# SVI Volatility Surface Calibration, 0DTE Gamma/Vanna & Dispersion (2026)

> Quantitative volatility surface desk note: SVI/SSVI butterfly & calendar arbitrage conditions, SPX 0DTE dealer gamma/vanna flows, and dispersion correlation.

Canonical: https://volsurfacedesk.com/desk-notes/svi-volatility-surface-arbitrage-free-calibration-0dte-gamma-vanna-dispersion-2026
Last updated: 2026-10-27T03:01:16.692Z

**Publication:** VolSurface Desk
**Category:** Volatility Surface & Derivatives Desk
**Published:** 2026-10-22T15:14:00.805Z
**Author:** Dr. Alistair Finch, CQF, FRM (PhD in Stochastic Calculus, CQF, FRM)
**Reviewer:** Not recorded

## Answer-first summary

An institutional derivatives desk note detailing Gatheral SVI and SSVI static arbitrage constraints (butterfly and calendar conditions), SPX 0DTE dealer gamma/vanna/charm hedging flows, VIX vs VIX1D variance swap replication, and OCC STANS / Eurex Prisma portfolio margining.

## Key takeaways

- SPX 0DTE options account for 51.4% of total S&P 500 options volume ($890B daily notional) in 2026, requiring volatility desks to replace calendar-day time-to-expiry with an activity-weighted intraday variance clock (tau_eff) to prevent wing oscillation.
- Static arbitrage freedom across an SVI / SSVI implied volatility surface requires simultaneous satisfaction of Roger Lee's asymptotic wing slope bound (b(1 + |rho|) <= 2), Durrleman's non-negative butterfly density condition (g(k) >= 0), and monotonic total variance across expiries (dw/dT >= 0).
- Intraday dealer delta-hedging in CME E-mini S&P 500 futures is governed by the coupled interaction of Spot Gamma, Vanna (dDelta/dVol), and Charm (-dDelta/dTau), generating $14.2B to $19.8B in deterministic afternoon rebalancing flows between 14:00 and 16:00 ET.
- VIX1D exhibits a realized vol-of-vol 3.45x higher than 30-day VIX due to two-zero-bid strike truncation on the 0DTE strip and cubic jump-diffusion discretization bias (-1/3 E[J^3]) during sudden intraday pullbacks.
- Clearing index dispersion and 0DTE books through OCC STANS (10,000 Monte Carlo copula scenarios) and CME cross-margining compresses initial margin requirements by 68% to 78% compared to strategy-based Reg-T, reducing capital usage from $21.56M to $4.85M per $100M notional.

## Analysis

## 1. Executive Summary & The 2026 Equity Volatility Surface Microstructure

By the first quarter of 2026, the topology of institutional equity and foreign exchange (FX) implied volatility surfaces has undergone a structural bifurcation driven by ultra-short-dated convexity trading and systematic correlation harvesting. Across US listed index options cleared through the Options Clearing Corporation (OCC) under the joint supervision of the [Commodity Futures Trading Commission (CFTC)](/regulators/cftc) and the [Securities and Exchange Commission (SEC)](/regulators/sec), Zero-Days-to-Expiration (0DTE) contracts on the S&P 500 Index (SPX) now account for **51.4% of total daily SPX options volume**, averaging **1.68 million contracts per session** ($890 billion in daily notional exposure) [[1]](https://www.cboe.com/insights/). Concurrently, institutional quantitative volatility desks and multi-manager hedge pods have expanded systematic index-versus-constituent dispersion books to record notional footprints, compressing the spread between the Cboe 3-Month Implied Correlation Index (COR3M) and realized pairwise S&P 500 correlation to **14.2 correlation points** during low-macro-variance regimes while triggering violent convexity squeezes during single-stock earnings shocks [[2]](https://cdn.cboe.com/resources/vix/vixwhite.pdf).

For institutional derivatives market makers, statistical arbitrage desks, and risk managers operating across [regulated institutional derivatives venues](/brokers) overseen by the [National Futures Association (NFA)](/regulators/nfa), the [European Securities and Markets Authority (ESMA)](/regulators/esma), the [UK Financial Conduct Authority (FCA)](/regulators/fca), and Germany's [Federal Financial Supervisory Authority (BaFin)](/regulators/bafin), these twin structural shifts expose fatal failure modes in legacy volatility surface fitting and Greek risk aggregation:

1. **Sub-Diurnal Time-Decay Breakdown:** Standard calendar-day Black-Scholes time parameterization ($\tau = T - t$ measured in 252 trading days or 365 calendar days) fails catastrophically inside the final 390 minutes of trading before the 16:00 ET SPX PM cash settlement. Without intraday variance-clock weighting that accounts for the opening auction burst, the midday liquidity trough, and the 15:50 ET Market-on-Close (MOC) imbalance release, raw implied volatility quotes exhibit artificial wing oscillations that violate static butterfly arbitrage bounds.
2. **Coupled Higher-Order Greek Feedback Loops:** Dealer delta-hedging flows in E-mini S&P 500 futures (ES) are no longer explainable solely through static spot-gamma ($\Gamma = \partial^2 V / \partial S^2$) profiles. Intraday interactions between **Vanna** ($\partial \Delta / \partial \sigma = \partial \mathcal{V} / \partial S$) and **Charm** ($-\partial \Delta / \partial \tau$, or delta bleed) create deterministic afternoon futures buying or selling waves that surpass $18.5 billion in directional notional over the final 90 minutes of the cash session [[1]](https://www.cboe.com/insights/) [[5]](https://www.cmegroup.com/clearing/risk-management/span-overview.html).
3. **Wing Extrapolation & Variance Swap Replication Bias:** Fair-value pricing of variance swaps, corridor variance swaps, and the Cboe Volatility Index (VIX and VIX1D) requires integrating out-of-the-money (OTM) option prices across a continuum of strikes $K \in (0, \infty)$ [[2]](https://cdn.cboe.com/resources/vix/vixwhite.pdf). Any violation of Roger Lee's moment formula ($|dw/dk| \le 2$) or Durrleman's non-negative local density condition ($g(k) \ge 0$) in the deep 1-delta and 5-delta put wings injects phantom convexity into synthetic variance strips and corrupts portfolio margin calculations under OCC STANS and Eurex Prisma [[3]](https://www.theocc.com/risk-management/margin-methodology) [[8]](https://www.bis.org/publ/qtrpdf/r_qt2309.htm).

