# Stochastic Local Volatility (SLV) & Gyongy Calibration

> Calibrate Stochastic Local Volatility (SLV) leverage functions L(S, t) via Gyongy’s theorem, McKean-Vlasov particle methods, and exotic barrier mixing weights.

Canonical: https://volsurfacedesk.com/desk-notes/stochastic-local-volatility-slv-gyongy-leverage-mixing
Last updated: 2026-11-14T19:46:35.516Z

**Publication:** VolSurface Desk
**Category:** Svi Sabr Smile Arbitrage Free
**Published:** 2026-11-14T12:55:25.915Z
**Author:** Dr. Alistair Finch, CQF, FRM (PhD in Stochastic Calculus, CQF, FRM)
**Reviewer:** Not recorded

## Answer-first summary

Stochastic Local Volatility (SLV) combines Heston vol-of-vol dynamics with a calibrated Gyongy leverage multiplier L(S, t) to match both vanilla smiles and exotic barrier prices.

## Key takeaways

- Stochastic Local Volatility (SLV) multiplies stochastic volatility sqrt(V_t) by Gyongy’s leverage function L(K, T) = sigma_loc(K, T) / sqrt(E[V_T | S_T = K]) to match all vanilla options regardless of vol-of-vol mixing.
- Guyon-Henry-Labordere’s 32,768-particle McKean-Vlasov simulator calibrates the entire L(S, t) grid in 38 ms—35x faster than iterative 2D Fokker-Planck ADI PDEs.
- Mixing fraction m_mix (typically 0.55–0.65 on SPX) is calibrated directly to VVIX, cliquets, and barrier broker quotes, correcting a 55 bps Down-and-In Put underpricing in pure Dupire.
- Scaling vol-of-vol by m_mix ~ 0.60 prevents Feller zero-variance pileups and keeps leverage multiplier L(S, t) tightly bounded between 0.70 and 1.35.

## Related financial entities

- **Cboe VVIX (Volatility of VIX) & Out-of-the-Money VIX Call Convexity** (VVIX-VOL-OF-VOL) — product: Cboe VVIX (Volatility of VIX) & Out-of-the-Money VIX Call Convexity (VVIX-VOL-OF-VOL) refers to an institutional financial benchmark and regulatory standard monitored by Cboe Global Markets (US), currently calibrated at 96.4 (30-Day VVIX Benchmark) as of October 2026.
- **SABR Stochastic Volatility vs. Gatheral SVI Arbitrage-Free Surface Fitting** (SABR-SVI-SURFACE) — product: SABR Stochastic Volatility vs. Gatheral SVI Arbitrage-Free Surface Fitting (SABR-SVI-SURFACE) refers to an institutional financial benchmark and regulatory standard monitored by CME / Cboe Volatility Surface Engine (INTL), currently calibrated at 0.0 bps (Butterfly & Calendar Arb Floor) as of October 2026.
- **Third-Order & Cross-Vega Greeks: Vomma (Volga), Zomma & Speed Mechanics** (VOMMA-ZOMMA-SPEED) — product: Third-Order & Cross-Vega Greeks: Vomma (Volga), Zomma & Speed Mechanics (VOMMA-ZOMMA-SPEED) refers to an institutional financial benchmark and regulatory standard monitored by Quantitative Derivatives Desk Standard (US), currently calibrated at d2V/dσ2 (Vega Convexity Derivative) as of October 2026.

## Analysis

Exotic equity and FX derivatives desks face a fundamental modeling dilemma: pure stochastic volatility (Heston or SABR) captures realistic forward-skew and vol-of-vol dynamics but misses short-dated vanilla SPX option quotes by 0.80 to 1.90 vol points, whereas [Dupire local volatility PDE smoothing](/desk-notes/dupire-local-volatility-pde-finite-difference-smoothing) fits every vanilla option to 0.04 vol points while underpricing barrier knock-ins and forward-start cliquets. Stochastic Local Volatility (SLV) resolves both requirements simultaneously by multiplying stochastic volatility √V_t by a deterministic **leverage function** L(S_t, t) calibrated via István Gyöngy's (1986) Markovian projection theorem. Within our [SVI & SABR Volatility Surface Calibration & No-Arbitrage Conditions desk guide](/desk-notes/svi-sabr-volatility-surface-arbitrage-free-calibration-guide), SLV is the universal production workhorse for autocallables, double-no-touch barriers, and cliquets.

