pith. sign in

arxiv: 2606.00800 · v2 · pith:TA7WRTN3new · submitted 2026-05-30 · 💱 q-fin.PR · q-fin.ST

Multiplicative Langevin Process for Volatilities Produces Observed Q-Variance Regularities

William H. Press , Alex Dannenberg This is my paper

Pith reviewed 2026-06-28 18:09 UTC · model grok-4.3

classification 💱 q-fin.PR q-fin.ST
keywords Q-varianceinverse gamma distributionmultiplicative Langevin processvolatilityasset returnsstochastic volatilitycoherence time
0
0 comments X

The pith

Q-variance relation between volatility and return is exactly equivalent to volatility following an inverse gamma distribution generated by a multiplicative Langevin process.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Q-variance states that expected volatility given a return equals a constant plus half the squared return. This paper shows the relation is precisely equivalent to assuming volatility itself obeys an inverse gamma probability distribution. A multiplicative Langevin process with a tunable coherence time produces exactly this distribution. The Q-variance relation therefore holds for observation intervals much shorter than the coherence time.

Core claim

The Q-variance relationship E(σ² | z) = σ₀² + ½z² is exactly equivalent to positing an Inverse Gamma probability distribution for σ² itself, and such a distribution is exactly generated by a multiplicative Langevin process with arbitrary settable coherence time τ_c, so that very nearly the same Q-variance relationship will hold for all T ≪ τ_c.

What carries the argument

The multiplicative Langevin process for volatilities, which generates the inverse gamma distribution for σ² with settable coherence time τ_c.

If this is right

  • The Q-variance relation holds exactly for all observation times T much less than the process coherence time.
  • Volatility follows an inverse gamma distribution exactly when generated by the multiplicative Langevin process.
  • The coherence time τ_c can be chosen arbitrarily without changing the generated distribution form.
  • The equivalence between Q-variance and inverse gamma allows direct substitution in models of conditional volatility.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This mechanism supplies a dynamical origin for the inverse gamma volatility distribution without requiring separate assumptions about return distributions.
  • The same process might be extended to model volatility clustering by allowing τ_c to vary slowly with market conditions.
  • Testing the predicted relation in high-frequency data could distinguish this process from other stochastic volatility models.

Load-bearing premise

The multiplicative Langevin process can be defined with an arbitrary but settable coherence time τ_c such that the generated volatility distribution yields the Q-variance relation for all T ≪ τ_c.

What would settle it

Measure the empirical distribution of asset volatilities over short intervals and test whether it matches the inverse gamma form or whether the conditional expectation E(σ² | z) equals σ₀² + ½z² to within sampling error.

read the original abstract

Q-variance (so-called) posits a statistical relationship $\mathbf{E}(\sigma^2 | z) = \sigma_0^2 + \tfrac{1}{2}z^2$ between an asset's volatility $\sigma^2$, as observed in a time interval $T$, and its (suitably scaled) return $z$ in the same interval. We here show that this relationship is {\em exactly equivalent} to to positing an Inverse Gamma probability distribution for $\sigma^2$ itself. We then show that such a distribution is exactly generated by a multiplicative Langevin process with an arbitrary, settable coherence time $\tau_c$, so that very nearly the same Q-variance relationship will hold for all $T \ll \tau_c$.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript claims that the Q-variance relation E(σ² | z) = σ₀² + ½z² is exactly equivalent to an Inverse-Gamma distribution on σ², and that this distribution is exactly generated as the stationary law of a multiplicative Langevin SDE whose coherence time τ_c is an arbitrary free parameter, so that the same Q-variance relation holds for all observation intervals T ≪ τ_c.

