Paper: arXiv 2404.08903
Authors: Anna Knezevic
Abstract
Enhancing the existing solution for pricing of fixed income instruments within Black-Karasinski model structure, with neural network at various parameterisation points to demonstrate that the method is able to achieve superior outcomes for multiple calibrations across extended projection horizons.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced mathematical concepts including path integrals, Taylor series expansions, and PDE approximations, but lacks empirical validation with backtests or statistical metrics, focusing instead on theoretical model formulation.
Research Flowchart
flowchart TD A["Research Goal"] --> B["Data & Calibration"] A --> C["Methodology"] B --> D["Path-Integral Approx."] C --> D D --> E["Neural Network Enh."] E --> F["Computational Process"] F --> G["Key Outcomes"] subgraph Inputs A B C end subgraph Processing D E F end subgraph Results G end
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 enhancing the existing Black‑Karasinski pricing solution for fixed‑income instruments by integrating a neural network to achieve superior outcomes across multiple calibrations and extended horizons.
Source sentence
Enhancing the existing solution for pricing of fixed income instruments within Black-Karasinski model structure, with neural network at various parameterisation points to demonstrate that the method is able to achieve superior outcomes for multiple calibrations across extended projection horizons.
Assumes 3
- The drift term is assumed to be an affine function. model
Source sentence
drift term being and affine function
- Normalization of parameters is performed relative to the entire calibration space. data
Source sentence
The normalisation was performed relative to the entire calibration space.
- The neural network relies on derivatives of the functions for optimisation. model
Source sentence
reliance on derivatives of the functions for optimisation
Data
Synthetic data only.
Costs
Not applicable
Finds 2
- For subsets with relaxed interaction between sigma and a, the neural network does not improve results; however, with stronger interaction between mean reversion and volatility, additional coefficients improve fit for longer horizons. pricing-calibration
Source sentence
more pronounced interaction between mean reversion parameter and volatility there seems to be additional coefficients in the equation that produce better fit to results for longer horizons.
- The main improvement stemmed from using biases in the optimisation, indicating the existing solution does not capture all underlying effects. theoretical
Source sentence
The main improvement came from the use of biases in the optimisation, which suggests that the structure of the existing solution is yet to capture all the effects of the underlying.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 9 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.
Cited References (35)
- Interest Rate Models - Theory and Practice (2001) — link
- Variational approach to quantum statistical mechanics of nonlinear systems with application to sine-Gordon chains (1985)
- Effective classical partition functions (1986)
- Term Structures: Interest Rate Models (2006)
- The Art of Term Structure Models: Volatility and Distribution — link
- Modelling interest rates in 2014 — link
- Stochastic Control for Asset Management (2013) — link
- A path-integral approximation for non-linear diffusions (2020)
and 27 more
- Closed-form Arrow-Debreu pricing for the Hull-White short rate model (2019)
- Neural networks, approximation theory, and finite precision computation (1995)
- Approximations Of Bond And Swaption Prices In A Black–Karasi 'n ski Model (2016)
- General short-rate analytics (2011)
- An effective approximation for zero-coupon bonds and Arrow–Debreu prices in the Black–Karasinski model (2014)
- A closed-form approximation of likelihood functions for discretely sampled diffusions: The exponent expansion (2007)
- Exact arrow-debreu pricing for the black-karasinski short rate model (2019)
- The Black-Karasinski Model: Thirty Years On (2021)
- Two-Factor Black-Karasinski Pricing Kernel (2019)
- Invariant approaches for the analytic solution of the stochastic Black-Derman toy model (2018)
- Multivariate binomial approximations for asset prices with nonstationary variance and covariance characteristics (1995)
- Numerical procedures for implementing term structure models I: Single-factor models (1994)
- Short rate models: Hull-White or Black-Karasinski? Implementation note and model comparison for ALM (2008)
- An introduction to stochastic processes in physics (2003)
- Numerical procedures for implementing term structure models II: Two-factor models (1994)
- Neural taylor approximations: Convergence and exploration in rectifier networks (2017)
- Tests of non linear Gaussian term structure models (2016)
- Interest Rate Models: Paradigm shifts in recent years (2007)
- An empirical comparison of forward-rate and spot-rate models for valuing interest-rate options (1999)
- Connectionist learning procedures. Artificial Intelligence, 40 1-3: 185 234, 1989. Reprinted in J. Carbonell, editor," (1990)
- Scikit-learn: Machine Learning in P ython (2011)
- A sigmoid approximation of the standard normal integral (1996)
- An introduction to numerical methods for stochastic differential equations (1999)
- A Two-Factor Black-Karasinski Sovereign Credit Default Swap Pricing Model. (2007)
- Bond and option pricing when short rates are lognormal (1991)
- Stochastic Gradient Descent (2010)
- Binomial term structure models (1998)
0 of 35 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.