![Figure 1: SVI / SSVI Arbitrage-Free Implied Volatility Surface Calibration & 0DTE Dealer Gamma Hedging Loop (2026)](/api/figure?title=SVI%20Implied%20Volatility%20Surface%20%26%200DTE%20Dealer%20Hedging%20Engine&subtitle=Gatheral%20SSVI%20No-Arbitrage%20Calibration%2C%20Vanna%2FCharm%20Rebalancing%20%26%20OCC%20STANS%20Margin&badge=EXHIBIT%2001&steps=Raw%20OPRA%20Quote%20Filtering|SSVI%20Butterfly%20%26%20Calendar%20Gate|Dealer%20Gamma%2FVanna%2FCharm%20Vector|OCC%20STANS%2010k-Scenario%20VaR&metrics=SPX%200DTE%20Volume%20Share:51.4%25%20of%20Total%20SPX%20ADV|Roger%20Lee%20Wing%20Slope%20Bound:%7Cdw%2Fdk%7C%20%3C%3D%202.0%20Asymptote|OCC%20STANS%20Margin%20Offset:68%25%20to%2078%25%20vs%20Reg-T|Dispersion%20Implied%20Corr%20Spread:14.2%20pts%20%28COR3M%20vs%20Realized%29&source=Cboe%20Global%20Markets%20%2F%20OCC%20STANS%20%2F%20CFTC%202026)

This desk note establishes the production mathematical and operational blueprint for arbitrage-free Stochastic Volatility Inspired (SVI) and Surface SVI (SSVI) calibration, sub-minute 0DTE dealer Greek decomposition, VIX/VIX1D replication diagnostics, and correlation dispersion structuring under 2026 clearinghouse margin frameworks.

---

## 2. Gatheral SVI & SSVI Parameterization: Static Arbitrage Mathematics

### 2.1 Raw SVI, Natural SVI, and Jump-Wing (SVI-JW) Formulations

Originally introduced by Jim Gatheral (2004) and formalized alongside Antoine Jacquier (2014), the **Stochastic Volatility Inspired (SVI)** parameterization models total implied variance $w(k, T) = \sigma_{\text{BS}}^2(k, T) \cdot T$ as a function of log-forward moneyness $k = \ln(K / F_T)$, where $K$ denotes the strike price and $F_T$ denotes the implied forward price at maturity $T$ extracted via put-call parity from at-the-money (ATM) box spreads [[4]](https://www.sec.gov/files/staff-report-equity-options-market-structure.pdf).

In its **Raw SVI** representation for a fixed expiration slice $T$, total implied variance is governed by five structural parameters $\chi_R = \{a, b, \rho, m, \sigma\}$:

```formula
w(k) = a + b \left\{ \rho (k - m) + \sqrt{(k - m)^2 + \sigma^2} \right\}, \quad g(k) = \left(1 - \frac{k w'(k)}{2 w(k)}\right)^2 - \frac{w'(k)^2}{4}\left(\frac{1}{w(k)} + \frac{1}{4}\right) + \frac{w''(k)}{2} \ge 0
- w(k) = \sigma_{\text{BS}}^2(k, T) \cdot T: Total implied Black-Scholes variance at log-moneyness k = \ln(K / F_T)
- a \in \mathbb{R}, b \ge 0, |\rho| < 1, m \in \mathbb{R}, \sigma > 0: Raw SVI level, slope, correlation skew, translation, and ATM curvature parameters
- g(k) \ge 0: Durrleman (2010) risk-neutral probability density positivity condition eliminating static butterfly arbitrage
```

Each parameter in $\chi_R$ admits an immediate geometric interpretation on the total variance smile:
- $a \in \mathbb{R}$ controls the vertical translation (overall variance level), constrained such that $\inf_{k \in \mathbb{R}} w(k) = a + b\sigma\sqrt{1 - \rho^2} \ge 0$.
- $b \ge 0$ governs the symmetric angle between the left (put) and right (call) asymptotes.
- $\rho \in (-1, 1)$ determines the orientation and asymmetry of the skew; negative values ($\rho \approx -0.72$ in SPX) steepen the downside put wing relative to the upside call wing.
- $m \in \mathbb{R}$ translates the vertex of the hyperbola horizontally along the log-moneyness axis.
- $\sigma > 0$ controls the smoothness and radius of curvature at the vertex $k = m$.

Because Raw SVI parameters are non-orthogonal—shifting $m$ or $\sigma$ simultaneously alters ATM implied volatility and wing slopes—institutional trading desks prefer **Jump-Wing SVI (SVI-JW)** for trader-facing surface controls. Defined for a specific expiry $T$, the five SVI-JW coordinates $\chi_{\text{JW}} = \{v_t, \psi_t, p_t, c_t, \tilde{v}_t\}$ map bijectively to Raw SVI via:
- **ATM Variance ($v_t$):** $v_t = w(0)/T = \frac{1}{T}\left(a + b\left(-\rho m + \sqrt{m^2 + \sigma^2}\right)\right)$
- **ATM Skew ($\psi_t$):** $\psi_t = \left.\frac{\partial \sqrt{w(k)/T}}{\partial k}\right|_{k=0} = \frac{1}{\sqrt{w(0)}} \frac{b}{2} \left(-\frac{m}{\sqrt{m^2 + \sigma^2}} + \rho\right)$
- **Left (Put) Wing Slope ($p_t$):** $p_t = \frac{b}{\sqrt{w(0)}}(1 - \rho)$
- **Right (Call) Wing Slope ($c_t$):** $c_t = \frac{b}{\sqrt{w(0)}}(1 + \rho)$
- **Minimum Variance ($\tilde{v}_t$):** $\tilde{v}_t = \frac{1}{T}\left(a + b\sigma\sqrt{1 - \rho^2}\right)$

### 2.1.1 Cross-Asset Mapping: FX 10-Delta / 25-Delta Risk Reversals & Butterflies to SVI-JW

While equity index options quote directly in strike space $K$ on centralized OPRA feeds, G10 and EM foreign exchange (FX) volatility surfaces quote over-the-counter in sticky-delta coordinates: At-The-Money Delta-Neutral Straddle ($
\sigma_{\text{ATM}}$), 25-Delta and 10-Delta Risk Reversals ($\text{RR}_{25\Delta} = \sigma_{25\Delta C} - \sigma_{25\Delta P}$), and 25-Delta and 10-Delta Market Strangles / Butterflies ($\text{BF}_{25\Delta} = \frac{1}{2}(\sigma_{25\Delta C} + \sigma_{25\Delta P}) - \sigma_{\text{ATM}}$). Whereas legacy Vanna-Volga three-point interpolation explodes beyond the 10-delta wings and violates Durrleman's $g(k) \ge 0$ condition whenever $|
\text{RR}_{25\Delta}| / \sigma_{\text{ATM}} > 0.22$, institutional cross-asset volatility desks map sticky-delta FX pillars directly into SVI-JW coordinates by solving the implicit Black-Scholes delta inversion $k(\Delta, w(k)) = \left(\frac{1}{2} + \text{sign}(\Delta)\right)w(k) - \sqrt{w(k)}\Phi^{-1}(\Delta e^{r_f T})$ via three-step Newton-Raphson iteration. This unifies equity index and FX volatility surfaces under a single arbitrage-free SSVI manifold, enabling consistent cross-asset correlation and quanto-hedging across EUR/USD, USD/JPY, and SPX/SX5E books.

