flowchart TD
A["Research Goal: Develop Monte Carlo methods for<br>Barndorff-Nielsen and Shephard model<br>under change of measure with infinite active jumps"] --> B["Methodology: Develop Two Simulation Approaches"]
B --> C["Input: Model Parameters<br>Volatility Jump Lévy Measure &<br>Infinitely Active Jumps"]
C --> D["Computational Process: Monte Carlo Simulation"]
D --> E{"Simulation Substeps"}
E --> E1["Discretize Time &<br>Generate Jump Times"]
E --> E2["Simulate Volatility Dynamics<br>under Change of Measure"]
E --> E3["Simulate Asset Price Paths<br>with Infinite Jumps"]
E --> E4["Calculate Option Payoffs<br>& Discount Values"]
E --> F["Computational Process: Average Results<br>across Many Simulations"]
F --> G["Outcomes: Numerical Option Prices<br>for Non-Martingale Case<br>Validated via Numerical Experiments"]
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 two Monte‑Carlo simulation algorithms for pricing options under the non‑martingale Barndorff‑Nielsen and Shephard model with infinite active jumps.
Source sentence
We develop two simulation methods for such a case under change of measure and conduct some numerical experiments.
Assumes 3
- The model parameters satisfy b2/2 > 2 (1−e^{−λT})/λ ∨|ρ| and α e^{−λT} σ0^2 + C_ρ^2 > −1. model
Source sentence
Assumption 2.1. Throughout this paper, we assume that b2/2 > 2 (1−e−λT)/λ ∨|ρ| and α e−λT σ2_0 + Cρ^2 > −1.
- The financial market consists of one riskless asset and one risky asset. market
Source sentence
Throughout this paper, we consider a financial market composed of one riskless asset and one risky asset.
- The risk‑free interest rate is set to zero for simplicity. market
Source sentence
Without loss of generality, we may assume that the interest rate is zero for simplicity.
Data
- Synthetic Monte‑Carlo paths for the BNS model synthetic unknown
Source sentence
Table 2: Results of Algorithm 4.3.
Costs
Not applicable
Finds 4
- Algorithm 3.1 yields errors below 1 % when the drift coefficient α ≤ 0.1, providing sufficient accuracy for small α. performance
Source sentence
As for Algorithm 3.1, when α ≤0.1, the errors are less than 1%, which ensures sufficient accuracy.
- Algorithm 4.3 remains sufficiently accurate for larger α values if the number of time steps M is increased. performance
Source sentence
On the other hand, Algorithm 4.3 is sufficiently accurate for larger values of α if M is increased.
- Even for the extreme case α = 100, errors can be kept below 1 % by using M = 20 000 time steps. performance
Source sentence
For example, even in the extreme case of α =100, the errors can be kept below 1% taking M =20,000.
- Two simulation algorithms for the non‑martingale IG‑OU type BNS model under the minimal martingale measure have been developed; Algorithm 3.1 is faster but limited to small α, while Algorithm 4.3 is slower but accurate for any α. theoretical
Source sentence
Two simulation algorithms for the non-martingale IG-OU type BNS model under the MMM have been developed in this paper. Algorithm 3.1 only works when the drift coefficient α is small, but it is faster than Algorithm 4.3.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 17 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.