A model-free backward and forward nonlinear PDEs for implied volatility
Pith reviewed 2026-05-24 20:16 UTC · model grok-4.3
The pith
Nonlinear PDEs govern implied volatility for convex-payoff claims without assuming asset dynamics.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We derive a backward and forward nonlinear PDEs that govern the implied volatility of a contingent claim whenever the latter is well-defined. This would include at least any contingent claim written on a positive stock price whose payoff at a possibly random time is convex. We also discuss suitable initial and boundary conditions for those PDEs. Finally, we demonstrate how to solve them numerically by using an iterative finite-difference approach.
What carries the argument
Nonlinear PDEs satisfied by implied volatility, obtained from the convexity of the claim payoff.
If this is right
- Implied volatility obeys a backward nonlinear PDE with payoff-derived conditions.
- Implied volatility also obeys a forward nonlinear PDE.
- The PDEs admit an iterative finite-difference numerical solution without specifying asset dynamics.
- The equations hold for claims with random maturity times.
Where Pith is reading between the lines
- The same PDEs might be used to test whether an observed volatility surface is consistent with some convex payoff.
- Links may exist to other model-free relations such as those connecting implied and local volatility.
- Numerical solution of the forward PDE could support calibration exercises that avoid full stochastic models.
Load-bearing premise
The payoff of the contingent claim is convex.
What would settle it
A numerical check showing that the implied volatility surface of a convex-payoff claim on a positive stock fails to satisfy the derived PDEs would falsify the result.
read the original abstract
We derive a backward and forward nonlinear PDEs that govern the implied volatility of a contingent claim whenever the latter is well-defined. This would include at least any contingent claim written on a positive stock price whose payoff at a possibly random time is convex. We also discuss suitable initial and boundary conditions for those PDEs. Finally, we demonstrate how to solve them numerically by using an iterative finite-difference approach.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper derives model-free backward and forward nonlinear PDEs governing the implied volatility of contingent claims with convex payoffs on positive stock prices (whenever the implied volatility is well-defined), discusses suitable initial and boundary conditions, and demonstrates numerical solution via an iterative finite-difference scheme.
Significance. If the central derivations are correct and the numerical method is stable, the result supplies a model-independent framework for implied-volatility dynamics that could be useful for pricing and hedging convex claims without committing to a specific underlying process; the explicit conditioning on well-definedness and convexity is a strength.
major comments (2)
- [§3] §3, Eq. (3.8): the transition from the general Dupire-type forward equation to the nonlinear PDE in implied volatility appears to invoke the chain rule on the Black-Scholes formula, but the manuscript does not explicitly verify that the resulting coefficients remain well-defined when the payoff convexity is only assumed at a random maturity rather than at a fixed T.
- [§4.2] §4.2, boundary-condition paragraph: the proposed far-field condition for the forward PDE is stated as σ(t,K)→0 as K→∞, yet no justification is given that this is compatible with the convexity assumption when the payoff is only convex at a random time; this condition is load-bearing for the well-posedness claim.
minor comments (2)
- [Abstract] The abstract contains the grammatical error 'a backward and forward nonlinear PDEs'; this should read 'backward and forward nonlinear PDEs'.
- [§2] Notation: the symbol σ(t,K) is used for implied volatility without an explicit reminder that it is the unique solution to the Black-Scholes equation for the given convex payoff; a short clarifying sentence in §2 would help.
Simulated Author's Rebuttal
We thank the referee for the careful reading and constructive comments. We respond to each major comment below.
read point-by-point responses
-
Referee: [§3] §3, Eq. (3.8): the transition from the general Dupire-type forward equation to the nonlinear PDE in implied volatility appears to invoke the chain rule on the Black-Scholes formula, but the manuscript does not explicitly verify that the resulting coefficients remain well-defined when the payoff convexity is only assumed at a random maturity rather than at a fixed T.
Authors: The derivation begins from the model-free Dupire forward equation for the price of any convex claim (which holds at a random maturity by the convexity assumption) and substitutes the Black-Scholes representation in terms of implied volatility. The resulting coefficients are therefore expressed solely in terms of the implied volatility surface and its derivatives, which exist by the standing assumption that implied volatility is well-defined. We nevertheless agree that an explicit remark confirming the coefficients remain well-defined under random maturity would improve clarity, and we will insert a short clarifying paragraph in the revised Section 3. revision: partial
-
Referee: [§4.2] §4.2, boundary-condition paragraph: the proposed far-field condition for the forward PDE is stated as σ(t,K)→0 as K→∞, yet no justification is given that this is compatible with the convexity assumption when the payoff is only convex at a random time; this condition is load-bearing for the well-posedness claim.
Authors: Under the convexity assumption the price function remains convex in strike at any (random) time, which implies that the call price C(t,K) satisfies the no-arbitrage bound C(t,K) = o(K) as K→∞; matching this price to the Black-Scholes formula forces the implied volatility to vanish at infinity. This asymptotic is therefore compatible with convexity at random time. We will add a brief justification of the far-field condition in the revised boundary-conditions paragraph. revision: partial
Circularity Check
No significant circularity detected in derivation chain
full rationale
The paper's central claim is a derivation of backward and forward nonlinear PDEs for implied volatility, explicitly scoped to cases where implied volatility is well-defined (convex payoffs on positive stock prices). This condition is stated as the domain of validity upfront rather than derived from or fitted to the result itself. No load-bearing steps reduce to self-definition, fitted inputs renamed as predictions, self-citation chains, uniqueness theorems imported from authors, ansatzes smuggled via citation, or renaming of known results. The derivation is presented as model-free and self-contained against the stated assumptions, with no evidence of circular reduction by the paper's own equations.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption The contingent claim has a convex payoff at a possibly random time on a positive stock price.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.