Paper: arXiv 2611.22604

Authors: Guolin Chen, Jiang Zhou, Richard Ma, Xia Sheng, Anson Tse, Terry Wei

Abstract

Options on leveraged ETFs (LETFs) are priced and hedged off the LETF volatility surface, but path-dependent variance drag makes that surface difficult to build consistently with the underlying index smile. We study three methods for constructing it: a closed-form first-order strike remapping, a new conditional-Gamma method, and a two-dimensional joint-law PDE that makes no distributional assumption and serves as the benchmark. The conditional-Gamma method is our main contribution. It conditions on the terminal underlying return, which preserves the state dependence of variance drag that a single remapped strike discards. Three one-dimensional forward PDEs under a calibrated local-volatility model give the conditional mean and variance of integrated variance, and a Gamma law matched to these moments keeps realised variance non-negative and prices the option in closed form. A shifted-Gamma variant that also matches the conditional third moment gives its clearest gains for inverse products at longer maturities, and a Gamma-mixture extension covers the full range of conditional skewness compatible with non-negative realised variance. On S&P 500 and Nikkei 225 surfaces the conditional-Gamma smile tracks the benchmark closely at every leverage tested, including the wings where the strike remapping drifts. A full LETF smile takes 0.85-1.42 seconds against 197-224 seconds for the benchmark, a 158- to 236-fold speed-up. Against listed TQQQ quotes the conditional-Gamma smile lies inside the market bid-ask at most strikes.

Complexity vs Empirical Score

  • Math Complexity: 9.0/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: This paper presents a highly sophisticated mathematical framework for pricing LETF options, introducing a novel conditional-Gamma method. It rigorously tests this method against a benchmark PDE and market quotes, demonstrating significant computational efficiency and accuracy. The combination of advanced theory and strong empirical validation places it firmly in the ‘Holy Grail’ quadrant.

Research Flowchart

  flowchart TD
    A["Research Goal: Consistently Price & Hedge LETF Options off Underlying Index Smile"] --> B{"Key Methodology: Conditional-Gamma Method"};
    B --> C1["Method 1: Closed-form Strike Remapping"];
    B --> C2["Method 2: Conditional-Gamma Method (Main Contribution)"];
    B --> C3["Method 3: 2D Joint-Law PDE (Benchmark)"];
    C2 --> D["Computational Processes: 3 1D Forward PDEs, Gamma Law Matching (1st/2nd/3rd moments)"];
    D --> E["Data/Inputs: S&P 500 & Nikkei 225 Surfaces, Calibrated Local-Volatility Model, Listed TQQQ Quotes"];
    E --> F["Key Findings: Conditional-Gamma tracks benchmark closely, 158-236x speed-up, lies within market bid-ask for TQQQ"];

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 a conditional‑Gamma pricing method for leveraged‑ETF options, which conditions on the terminal underlying return and uses a Gamma law matched to conditional mean and variance to price options in closed form.

Source sentence
The conditional-Gamma method is our main contribution. It conditions on the terminal underlying return, which preserves the state dependence of variance drag that a single remapped strike discards.

Assumes 5

  • A calibrated local‑volatility (LV) model for the underlying index is assumed. model
    Source sentence
    The practical question is how the LETF implied-volatility smile relates to the smile of the underlying, given a calibrated local-volatility (LV) model.
  • Discrete rebalancing, tracking error, jumps and random fees are excluded from the model. execution
    Source sentence
    Discrete rebalancing, tracking error, jumps, and random fees are outside the present idealisation.
  • The underlying asset follows the local‑volatility dynamics dS_t/S_t = μ_S(t)dt + σ(t,S_t)dW_t with deterministic drift μ_S(t). model
    Source sentence
    2 Definition 2.1 (Local-volatility dynamics). Let St denote the underlying asset price satisfying dSt St = µS t dt + σ(t, St) dWt, (1) where µS t is a deterministic drift (determined by rates, dividends, and borrow costs), σ(t, s) is the local volatility function, and Wt is a standard Brownian motion.
  • The leveraged ETF satisfies dL_t/L_t = β dS_t/S_t + φ_t dt, where φ_t is deterministic financing cost and fee term. model
    Source sentence
    2.2 Leveraged ETF A leveraged ETF with constant leverage ratio β satisfies [2, 12, 11] dLt Lt = β dSt St + φt dt, (2) where φt collects deterministic financing costs and fees.
Show 1 more
  • Realised integrated variance I_T is non‑negative (I_T ≥ 0). model
    Source sentence
    7 Conditional-Gamma Closure The realised variance is a non-negative additive functional, IT = Z T 0 a(t, Xt) dt ≥0.