### 2.2 Static Arbitrage Conditions: Roger Lee Moment Bounds, Butterfly Density, and SSVI Calendar Constraints

A parameterized volatility surface $w(k, T)$ is free of **static arbitrage** if and only if it is simultaneously free of **calendar spread arbitrage** and **butterfly spread arbitrage** across all $(k, T) \in \mathbb{R} \times (0, \infty)$ [[7]](https://www.esma.europa.eu/publications-and-data/all-publications).

1. **Roger Lee's (2004) Wing Slope Bounds:** As $k \to \pm\infty$, the Raw SVI specification converges to linear asymptotes with slopes $\lim_{k \to -\infty} dw/dk = b(\rho - 1)$ and $\lim_{k \to +\infty} dw/dk = b(\rho + 1)$. Roger Lee's moment formula proves that for the underlying asset price $S_T$ to possess a finite first moment $\mathbb{E}^\mathbb{Q}[S_T] < \infty$ and avoid exploding call prices, the asymptotic slopes cannot exceed 2 in absolute value:
 $$b(1 + |\rho|) \le 2$$
2. **Butterfly Arbitrage & Durrleman's Condition ($g(k) \ge 0$):** By Breeden-Litzenberger (1978), the risk-neutral probability density $p(K) = e^{rT} \frac{\partial^2 C(K,T)}{\partial K^2}$ is non-negative if and only if the function $g(k)$ defined in the formula block above satisfies $g(k) \ge 0$ for all $k \in \mathbb{R}$, alongside $b(1 + |\rho|) \le 2$. A classic sufficient condition established by Gatheral and Jacquier (2014) guarantees $g(k) \ge 0$ globally on a slice if $b(1 + |\rho|) \le 4$ and $b^2(1 + |\rho|)\sigma^{-1}$ remains bounded, though numerical optimization in production enforces $g(k_i) \ge \epsilon = 10^{-7}$ across a dense 400-node Gauss-Lobatto grid spanning $k \in [-3.5, +2.0]$.
3. **Calendar Spread Arbitrage & Surface SVI (SSVI):** For two expirations $T_1 < T_2$, calendar arbitrage absence requires that total variance is non-decreasing along lines of constant log-forward moneyness:
 $$\frac{\partial w(k, T)}{\partial T} \ge 0 \quad \forall k \in \mathbb{R}, T > 0$$
 To enforce calendar and butterfly arbitrage freedom jointly across the entire $(k, T)$ manifold, Gatheral and Jacquier introduced **Surface SVI (SSVI)**, parameterizing the surface via ATM total variance $\theta_t = w(0, T)$ and a curvature function $\varphi(\theta_t)$:
 $$w(k, \theta_t) = \frac{\theta_t}{2}\left(1 + \rho \varphi(\theta_t) k + \sqrt{\left(\varphi(\theta_t) k + \rho\right)^2 + (1 - \rho^2)}\right)$$
 Under the power-law kernel $\varphi(\theta) = \eta \theta^{-\gamma} (1 + \beta \theta)^{\gamma - 1}$ with $\gamma \in (0, 1/2]$, SSVI is guaranteed free of calendar spread arbitrage whenever $\partial_t \theta_t \ge 0$ and $0 \le \partial_\theta (\theta \varphi(\theta)) \le \frac{1}{\rho^2}\left(1 + \sqrt{1 - \rho^2}\right)\varphi(\theta)$, and free of butterfly arbitrage whenever $\theta \varphi(\theta)(1 + |\rho|) < 4$ and $\theta \varphi(\theta)^2 (1 + |\rho|) \le 4$.

### 2.3 Real-Time Production Calibration Pipeline on OPRA Feeds

```flow
Stage 1: OPRA Tick Filtering & Micro-Price Mid Extraction | Ingest sub-millisecond OPRA multicast feeds; discard crossed/locked quotes, enforce bid-ask spread gates (< 18% of vega-weighted mid), and compute volume-weighted micro-price mids across all strikes.
Stage 2: Put-Call Parity Box-Spread Forward & Borrow Curve Fit | Solve simultaneously for implied forward F_T and synthetic dividend/borrow yield q_T per expiry using ATM box spreads within |k| <= 0.025 to eliminate early-exercise and dividend distortion.
Stage 3: SSVI Global Seed + L-BFGS-B Slice Polishing with Durrleman Penalty | Fit global SSVI power-law skeleton across all tenors, then run constrained L-BFGS-B per slice minimizing inverse-spread vega-weighted squared error subject to g(k) >= 1e-7 and dw/dt >= 0.
Stage 4: Sub-Millisecond Greek Vectorization & STANS Scenario Export | Broadcast arbitrage-free SVI-JW coordinates, local volatility surfaces (Dupire), and 2nd/3rd-order Greek sensitivities (Gamma, Vanna, Charm, Vomma, Speed) to automated hedging and margin engines.
```

Table 1 benchmarks calibrated Raw SVI and SSVI-JW parameters across six canonical SPX maturity slices as observed during the Q1 2026 US equity derivatives cycle, demonstrating the extreme steepening of ATM skew ($\psi_t$) and wing slope ($b$) in sub-24-hour expirations.