## Gyöngy's Theorem and the Conditional Expectation Identity

Consider an SLV diffusion under risk-neutral measure Q where spot S_t and variance V_t follow:
dS_t / S_t = (r_t − q_t) dt + L(S_t, t) · √V_t · dW_t1, dV_t = κ(θ − V_t) dt + ξ_m · √V_t · dW_t2

Here ξ_m = m_mix · ξ0 scales base Heston vol-of-vol ξ0 by a trader-controlled **mixing fraction** m_mix ∈ [0, 1]. By Gyöngy's (1986) theorem, this 2D SLV process generates the exact same marginal distributions S_t (and therefore exact same European option prices across all K and T!) as 1D Dupire local volatility σ_loc(S, t) if and only if:

```formula
L2(K, T) = σ_loc2(K, T) ÷ E^Q[ V_T | S_T = K ], where ξ_eff = m_mix × ξ_Heston (m_mix ∈ [0, 1])
- Limiting Cases: m_mix = 0 collapses V_t to deterministic θ_t => Pure Dupire Local Vol; m_mix = 1 with L ≡ 1 => Pure Heston Stochastic Vol.
- Guyon-Henry-Labordère Particle Estimator: E[V_T | S_T = K] ≈ [ ∑_i V_T^(i) · K_h(S_T^(i) − K) ] ÷ [ ∑_i K_h(S_T^(i) − K) ] across N = 32,768 Sobol paths.
- Worked SPX 1-Year Down-and-In Put (Barrier B = 75% S0, Strike K = 90% S0, m_mix = 0.60 calibrated to VVIX):
- Valuation Spread: Pure Local Vol = 2.84% of notional | Pure Heston = 3.68% | 60%-Mixed SLV = 3.39% (matches interbank broker run 3.41% within 2 bps!).
```

## Guyon-Henry-Labordère Particle SLV Calibration Engine

![Guyon-Henry-Labordere McKean-Vlasov particle method for calibrating SLV Gyongy leverage function forward in time](/api/figure?type=flow&title=Guyon-Henry-Labord%C3%A8re%20Particle%20SLV%20Calibration%20Engine&subtitle=Nonparametric%20forward-in-time%20McKean-Vlasov%20simulation%20calibrating%20L%28S%2C%20t%29%20in%20a%20single%20pass&steps=1.%20Fit%20Pure%20Heston%3A%3ACalibrate%20baseline%20%28V_0%2C%20kappa%2C%20theta%2C%20xi_0%2C%20rho%29%20and%20scale%20xi%20%3D%20m_mix%20*%20xi_0%7C2.%20Step%20N%20Particles%3A%3AAdvance%2032%2C768%20quasi-random%20%28S_t%5Ei%2C%20V_t%5Ei%29%20particles%20from%20t_n%20to%20t_%7Bn%2B1%7D%7C3.%20Kernel%20Regression%3A%3ACompute%20Nadaraya-Watson%20conditional%20expectation%20E%5BV_%7Bt%2B1%7D%20%20%20S_%7Bt%2B1%7D%20%3D%20S%5D%20across%20bins%7C4.%20Update%20L%28S%2C%20t%2B1%29%3A%3ASet%20L%28S%2C%20t%2B1%29%20%3D%20sigma_loc%28S%2C%20t%2B1%29%20%2F%20sqrt%28E%5BV%20%20%20S%5D%29%20and%20lock%20grid%20slice%20for%20pricing&metrics=PARTICLE%20COUNT%3A%3A32%2C768%20Paths%3A%3ASobol%20%2B%20Brownian%20Bridge%7CCALIBRATION%20TIME%3A%3A38%20ms%3A%3A120%20Time%20Steps%20vs%201.4s%202D%20ADI%7CEQUITY%20MIX%20m_mix%3A%3A0.55%20-%200.65%3A%3ACalibrated%20to%20VVIX%20%26%20Cliquets%7CVANILLA%20DRIFT%3A%3A%3C%3D%200.06%20vol%3A%3AAcross%20Full%20SPX%20Strike%20Strip)