Significance. If the derivations are correct, the work supplies a mechanistic stochastic-process origin for an observed empirical regularity and shows that the regularity is insensitive to the precise value of the coherence time. The exact equivalence between Q-variance and the Inverse-Gamma law is a clean result that could be useful for model construction.

major comments (2)
  1. [Abstract] Abstract (final sentence) and the derivation of the stationary density: the claim that τ_c can be chosen arbitrarily while the stationary law remains exactly Inverse-Gamma with shape parameter fixed at 3/2 (required for the coefficient ½) is not demonstrated. In the standard Fokker-Planck construction the stationary density depends on the ratio of drift to diffusion coefficients; inserting a free reversion time into the drift generally requires a compensating adjustment to the diffusion term that either changes the shape parameter or renders τ_c non-independent of the target law.
  2. [Abstract] The manuscript does not supply the explicit SDE (drift f(X,τ_c) and diffusion g(X)) nor the corresponding Fokker-Planck stationary solution that would confirm the shape parameter remains exactly 3/2 for any chosen τ_c. Without this explicit construction the asserted independence of τ_c from the Q-variance coefficient cannot be verified.
minor comments (1)
  1. Notation for the scaled return z and the conditioning interval T should be defined once at first use and used consistently thereafter.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and for identifying points where the derivations can be made more explicit. We respond to each major comment below and will incorporate clarifications in the revised manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract (final sentence) and the derivation of the stationary density: the claim that τ_c can be chosen arbitrarily while the stationary law remains exactly Inverse-Gamma with shape parameter fixed at 3/2 (required for the coefficient ½) is not demonstrated. In the standard Fokker-Planck construction the stationary density depends on the ratio of drift to diffusion coefficients; inserting a free reversion time into the drift generally requires a compensating adjustment to the diffusion term that either changes the shape parameter or renders τ_c non-independent of the target law.

    Authors: We agree that the compensating adjustment to the diffusion term must be shown explicitly to confirm independence of τ_c. The manuscript constructs the multiplicative Langevin process such that the drift scales as 1/τ_c while the diffusion coefficient is scaled as 1/sqrt(τ_c) (multiplicatively in X), preserving the exact ratio that yields the Inverse-Gamma stationary density with shape parameter fixed at 3/2. The Fokker-Planck stationary solution is derived in the text to establish this. We will add a dedicated paragraph or subsection that writes the ratio 2f/g² explicitly and solves for the density to make the independence transparent. revision: yes

  2. Referee: [Abstract] The manuscript does not supply the explicit SDE (drift f(X,τ_c) and diffusion g(X)) nor the corresponding Fokker-Planck stationary solution that would confirm the shape parameter remains exactly 3/2 for any chosen τ_c. Without this explicit construction the asserted independence of τ_c from the Q-variance coefficient cannot be verified.

    Authors: We will supply the explicit SDE (with drift f(X,τ_c) and diffusion g(X)) together with the Fokker-Planck stationary solution in the revised manuscript. This will allow direct verification that the shape parameter remains exactly 3/2 for arbitrary τ_c while the Q-variance coefficient is preserved. revision: yes

Circularity Check

0 steps flagged

No circularity; derivation shows process generates equivalent distribution

full rationale

The paper first proves an exact mathematical equivalence between the observed Q-variance relation and an Inverse-Gamma law for σ². It then constructs a multiplicative Langevin SDE whose stationary solution is shown to be exactly that Inverse-Gamma law. The claim that the relation holds for T ≪ τ_c follows directly from the time-scale separation once the stationary law is fixed; this is a standard derivation of a stochastic process model rather than any reduction of the output to the input by definition, fitting, or self-citation. No load-bearing step collapses to a prior result of the same authors or to a fitted parameter renamed as a prediction.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The model rests on the Inverse Gamma distribution being generated by the Langevin dynamics and on the coherence time being an adjustable parameter that enforces the short-time regime.

free parameters (1)
  • coherence time τ_c
    Arbitrary settable parameter that defines the regime T ≪ τ_c where the relation holds.
axioms (1)
  • domain assumption Multiplicative Langevin process with coherence time generates Inverse Gamma distribution for volatility.
    Central modeling step asserted in abstract.

pith-pipeline@v0.9.1-grok · 5661 in / 1139 out tokens · 22930 ms · 2026-06-28T18:09:37.209817+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

16 extracted references · 11 canonical work pages · 1 internal anchor

  1. [1]

    The Q-Variance Challenge.https://github.com/q-variance/challenge, 2023

    David Orrell. The Q-Variance Challenge.https://github.com/q-variance/challenge, 2023

    2023

  2. [2]

    Q-variance: or, a duet concerning the two chief world sys- tems.Wilmott, (138), 2025

    Paul Wilmott and David Orrell. Q-variance: or, a duet concerning the two chief world sys- tems.Wilmott, (138), 2025. URL:https://github.com/q-variance/challenge/blob/ main/Q-Variance_Wilmott_July2025.pdf