Data

  • S&P 500 implied‑volatility surface options public
    SPX index
    Source sentence
    On S&P 500 and Nikkei 225 surfaces the conditional-Gamma smile tracks the benchmark closely at every leverage tested, including the wings where the strike remapping drifts. Against it, the conditional-Gamma smile stays close at every leverage tested on SPX and NKY, including the wings where the strike remapping drifts.
  • Nikkei 225 implied‑volatility surface options public
    NKY index
    Source sentence
    On S&P 500 and Nikkei 225 surfaces the conditional-Gamma smile tracks the benchmark closely at every leverage tested, including the wings where the strike remapping drifts. Against it, the conditional-Gamma smile stays close at every leverage tested on SPX and NKY, including the wings where the strike remapping drifts.
  • Listed QQQ and TQQQ option quotes options public
    20 Mar 2026, 22 Jun 2026, 25 Sep 2026 · QQQ/TQQQ ETFs
    Source sentence
    Three market dates are tested, 20 March 2026, 22 June 2026 and 25 September 2026. They are spaced roughly one quarter apart.
  • Monte‑Carlo simulation of LETF dynamics synthetic unknown
    LETF under calibrated LV model
    Source sentence
    13.2.2 Monte Carlo benchmark The Monte Carlo benchmark evolves the LETF as its own process. Under Dupire local volatility the two forward-normalised processes are driven by the same Brownian motion, deSt = p a(t, Xt) eSt dWt, deLt = β p a(t, Xt) eLt dWt, eS0 = eL0 = 1, (71) 15 Table 1: Grid configurations for the forward joint-law solve.

Costs

Unclear - sources disagree The platform classifier and the paper-text extraction give different answers, so no value is asserted.

Platform classifier: Not applicable

Paper text extraction: Not modeled

Source sentence (the authors' words)
Discrete rebalancing, tracking error, jumps, and random fees are outside the present idealisation.

Finds 3

  • The conditional‑Gamma smile tracks the two‑dimensional joint‑law benchmark closely at every leverage tested on SPX and NKY, including the wings where the first‑order strike remapping drifts. pricing-calibration
    Source sentence
    On S&P 500 and Nikkei 225 surfaces the conditional-Gamma smile tracks the benchmark closely at every leverage tested, including the wings where the strike remapping drifts. Against it, the conditional-Gamma smile stays close at every leverage tested on SPX and NKY, including the wings where the strike remapping drifts.
  • The conditional‑Gamma method is 158‑ to 236‑fold faster than the benchmark, producing a full LETF smile in about one second versus 197‑224 seconds for the benchmark. performance
    Source sentence
    A full LETF smile takes 0.85-1.42 seconds against 197-224 seconds for the benchmark, a 158- to 236-fold speed-up. Gamma (s) Shifted Gamma (s) Joint law (s) NKY −2 0.85 1.06 200.97 NKY −1 0.88 1.09 196.82 NKY 2 0.85 1.06 196.61 NKY 3 0.89 1.07 197.38 SPX −2 1.30 1.54 218.69 SPX −1 1.42 1.53 223.94 SPX 2 1.23 1.49 223.19 SPX 3 1.28 1.43 215.30 Both conditional methods run on the order of one second: 0.85 to 1.42
  • Against listed TQQQ quotes, the conditional‑Gamma smile lies inside the market bid‑ask spread at most strikes. pricing-calibration
    Source sentence
    Against listed TQQQ quotes the conditional-Gamma smile lies inside the market bid-ask at most strikes. 13.5.2 Results At strikes with a two-sided quote in Figure 8, the joint-law and conditional-Gamma curves overlap over essentially the whole quoted range, with a separation orders of magnitude smaller than the spread.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 12 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-18 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (18)

  1. Stochastic Volatility Modeling (2017) — link
  2. Leveraged Exchange-Traded Funds: Price Dynamics and Options Valuation (2016)
  3. Implied Volatility of Leveraged ETF Options (2015) — link
  4. Consistent Pricing of Options on Leveraged ETFs (2014) — link
  5. LEVERAGED ETF IMPLIED VOLATILITIES FROM ETF DYNAMICS (2014) — link
  6. Gamma expansion of the Heston stochastic volatility model (2011) — link
  7. Path-Dependence of Leveraged ETF Returns (2010) — link
  8. The Volatility Surface: A Practitioner’s Guide (2006)
and 10 more
  1. Exact Simulation of Stochastic Volatility and Other Affine Jump Diffusion Processes (2006) — link
  2. Pricing with a Smile (1994)
  3. Prices of State-Contingent Claims Implicit in Option Prices (1978) — link
  4. On distributions of certain Wiener functionals (1949) — link
  5. If the approximation does not preserve E [ e Y ] = 1, compute (60), or (63) for the shifted closure, and apply the corresponding substitutions consistently to every strike
  6. Evaluate the conditional price integral (33), or its shifted counterpart (46), for each LETF strike, using stable library implementations of the regularised incomplete gamma function
  7. Conditional-Gamma. The price of Theorem 7.1, with m ( x ) and v ( x ) from the triangular moment system (55) and the martingale normalisation (60) applied before implied-volatility inversion
  8. For the three-moment closure, form λ 3 ( x ) from (19), test the admissibility condition (42), and on the admissible set (43) form α s ( x ), ϑ s ( x ) and ζ ( x ) from (41)
  9. Verify probability conservation and non-negativity of M 0 , and check the unconditional identities M n ( T, x ) dx = E [ I nT ] against an independent calculation or Monte Carlo
  10. Invert the corrected prices to Black implied volatilities

0 of 18 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.