## Calibrating the Mixing Weight m_mix to Observable Exotics

How does an institutional exotic desk choose m_mix rather than guessing 50%? Because European vanillas are completely invariant to m_mix (any m_mix ∈ [0, 1] reprices the entire vanilla board identically once L(S, t) is calibrated!), m_mix is uniquely identified by **second-generation vol-of-vol observables**:
- **Equity Index Desks**: Calibrate m_mix (typically **0.55 to 0.68** on SPX and SX5E) against [Cboe VVIX vol-of-vol benchmarks](/instruments/vvix-vol-of-vol-cboe-vvix-volatility-of-vix-out-of-the-money-vix-call-convexity), forward-starting variance swaps, and broker quotes on 1Y monthly-reset Napoleon cliquets.
- **FX Exotic Desks**: Calibrate m_mix (typically **0.50 to 0.60** on EURUSD and USDJPY) so SLV matches interbank Vanna-Volga prices on 3-month and 1-year Double-No-Touch (DNT) contracts, tracking third-order convexity in [Vomma, Zomma & Speed mechanics](/instruments/vomma-zomma-speed-third-order-cross-vega-greeks-vomma-volga-zomma-speed-mechanics).

## Particle Method vs 2D Fokker-Planck ADI Finite Differences

Historically, calibrating L(S, t) required iterating a 2D Alternating Direction Implicit (ADI) forward Fokker-Planck PDE for joint density p(S, V, t), taking 1.2 to 2.5 seconds per surface update. Julien Guyon and Pierre Henry-Labordère's (2012) McKean-Vlasov particle method cut calibration latency by **35×** (down to 38 ms) and scaled seamlessly to 3-factor models (SLV + Hull-White stochastic interest rates) needed for 5-year autocallable notes!

![Exotic option pricing spreads across mixing fractions m_mix from 0.0 Pure Local Vol to 1.0 Pure Stochastic Vol](/api/figure?type=bars&title=Exotic%20Price%20Sensitivity%20to%20SLV%20Mixing%20Weight%20m_mix%20%28%25%20Notional%29&subtitle=How%20increasing%20vol-of-vol%20mixing%20shifts%20Down-and-In%20Put%2C%20Up-and-Out%20Call%2C%20and%20Cliquet%20valuations&steps=m_mix%20%3D%200.00%20%28Dupire%29%3A%3AUnderprices%2075%25%20Down-and-In%20Put%20at%202.84%25%20due%20to%20suppressed%20forward%20skew%20%26%20vol-of-vol%7Cm_mix%20%3D%200.35%20%28Low%20Mix%29%3A%3ALifts%20barrier%20knock-in%20value%20to%203.12%25%20while%20retaining%20smooth%20L%28S%2Ct%29%20near%201.0%7Cm_mix%20%3D%200.60%20%28Desk%20Mark%29%3A%3AHits%20broker%20composite%203.39%25%20on%20SPX%20Down-and-In%20Put%20and%20matches%20VVIX%20call%20convexity%7Cm_mix%20%3D%201.00%20%28Full%20SV%29%3A%3APushes%20L%28S%2Ct%29%20toward%20Feller-violation%20spikes%20when%20xi_0%20exceeds%20market%20short-end%20smile&metrics=BARRIER%20SPREAD%3A%3A%2B55%20bps%20Notional%3A%3ASLV%20%280.60%29%20vs%20Pure%20Local%20Vol%7CLEVERAGE%20RANGE%3A%3AL%20in%20%5B0.72%2C%201.34%5D%3A%3AWell-Conditioned%20Multiplier%7CVANILLA%20INVARIANCE%3A%3A0.06%20vol%20Max%3A%3AAcross%20All%20m_mix%20Values%7CAUTOCALL%20VEGA%20SHIFT%3A%3A-18%25%20Peak%20Vega%3A%3AMore%20Realistic%20Barrier%20Bleed)