| SPX Maturity Tenor | ATM Implied Vol (%) | ATM Skew Psi (pts/10% k) | SVI Correlation Rho (ratio) | Wing Slope b (rad) | Min Durrleman Density g_min (bps) | Vega-Weighted Fit RMSE (bps) |
| :--- | :---: | :---: | :---: | :---: | :---: | :---: |
| 0DTE (Intraday 2h Left) | 18.45% | -14.80 pts | -0.84 | 0.42 | 12.4 bps | 4.8 bps |
| 1DTE (Overnight Slice) | 16.90% | -11.20 pts | -0.81 | 0.38 | 18.5 bps | 3.6 bps |
| 7DTE (1-Week Weekly) | 15.60% | -8.65 pts | -0.78 | 0.33 | 24.0 bps | 2.9 bps |
| 30DTE (Monthly AM VIX) | 15.10% | -6.40 pts | -0.75 | 0.28 | 31.2 bps | 2.1 bps |
| 90DTE (Quarterly IMM) | 15.85% | -5.10 pts | -0.72 | 0.24 | 38.6 bps | 1.8 bps |
| 180DTE (6-Month Semiannual) | 16.40% | -4.35 pts | -0.69 | 0.21 | 44.0 bps | 1.6 bps |

---

## 3. 0DTE Intraday Options Microstructure: Gamma, Vanna, Charm & Speed Mechanics

### 3.1 Intraday Variance Clock ($\tau_{\text{eff}}$) vs. Calendar Time

When calibrating an SPX 0DTE slice between 09:30 ET and 16:00 ET, measuring time-to-expiration via uniform wall-clock seconds introduces severe mispricing across mid-session lulls and macro release windows (such as 08:30 ET CPI/NFP prints or 14:00 ET FOMC announcements) [[1]](https://www.cboe.com/insights/) [[6]](https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm). Institutional volatility desks replace linear time $T - t$ with an **Activity-Weighted Intraday Variance Clock**:
$$\tau_{\text{eff}}(t, T) = \int_t^T \lambda(u) \, du + \sum_{j: t < t_j \le T} \mathcal{J}_j$$
where $\lambda(u)$ is a piecewise-smooth diurnal variance intensity spline calibrated to high-frequency 5-second E-mini S&P 500 realized variance (typically assigning **28% of regular trading hours variance** to 09:30–10:30 ET, **34%** to 10:30–14:30 ET, and **38%** to 14:30–16:00 ET), and $\mathcal{J}_j$ represents discrete jump-variance weights assigned to scheduled macroeconomic releases.

![Figure 2: SPX 0DTE Intraday Greek Decay Trajectory & Index Dispersion Convexity Architecture](/api/figure?title=0DTE%20Greek%20Decay%20Mechanics%20%26%20Index%20Dispersion%20Architecture&subtitle=Intraday%20Gamma%2FVanna%2FCharm%20Rebalancing%20Vector%20vs%20Single-Stock%20Variance%20Replication&badge=EXHIBIT%2002&steps=09%3A30%20ET%200DTE%20Customer%20Overwrite|13%3A00%20ET%20Gamma%20Pin%20%26%20Charm%20Bleed|15%3A00%20ET%20Vanna-Gamma%20Cash%20Squeeze|16%3A00%20ET%20MOC%20%26%20Dispersion%20Reset&metrics=Peak%20ATM%200DTE%20Gamma%20%2815%3A30%20ET%29:8.4x%20vs%2009%3A30%20ET%20Open|15%3A00-16%3A00%20ET%20Charm%20Flow:%2414.2B%20to%20%2419.8B%20ES%20Notional|VIX%20vs%20VIX1D%20Vol-of-Vol%20Ratio:3.45x%20Intraday%20Realized%20Beta|Top-50%20Dispersion%20Vega%20Efficiency:86.5%25%20Variance%20Capture&source=Desk%20Quantitative%20Research%20%2F%20Cboe%20OPRA%20%2F%20CME%202026)

### 3.2 Second- and Third-Order Greeks Governing Dealer Futures Rebalancing

Market makers warehousing net customer order flow in SPX 0DTE options manage delta exposure dynamically using CME E-mini S&P 500 futures (ES) and Micro E-mini futures (MES) [[5]](https://www.cmegroup.com/clearing/risk-management/span-overview.html). By applying a second-order Taylor expansion to the dealer portfolio delta $\Delta_{\text{port}}(S, \sigma, \tau_{\text{eff}})$ over an infinitesimal interval $(t, t + dt)$, the instantaneous hedging demand $dH_t = -d\Delta_{\text{port}}$ decomposes into four distinct structural channels:

$$d\Delta_{\text{port}} = \underbrace{\Gamma_{\text{port}} \, dS_t}_{\text{Spot Gamma Flow}} + \underbrace{\text{Speed}_{\text{port}} \, \frac{1}{2}(dS_t)^2}_{\text{Gamma Convexity}} + \underbrace{\text{Vanna}_{\text{port}} \, d\sigma_t}_{\text{Skew/Vol Repricing Flow}} - \underbrace{\text{Charm}_{\text{port}} \, d\tau_{\text{eff}}}_{\text{Diurnal Delta Bleed}}$$

1. **Gamma ($\Gamma = \frac{\partial^2 V}{\partial S^2} = \frac{\phi(d_1)}{S \sigma \sqrt{\tau_{\text{eff}}}}$):** Scales at the ATM strike as $\tau_{\text{eff}}^{-1/2}$. Between 09:30 ET ($\tau_{\text{eff}} = 1$ session) and 15:30 ET ($\tau_{\text{eff}} \approx 0.08$ session), ATM 0DTE gamma amplifies by **$3.5\times$ to $8.4\times$** depending on realized variance consumption. When dealers are **net long 0DTE gamma** (resulting from retail and systematic call/put overwriting), they sell ES futures into rallies and buy ES futures on dips, suppressing intraday realized volatility and creating high-probability "gamma pins" around heavy open-interest strikes.
2. **Vanna ($\text{Vanna} = \frac{\partial \Delta}{\partial \sigma} = \frac{\partial \mathcal{V}}{\partial S} = -\frac{\phi(d_1) d_2}{\sigma}$):** Measures the sensitivity of delta to changes in implied volatility. For OTM puts ($k < 0, d_2 < 0$), Vanna is strictly positive; for OTM calls ($k > 0, d_2 > 0$), Vanna is negative. Because spot equity returns and implied volatility exhibit a strong negative correlation ($\text{Corr}(dS, d\sigma) \approx -0.78$), an afternoon compression in 0DTE implied volatility ($d\sigma < 0$) reduces the negative delta of dealer short OTM puts, compelling dealers who are short downside puts to **buy ES futures** relentlessly into the close (the classic "Vanna tailwind").
3. **Charm ($\text{Charm} = -\frac{\partial \Delta}{\partial \tau_{\text{eff}}} = \phi(d_1)\left[\frac{q}{\sigma\sqrt{\tau_{\text{eff}}}} - \frac{d_2}{2\tau_{\text{eff}}}\right]$):** Quantifies the rate at which OTM option deltas decay toward zero and ITM option deltas converge toward $\pm 1.0$ purely from the passage of time, scaling as $\tau_{\text{eff}}^{-1}$ near ATM wings ($|d_2| \approx 1$). Even if spot $S_t$ and implied vol $\sigma_t$ remain completely static between 13:00 ET and 15:30 ET, Charm bleeds away the delta of OTM 25-delta puts at a rate exceeding **12 delta points per hour**, forcing dealers who are short OTM puts (long positive delta via hedges) to unwind long hedges or, when dealers are **long OTM puts** against customer overwriters, to buy back short futures hedges mechanically.
4. **Speed ($\text{Speed} = \frac{\partial \Gamma}{\partial S} = -\frac{\Gamma}{S}\left(\frac{d_1}{\sigma\sqrt{\tau_{\text{eff}}}} + 1\right)$) and Vomma ($\text{Vomma} = \frac{\partial^2 V}{\partial \sigma^2} = \mathcal{V}\frac{d_1 d_2}{\sigma}$):** Speed governs the asymmetry of gamma transition as spot breaches strike boundaries. During rapid late-day selloffs ($>1.2\%$ after 14:30 ET), Speed causes dealer gamma to flip from stabilizing long-gamma near ATM iron condor wings to destabilizing **short-gamma** in the 10-delta tail, accelerating intraday flash cascades [[4]](https://www.sec.gov/files/staff-report-equity-options-market-structure.pdf).