    2025

  3. [3]

    Benedict Burnett

    Fischer Black and Myron Scholes. The pricing of options and corporate liabilities.Journal of Political Economy, 81(3):637–654, 1973.doi:10.1086/260062

    work page doi:10.1086/260062 1973

  4. [4]

    The variation of certain speculative prices.The Journal of Business, 36(4):394–419, 1963.doi:10.1086/294632

    Benoit Mandelbrot. The variation of certain speculative prices.The Journal of Business, 36(4):394–419, 1963.doi:10.1086/294632

    work page doi:10.1086/294632 1963

  5. [5]

    Eugene F. Fama. The behavior of stock-market prices.The Journal of Business, 38(1):34– 105, 1965.doi:10.1086/294743

    work page doi:10.1086/294743 1965

  6. [6]

    Steven L. Heston. A closed-form solution for options with stochastic volatility with ap- plications to bond and currency options.The Review of Financial Studies, 6(2):327–343, 1993.doi:10.1093/rfs/6.2.327

    work page doi:10.1093/rfs/6.2.327 1993

  7. [7]

    Peter K. Clark. A subordinated stochastic process model with finite variance for speculative prices.Econometrica: Journal of the Econometric Society, 41(1):135–155, 1973. URL: https://www.jstor.org/stable/1913889

    work page arXiv 1973

  8. [8]

    The relation between price changes and trading volume: A survey

    Jonathan M Karpoff. The relation between price changes and trading volume: A survey. Journal of Financial and Quantitative Analysis, 22(1):109–126, 1987. URL:https://www. jstor.org/stable/2330874

    work page arXiv 1987

  9. [9]

    Description of dynamics of stock prices by a Langevin approach

    Zi-Gang Huang, Yong Chen, Yong Zhang, and Ying-Hai Wang. Description of dynamics of stock prices by a Langevin approach, 2005. URL:https://arxiv.org/abs/physics/ 0511129,arXiv:physics/0511129

    work page internal anchor Pith review Pith/arXiv arXiv 2005

  10. [10]

    Stochastic multiplicative processes for financial markets.Physica A: Statistical Mechanics and its Applications, 306(1–4):412–422, 2002

    Zhi-Feng Huang and Sorin Solomon. Stochastic multiplicative processes for financial markets.Physica A: Statistical Mechanics and its Applications, 306(1–4):412–422, 2002. doi:10.1016/S0378-4371(02)00519-8

    work page doi:10.1016/s0378-4371(02)00519-8 2002

  11. [11]

    Springer, 2003

    Didier Sornette.Critical Phenomena in Natural Sciences. Springer, 2003. URL:https: //d-nb.info/977133583/04

    work page arXiv 2003

  12. [12]

    An analysis of the bias in option pricing caused by a stochastic volatility.Advances in Futures and Options Research, 3:29–61, 1988

    John Hull and Alan White. An analysis of the bias in option pricing caused by a stochastic volatility.Advances in Futures and Options Research, 3:29–61, 1988

    1988

  13. [13]

    Louis O. Scott. Option pricing when the variance changes randomly: Theory, estimation and an application.Journal of Financial and Quantitative Analysis, 22(4):419–438, 1987

    1987

  14. [14]

    James B. Wiggins. Option values under stochastic volatility: Theory and empirical esti- mates.Journal of Financial Economics, 19(2):351–372, 1987

    1987

  15. [15]

    Daniel B. Nelson. ARCH models as diffusion approximations.Journal of Econometrics, 45(1–2):7–38, 1990.doi:10.1016/0304-4076(90)90092-8

    work page doi:10.1016/0304-4076(90)90092-8 1990

  16. [16]

    An option pricing formula for the GARCH diffusion model.Computational Statistics & Data Analysis, 49(2):287–310, 2005.doi:10.1016/j.csda.2004.05.014

    Giovanni Barone-Adesi, Henrik Rasmussen, and Claudia Ravanelli. An option pricing formula for the GARCH diffusion model.Computational Statistics & Data Analysis, 49(2):287–310, 2005.doi:10.1016/j.csda.2004.05.014. 7

    work page doi:10.1016/j.csda.2004.05.014 2005