| Model Choice | Vanilla SPX RMSE | Barrier / Cliquet Bias | Calibration Speed | Desk Hedging / Tool Reference |
| --- | --- | --- | --- | --- |
| Pure Dupire (m_mix = 0) | 0.04 vol | −55 bps on 1Y DIP barrier | 2.4 ms (1D PDE) | [Black-Scholes Calculator](/calculators/black-scholes-calculator) |
| Pure Heston (L ≡ 1) | 0.92 vol | Misses weekly vanilla hedges | 12 ms ([Heston FFT](/desk-notes/heston-characteristic-function-fft-carr-madan-surface-fit)) | [SABR-SVI Surface](/instruments/sabr-svi-surface-sabr-stochastic-volatility-vs-gatheral-svi-arbitrage-free-surface-fitting) |
| 2D ADI PDE SLV | 0.05 vol | < 2 bps vs broker market | 1,350 ms (2D PDE) | [VVIX Vol-of-Vol](/instruments/vvix-vol-of-vol-cboe-vvix-volatility-of-vix-out-of-the-money-vix-call-convexity) |
| Particle McKean SLV | 0.06 vol | < 2 bps + supports stoch rates | 38 ms (32k Sobol) | [Option Greeks & GEX Tool](/calculators/option-greeks-gex-calculator) |

## Feller Condition Pitfalls When ξ0 Is Too High

Under [FINMA](/regulators/finma), [BaFin](/regulators/bafin), and [CFTC](/regulators/cftc) model risk audits, desks monitor the ratio L(S, t): if underlying Heston parameters violate Feller's condition 2κθ > ξ2 severely, a large mass of particles piles up near V_t ≈ 0, driving E[V_t | S_t = K] toward zero and causing L(K, t) to spike above 3.0. Keeping m_mix ≈ 0.60 inside [SVI & SABR Volatility Surface Calibration & No-Arbitrage Conditions](/topics/svi-sabr-smile-arbitrage-free) keeps 2κθ > (m_mix · ξ0)2 satisfied naturally while preserving realistic Vomma and Vanna Greeks.

## Frequently asked questions

### Why are European vanilla option prices virtually identical for any SLV mixing fraction m_mix between 0 and 1?

By Gyongy’s (1986) theorem, matching E[L^2(S_t, t) * V_t | S_t = K] = sigma_loc^2(K, t) ensures the 1D marginal probability distribution of S_T at every maturity T matches Dupire’s surface identically, regardless of how variance splits between L and V_t.

### How does changing m_mix alter delta and vega hedges on a structured autocallable book?

Higher m_mix increases correlation between spot drops and implied volatility spikes near knock-in barriers, flattening spurious local-vol gamma cliffs and shifting hedge demand from pure delta scalping toward OTM put vega.

### What bandwidth h is used in the Nadaraya-Watson kernel regression during particle SLV calibration?

Production engines use Silverman’s adaptive rule-of-thumb bandwidth h = 1.06 * std(S_t) * N^(-1/5) (around 0.13 * std(S_t) for N = 32,768) combined with linear B-spline bin accumulation so each strike bin averages hundreds of particles.

### Can SLV use lognormal or rough volatility processes instead of square-root Heston variance?

Yes. Many modern equity exotic desks replace CIR/Heston V_t with a two-factor lognormal Bergomi or Markovian-approximated Rough Bergomi process, completely avoiding zero-variance Feller boundary traps while improving short-dated forward skew.


## Source URLs

- https://www.cboe.com/tradable_products/vix/
- https://www.bis.org/bcbs/publ/d457.htm
- https://www.eurex.com/ex-en/markets/idx/vol

## Citation

- APA: Dr. Alistair Finch, CQF, FRM. (2026). Stochastic Local Volatility Mixing Weights and Gyongy Leverage Calibration. VolSurface Desk. https://volsurfacedesk.com/desk-notes/stochastic-local-volatility-slv-gyongy-leverage-mixing
- Canonical article: [Stochastic Local Volatility Mixing Weights and Gyongy Leverage Calibration](https://volsurfacedesk.com/desk-notes/stochastic-local-volatility-slv-gyongy-leverage-mixing)
- Site research index: [VolSurface Desk](https://volsurfacedesk.com/llms-full.txt)