Table 2 contrasts the empirical Greek scaling and dealer ES hedging flow profiles across five SPX expiration buckets per $10 billion of option notional open interest.

| Option Expiration Bucket | ATM Gamma per 1% Spot Move ($B) | Vanna Flow per 1 Vol Pt Drop ($B) | Afternoon Charm Flow 14:00-16:00 ($B) | Speed Convexity Ratio (pts) | Intraday Realized Vol Impact (bps) |
| :--- | :---: | :---: | :---: | :---: | :---: |
| 0DTE (15:00 ET Window) | $6.85B | $1.92B | $4.40B | 18.40 pts | -145.0 bps |
| 0DTE (10:00 ET Window) | $2.40B | $1.15B | $1.35B | 6.20 pts | -62.0 bps |
| 1DTE (Overnight Expiry) | $1.55B | $1.48B | $0.72B | 3.85 pts | -28.0 bps |
| 7DTE (Weekly Strip) | $0.68B | $1.85B | $0.24B | 1.40 pts | -12.0 bps |
| 30DTE (Monthly Strip) | $0.32B | $2.10B | $0.06B | 0.55 pts | -4.5 bps |

---

## 4. VIX vs. VIX1D Construction, Variance Swap Replication & Corridor Variance

### 4.1 Model-Free Variance Swap Replication and the OTM Strip Integral

By Carr and Madan (1998) and Demeterfi et al. (1999), the fair strike $K_{\text{var}}$ of a continuously sampled log-return variance swap over $[0, T]$ can be replicated statically without specifying a parametric stochastic volatility process by holding a continuum of OTM puts and calls weighted inversely by the square of their strike price $K^2$ [[2]](https://cdn.cboe.com/resources/vix/vixwhite.pdf):

$$\sigma_{\text{var}}^2(T) = \frac{2 e^{rT}}{T} \left( \int_0^{F_T} \frac{P(K, T)}{K^2} \, dK + \int_{F_T}^{\infty} \frac{C(K, T)}{K^2} \, dK \right)$$

Both the benchmark **30-day Cboe Volatility Index (VIX)** and the **1-Day Cboe Volatility Index (VIX1D)** implement discretized approximations of this strip:
$$\sigma_{\text{VIX}}^2 = \frac{2}{T} \sum_{i} \frac{\Delta K_i}{K_i^2} e^{rT} Q(K_i) - \frac{1}{T}\left[\frac{F}{K_0} - 1\right]^2$$
where $K_0$ is the highest strike below the forward index level $F$, $Q(K_i)$ is the midpoint quote of the OTM option at strike $K_i$, and $\Delta K_i = (K_{i+1} - K_{i-1})/2$ [[2]](https://cdn.cboe.com/resources/vix/vixwhite.pdf).

### 4.2 Discretization Truncation Bias, Jump Tails, and VIX1D Microstructure

While VIX interpolates between near-term and next-term SPX weeklys (with $23 < \text{DTE} < 37$ days), **VIX1D** interpolates across the current day's PM-settled 0DTE strip and the next trading session's 1DTE PM-settled strip. Institutional variance desks track three structural divergences between VIX1D, 30-day VIX, and OTC variance swaps:

1. **Strike Grid Truncation & the Two-Zero-Bid Rule:** The Cboe methodology truncates the OTM option strip once two consecutive strikes post a zero bid ($\text{Bid} = 0.00$). During calm intraday sessions, the 0DTE strip truncates within $\pm 2.5\%$ of spot, omitting deep tail jump risk. Conversely, when a sudden 1% afternoon pullback illuminates bids across 15 previously dormant OTM put strikes, the discretized sum $\sum (\Delta K_i / K_i^2) Q(K_i)$ jumps discontinuously, causing VIX1D to exhibit a **vol-of-vol $3.45\times$ higher than 30-day VIX**.
2. **Jump-Diffusion Bias in Variance Swaps:** If the underlying index experiences a discontinuous Poisson jump of size $J = \ln(S_{t+} / S_{t-})$, the realized variance increment vs. the continuous replication strip diverges by the cubic jump error:
 $$\mathbb{E}^\mathbb{Q}[\text{RV}_{[0,T]}] - K_{\text{var}}^2 = -\frac{2}{T} \mathbb{E}^\mathbb{Q}\left[ \sum_{0 < s \le T} \left( e^{J_s} - 1 - J_s - \frac{1}{2}J_s^2 \right) \right] \approx -\frac{1}{3T} \mathbb{E}^\mathbb{Q}\left[\sum J_s^3\right]$$
 For negatively skewed equity jumps ($J_s < 0$), $-\frac{1}{3}J_s^3 > 0$, meaning true realized variance swap payoffs exceed the static $1/K^2$ log-contract strip value by **65 to 140 volatility bps** in crisis regimes.
3. **OTC Corridor Variance & Capped Variance Migration:** Following severe dealer losses on uncapped single-stock variance swaps during retail squeeze episodes, institutional OTC flow in 2026 has migrated heavily toward **corridor variance swaps** (which accrue variance only when $S_t \in [B_L, B_U]$) and **dispersion swaps packaged as volatility swaps** (linear in $\sigma$ rather than $\sigma^2$) to eliminate extreme left-tail $1/K^2$ convexity exposure [[8]](https://www.bis.org/publ/qtrpdf/r_qt2309.htm).

---

## 5. Equity Index Dispersion & Implied Correlation Trading Mechanics

### 5.1 The Fundamental Variance Decomposition Identity

Index dispersion trading arbitrages the structural risk premium embedded between index implied volatility ($\sigma_I$) and the basket of constituent single-stock implied volatilities ($\sigma_i$). Let $w_i$ denote the market-capitalization weight of constituent $i \in \{1, \dots, N\}$ in the S&P 500 ($\sum_{i=1}^N w_i = 1$). The exact index variance decomposition is:

$$\sigma_I^2 = \sum_{i=1}^N w_i^2 \sigma_i^2 + \sum_{i=1}^N \sum_{j \neq i} w_i w_j \rho_{ij} \sigma_i \sigma_j$$

Assuming a homogeneous equicorrelation parameter $\bar{\rho}_{\text{imp}}$ across all off-diagonal pairs $(i \neq j)$, the **Implied Correlation** of the index surface at maturity $T$ and moneyness $k$ is solved directly as:

$$\bar{\rho}_{\text{imp}}(k, T) = \frac{\sigma_I^2(k, T) - \sum_{i=1}^N w_i^2 \sigma_i^2(k, T)}{\left(\sum_{i=1}^N w_i \sigma_i(k, T)\right)^2 - \sum_{i=1}^N w_i^2 \sigma_i^2(k, T)} \approx \frac{\sigma_I^2(k, T)}{\left(\sum_{i=1}^N w_i \sigma_i(k, T)\right)^2}$$

### 5.2 Structural Supply-Demand Imbalance & Dispersion Carry Mechanics

Across institutional equity markets, a persistent structural imbalance governs correlation pricing:
- **Index Volatility Demand (Long $\sigma_I$):** Pension funds, insurance variable annuity hedgers, macro hedge funds, and retail buffer ETFs systematically purchase SPX OTM puts and collars for portfolio tail protection, bidding up $\sigma_I$ and steepening index skew.
- **Single-Stock Volatility Supply (Short $\sigma_i$):** Yield-enhancement funds, private wealth structured note issuers (autocallables), and systematic call-overwriting ETFs sell single-stock upside calls and downside puts on mega-cap constituents (NVDA, AAPL, MSFT, AMZN, META, GOOGL), depressing $\sigma_i$ relative to fundamental earnings variance.

A standard **Long Dispersion (Short Correlation)** package sells index volatility (via SPX variance swaps, straddles, or iron condors) and buys constituent volatility across a proxy basket of the top $M = 50$ or $M = 100$ S&P 500 components (which capture **86.5% to 92.4% of total index variance sensitivity** while minimizing execution slippage). Desk execution differentiates between two primary weighting schemes:

1. **Vega-Weighted Dispersion:** Sets the constituent vega notional equal to the index vega notional ($\sum_{i=1}^M w_i^* \mathcal{V}_i = \mathcal{V}_I$). Because $\sigma_i > \sigma_I$ (on average $\sum w_i \sigma_i / \sigma_I \approx 1.75$), a vega-neutral book is **net short dollar gamma**, losing money if all stocks move in lockstep ($\rho_{\text{realized}} \to 1$) during a macro crash.
2. **Theta-Weighted (Gamma-Neutral) Dispersion:** Scales the index leg up by the ratio $\left(\sum w_i \sigma_i\right) / \sigma_I$ so that daily theta bleed on the long single-stock legs is fully financed by the short index theta ($\Theta_{\text{net}} = 0$ and $\$\Gamma_{\text{net}} = 0$ at inception). This leaves the desk **net short vega** on the index relative to constituents, requiring explicit tail-hedging via VIX out-of-the-money calls or 5-delta SPX put ratios.

### 5.3 Implied Correlation Skew ($\bar{\rho}(k)$) & Single-Stock Earnings Convexity

Evaluating $\bar{\rho}_{\text{imp}}(k, T)$ across moneyness strikes reveals a steep **Implied Correlation Skew**: while 25-delta upside call correlation in SPX trades near **0.24 to 0.31**, 25-delta downside put correlation trades between **0.58 and 0.72**, and 10-delta tail put correlation approaches **0.88** [[1]](https://www.cboe.com/insights/) [[2]](https://cdn.cboe.com/resources/vix/vixwhite.pdf). This reflects the empirical stylized fact that upside equity rallies are driven by idiosyncratic mega-cap dispersion (where top technology constituents decouple on earnings beats), whereas downside market liquidations drive cross-sectional equity correlations toward unity ($\rho_{ij} \to 1$). Consequently, institutional dispersion desks increasingly structure **Skew-Adjusted Conditional Dispersion**, selling 25-delta SPX put spreads against long 25-delta single-stock strangles around quarterly earnings cycles, capturing both the elevated downside correlation risk premium and single-name post-earnings realized jump variance.

---

## 6. Portfolio Margining & Regulatory Capital Architecture: OCC STANS, CME SPAN 2, and Eurex Prisma

### 6.1 Institutional Margin Compression vs. Strategy-Based Reg-T

For institutional market makers and quantitative volatility funds operating through [tier-1 prime brokers and clearing members](/brokers) regulated across global jurisdictions (see our [comprehensive regulatory directory](/regulators)), capital efficiency on SVI surface arbitrage and dispersion books depends entirely on portfolio-level risk methodologies rather than rigid strategy-based Regulation T rules [[3]](https://www.theocc.com/risk-management/margin-methodology) [[5]](https://www.cmegroup.com/clearing/risk-management/span-overview.html).

- **OCC STANS (System for Theoretical Analysis and Numerical Simulations):** Used by the Options Clearing Corporation under [SEC](/regulators/sec) and [CFTC](/regulators/cftc) oversight, STANS executes **10,000 Monte Carlo copula scenarios** daily across 500+ underlying risk factors, incorporating heavy-tailed Student-$t$ innovations, GARCH volatility updating, and nine-point implied volatility surface shocks per expiry. STANS computes a 99% two-day Expected Shortfall (ES) alongside stress-test add-ons, granting **68% to 78% cross-product margin offsets** between short SPX index options, long constituent equity options, and CME E-mini S&P 500 futures held in cross-margin accounts [[3]](https://www.theocc.com/risk-management/margin-methodology).
- **CME SPAN 2 (Historical Value-at-Risk & Vol Surface Shocks):** Replacing legacy 16-scenario SPAN, CME SPAN 2 models joint futures and options books across ES, NQ, and RTY using filtered historical simulation (HVaR) combined with explicit liquidity, concentration, and **0DTE intraday expiry add-ons** designed to capture 15:00–16:00 ET pin-risk and post-settlement exercise/assignment slippage [[5]](https://www.cmegroup.com/clearing/risk-management/span-overview.html).
- **Eurex Prisma & EMIR 3.0 Anti-Procyclicality Buffers:** Under [ESMA](/regulators/esma), [BaFin](/regulators/bafin), and [FCA](/regulators/fca) supervision, Eurex Prisma calculates portfolio Expected Shortfall over a 2-to-5 day liquidation horizon blending a 3-year lookback window with a **25% stressed-period weight** (per EMIR RTS anti-procyclicality mandates) to prevent sudden margin procyclicality spirals during Euro Stoxx 50 (SX5E) volatility spikes [[7]](https://www.esma.europa.eu/publications-and-data/all-publications).
- **SA-CCR & Basel III Endgame (FRTB) CVA/Market Risk:** For bank-affiliated derivatives desks, non-cleared OTC variance and dispersion swaps attract Standardized Approach for Counterparty Credit Risk (SA-CCR) Exposure-at-Default ($\text{EAD} = 1.4 \times (\text{RC} + \text{PFE})$) with a **1.5% supervisory factor** on equity indices and **32% supervisory volatility** on single-name equities, driving rapid migration toward listed SPX box spreads, listed CBOT/Cboe variance futures, and centrally cleared total return swaps [[8]](https://www.bis.org/publ/qtrpdf/r_qt2309.htm).

Table 3 details empirical initial margin requirements, cross-margin offsets, and stressed liquidity add-ons across the four primary clearing regimes for a standardized **$100 million notional SPX vs. Top-50 Dispersion & 0DTE Gamma Book**.

| Clearing & Margin Regime | Base Initial Margin (% Notional) | Cross-Asset Dispersion Offset (%) | 0DTE Intraday Pin Add-On (bps) | Stressed Vol Spike Buffer (% Margin) | Net Capital Requirement ($M per $100M) |
| :--- | :---: | :---: | :---: | :---: | :---: |
| Legacy Strategy Reg-T | 24.50% | 12.0% | 185.0 bps | 35.0% | $21.56M |
| FINRA TIMS Portfolio Margin | 12.80% | 52.5% | 110.0 bps | 25.0% | $9.40M |
| OCC STANS + CME Cross-Margin | 6.40% | 76.2% | 65.0 bps | 18.5% | $4.85M |
| Eurex Prisma (EMIR 3.0 APC) | 7.15% | 71.8% | 72.0 bps | 25.0% | $5.38M |

---

## 7. Production Implementation Checklist for Institutional Vol Desks

To deploy an institutional-grade SVI/SSVI calibration and 0DTE/dispersion execution engine capable of surviving 2026 intraday liquidity shocks without violating static arbitrage or clearinghouse risk gates, quantitative engineering teams must enforce six operational invariants:

1. **Sub-Tick Forward & Borrow Extraction:** Never calibrate SVI against spot $S_t$ and a textbook risk-free rate. Extract the exact implied forward $F_T$ and implied repo/dividend rate $q_T$ simultaneously on every 100-millisecond OPRA snapshot using a robust Huber-loss regression across the 5 nearest-to-ATM put-call pairs ($C_i - P_i = e^{-rT}(F_T - K_i)$).
2. **Hard Durrleman & Roger Lee Barrier Enforcement:** Embed exact analytical gradients of $g(k)$ and $\partial_T w(k, T)$ inside the L-BFGS-B or ADMM optimizer. Reject any surface update where $\min_k g(k) < 10^{-7}$ or $b(1 + |\rho|) > 1.995$, falling back to the prior valid SSVI state vector if corrupted OPRA quotes appear during fast markets.
3. **Diurnal Variance Clock Integration ($\tau_{\text{eff}}$):** Map every 0DTE and 1DTE timestamp through a high-frequency variance clock spline anchored to historical 1-minute ES futures realized variance and scheduled macro event weights (CPI, FOMC, NFP, NVDA/AAPL earnings) so that ATM implied volatility does not exhibit artificial intraday step-functions.
4. **Joint Gamma-Vanna-Charm Hedging Controller:** Replace naive spot-delta threshold hedging with a stochastic optimal control hedger that forecasts the deterministic afternoon Charm bleed ($\int_{t}^{T} \text{Charm}_u \, d\tau_{\text{eff}}$) and expected Vanna flow ($\text{Vanna} \cdot \beta_{\sigma, S} \, dS$), pre-positioning limit orders in CME ES/MES futures to capture spread rather than crossing the spread during the 15:50 ET MOC imbalance window.
5. **Truncated-Wing Variance Strip Augmentation:** When pricing OTC variance swaps or hedging VIX/VIX1D futures against SPX option strips, splice an SVI-parameterized tail integral onto the discretized Cboe sum beyond the two-zero-bid cutoff strike $K_{\min}$ to prevent sudden vega steps when deep OTM bids illuminate.
6. **Real-Time STANS / SPAN 2 Margin Shadowing:** Run a GPU-accelerated 10,000-path Student-$t$ copula shadow engine mirroring OCC STANS and CME SPAN 2 intraday margin calls every 5 minutes, dynamically trimming 0DTE short-wing exposure before clearing member intraday risk limits trigger forced liquidation.

---

## Primary Regulatory Filings, Statutory Citations & Empirical References

- **[1]** Cboe Global Markets (2026), *Market Intelligence & 0DTE S&P 500 Index Options Microstructure Analytics*. Available at: [https://www.cboe.com/insights/](https://www.cboe.com/insights/)
- **[2]** Cboe Exchange, Inc. (2025/2026), *Cboe Volatility Index (VIX) & 1-Day Volatility Index (VIX1D) Mathematics Methodology White Paper*. Available at: [https://cdn.cboe.com/resources/vix/vixwhite.pdf](https://cdn.cboe.com/resources/vix/vixwhite.pdf)
- **[3]** The Options Clearing Corporation (OCC) (2026), *STANS (System for Theoretical Analysis and Numerical Simulations) Margin Methodology & Monte Carlo Risk Framework*. Available at: [https://www.theocc.com/risk-management/margin-methodology](https://www.theocc.com/risk-management/margin-methodology)
- **[4]** U.S. Securities and Exchange Commission (SEC) Division of Trading and Markets (2025/2026), *Staff Report on Equity and Options Market Structure Conditions*. Available at: [https://www.sec.gov/files/staff-report-equity-options-market-structure.pdf](https://www.sec.gov/files/staff-report-equity-options-market-structure.pdf)
- **[5]** CME Group Clearing (2026), *CME SPAN 2 Risk Management & Equity Index Options Portfolio Margining Framework*. Available at: [https://www.cmegroup.com/clearing/risk-management/span-overview.html](https://www.cmegroup.com/clearing/risk-management/span-overview.html)
- **[6]** Commodity Futures Trading Commission (CFTC) Division of Market Oversight (2026), *Commitments of Traders (COT) & Listed Index Derivatives Market Surveillance Reports*. Available at: [https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm](https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm)
- **[7]** European Securities and Markets Authority (ESMA) (2026), *EMIR 3.0 Regulatory Technical Standards on CCP Anti-Procyclicality Margin Buffers & Eurex Prisma Validation*. Available at: [https://www.esma.europa.eu/publications-and-data/all-publications](https://www.esma.europa.eu/publications-and-data/all-publications)
- **[8]** Bank for International Settlements (BIS) Monetary and Economic Department (2025/2026), *Quarterly Review: Zero-Day Options, Dealer Gamma Hedging, and Non-Bank Leverage in Equity Volatility Markets*. Available at: [https://www.bis.org/publ/qtrpdf/r_qt2309.htm](https://www.bis.org/publ/qtrpdf/r_qt2309.htm)

---

> **Flagship Cornerstone Authority Reference**: For the master institutional framework, 2026 benchmark matrix, and mathematical calibration governing this domain, consult our flagship pillar guide: **[S&P 500 SVI Volatility Surface & Dealer Gamma Exposure Desk (SVI-DEALER-GEX)](/desk-notes/sp500-svi-volatility-surface-gamma-exposure-desk)** and the canonical entity dossier **[S&P 500 SVI Volatility Surface & Dealer GEX](/instruments/sp500-svi-volatility-surface-gamma-exposure-desk)**.

## Frequently asked questions

### How does Gatheral's SSVI parameterization eliminate both butterfly and calendar spread arbitrage across the volatility surface?

Surface SVI (SSVI) parameterizes total implied variance w(k, theta_t) globally using at-the-money total variance theta_t = w(0, T) and a power-law curvature kernel phi(theta_t). Calendar spread arbitrage (dw/dT < 0) is eliminated whenever ATM total variance is non-decreasing in maturity (d_t theta_t >= 0) and the derivative d_theta(theta * phi(theta)) remains within explicit correlation-dependent bounds. Static butterfly arbitrage (negative Breeden-Litzenberger risk-neutral density) is prevented by enforcing Durrleman's condition g(k) >= 0 and Roger Lee's asymptotic wing slope constraint theta * phi(theta) * (1 + |rho|) <= 2.

### Why do Vanna and Charm flows dominate SPX 0DTE futures hedging during the final two hours of the US cash equity session?

As time-to-expiration approaches zero at 16:00 ET, Charm (-dDelta/dTau) scales inversely with remaining variance time (tau_eff^-1), bleeding away out-of-the-money option deltas at over 12 delta points per hour even if spot remains flat. Simultaneously, afternoon compression in 0DTE implied volatility triggers Vanna (dDelta/dVol) flows across out-of-the-money puts. When dealers are net short downside 0DTE puts, both diurnal Charm decay and implied volatility compression reduce their negative option delta, compelling mechanical buying of $14.2B to $19.8B in CME E-mini S&P 500 futures into the close.

### What drives the structural implied correlation risk premium in S&P 500 dispersion trading?

The S&P 500 implied correlation premium arises from segmented institutional order flow: macro hedgers, pensions, and buffer ETFs systematically buy SPX index protection (elevating index implied volatility sigma_I), while yield-enhancement funds and structured note issuers systematically overwrite single-stock calls and puts on mega-cap constituents (depressing constituent volatility sigma_i). Dispersion desks harvest this spread—averaging 14.2 correlation points between COR3M and realized correlation—by selling index variance and buying a vega- or theta-weighted basket of the top 50 constituent options.

### How does OCC STANS portfolio margining treat index dispersion and 0DTE gamma books compared to FINRA Reg-T?

Whereas legacy Regulation T evaluates option legs in isolated strategy buckets and charges roughly 24.50% of notional ($21.56M per $100M dispersion portfolio), OCC STANS executes 10,000 daily Monte Carlo copula scenarios incorporating Student-t fat tails, GARCH volatility updates, and nine-point implied volatility surface shocks. By recognizing the empirical correlation between short SPX index vega/gamma, long constituent equity options, and CME E-mini futures, STANS and CME cross-margining grant 68% to 78% margin offsets, reducing net capital requirements to roughly $4.85M per $100M notional.


## Source URLs

- https://www.cboe.com/insights/
- https://cdn.cboe.com/resources/vix/vixwhite.pdf
- https://www.theocc.com/risk-management/margin-methodology
- https://www.sec.gov/files/staff-report-equity-options-market-structure.pdf
- https://www.cmegroup.com/clearing/risk-management/span-overview.html
- https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm
- https://www.esma.europa.eu/publications-and-data/all-publications
- https://www.bis.org/publ/qtrpdf/r_qt2309.htm

## Citation

- APA: Dr. Alistair Finch, CQF, FRM. (2026). Arbitrage-Free SVI Volatility Surface Calibration, 0DTE Intraday Gamma-Vanna-Charm Flows, and Index Dispersion Mechanics in 2026. VolSurface Desk. https://volsurfacedesk.com/desk-notes/svi-volatility-surface-arbitrage-free-calibration-0dte-gamma-vanna-dispersion-2026
- Canonical article: [Arbitrage-Free SVI Volatility Surface Calibration, 0DTE Intraday Gamma-Vanna-Charm Flows, and Index Dispersion Mechanics in 2026](https://volsurfacedesk.com/desk-notes/svi-volatility-surface-arbitrage-free-calibration-0dte-gamma-vanna-dispersion-2026)
- Site research index: [VolSurface Desk](https://volsurfacedesk.com/llms-full.txt)
