What Is Vanna and How Does It Affect Markets?

The cross-Greek behind slow grinds and squeezes.

Vanna explained: it is a second-order option Greek that measures how delta changes as implied volatility moves, or equivalently how vega changes as the underlying price moves. Mathematically, vanna = ∂delta/∂σ = ∂vega/∂spot. That dual identity is what makes it powerful — it links the volatility surface to spot-price dynamics in a way that forces real, mechanical hedging flows from dealers and market-makers. Those flows, aggregated across large open-interest strikes, can visibly tilt intraday and multi-day market structure. Understanding vanna is essential context for reading slow-grind rallies, sharp vol-spike reversals, and the rhythm of options expiries. You can explore how the underlying Greeks feed into it using the options pricing calculator.

Vanna explained: the math and the intuition

Start with delta. A call option's delta is roughly the probability — under the risk-neutral measure — that it expires in the money. As implied volatility (IV) rises, that probability distribution widens: out-of-the-money options gain delta because the market assigns more weight to large moves reaching the strike. As IV falls, the distribution narrows and OTM deltas shrink back toward zero.

For a dealer who sold options and is delta-hedging, this creates a headache even if spot doesn't move. A fall in IV changes every option's delta on the book. The dealer must rebalance — buying or selling the underlying — not because prices moved, but because volatility did. That rebalancing is a vanna flow.

For a European call, vanna under Black-Scholes is:

Vanna = −e^−qT · N'(d₁) · d₂ / σ

where N'(d₁) is the standard normal density and d₂ = d₁ − σ√T. The sign depends on whether the option is OTM or ITM. For an OTM call, d₂ is negative, so vanna is positive: as IV rises, delta rises. For a deep-ITM call, d₂ is large and positive, so vanna is negative: IV changes matter little because delta is already near 1.

Why OTM options carry the most vanna

Vanna peaks for options sitting near but outside the money — roughly in the 20–40 delta range. The intuition: at-the-money options already have delta close to 0.5 and don't move much when vol shifts. Deep-in-the-money options have deltas pinned near 1. It is the borderline cases — the strikes where the outcome is genuinely in question — that see the largest delta re-rating when IV moves.

Consider a put with a strike 8% below spot, currently 25-delta. If IV compresses from 22% to 16%, the market is saying large moves are less likely. That put's delta might fall from 0.25 to 0.16. A dealer short that put who was hedging by selling the underlying now has too much short hedge — they must buy stock back. Multiply that across millions of contracts and you have a coherent directional bid.

A rough numerical feel: for a 30-delta put with σ = 0.20, T = 30 days, and spot = 100, vanna is approximately −0.10 per point of vol. A 4-vol-point IV compression implies a delta shift of roughly 0.04 on that option. On 50,000 contracts (5 million shares of notional delta per 0.01), that is a 200,000-share mechanical bid — not trivial.

The vanna rally and the vol-spike reversal

The term "vanna rally" refers to a specific regime: IV is falling (or stable at low levels), and the market grinds higher with low realized volatility. The causality runs like this:

1. Dealers are typically net short puts relative to the open interest they carry — retail and institutional buyers of downside protection leave dealers holding short gamma/short vanna positions on OTM puts.
2. As IV falls, the deltas of those short puts collapse. Dealers are over-hedged short and must buy spot to neutralize delta.
3. Buying lifts spot, which further compresses the IV that gets priced into those puts (lower spot vol, positive drift). More delta re-hedging follows.
4. The result is a slow, low-vol grind — characteristically narrow daily ranges with a persistent upward bias.

The reverse fires on vol spikes. A sharp IV expansion — a geopolitical shock, a macro surprise — forces dealers to sell spot aggressively as their put deltas balloon. That selling can create a feedback loop: falling spot, rising IV, more delta selling. Vanna amplifies moves in both directions; it merely tends to be more visible on the upside because vol compression regimes last longer than vol spikes.

Vanna around expiries and OpEx

Vanna flows are not distributed evenly through time. They concentrate around large open-interest events: monthly and quarterly options expiration (OpEx), and especially quad witching, when equity index futures, equity index options, stock options, and single-stock futures all settle simultaneously.

As expiry approaches, options that are OTM bleed extrinsic value rapidly. Their effective vanna exposure collapses to near zero as they become worthless. This removes a structural hedging pressure — dealers no longer need to manage those delta sensitivities — and the market can see a notable character shift shortly after a large expiry. The vanna and charm flows around these windows are among the most predictable structural influences on short-term equity index behavior.

Charm — the rate at which delta changes with time (∂delta/∂T) — interacts directly with vanna. In the final days before expiry, charm dominates: even if IV stays flat, delta on OTM options decays rapidly with time. Vanna and charm can reinforce each other when IV is also falling into expiry, compressing dealer hedging needs from two directions simultaneously, or they can partially offset when vol spikes into OpEx.

What vanna flows can and cannot tell you

Dealer positioning models that estimate vanna exposure aggregate open interest by strike, assign a sign to dealer inventory, and calculate the net delta sensitivity to a vol move. These are estimates with real limitations:

• Positioning is inferred, not observed. Dealer net exposure is reverse-engineered from public open-interest data, with assumptions about who is on which side of each trade. Those assumptions can be wrong.
• Flows compete with fundamentals. A strong macro catalyst overwhelms mechanical vanna hedging. The framework is most useful in low-information, drift-dominated regimes — not around earnings seasons or policy announcements.
• Realized IV vs. implied IV divergence. Vanna flows depend on changes in implied, not realized, volatility. If implied vol is sticky while realized vol falls, the mechanical pressure is muted even though the market is calm.
• The model is Black-Scholes. Real dealer hedging uses more complex models with vol-of-vol adjustments that alter the precise magnitude of vanna sensitivity.

Used correctly, vanna is a structural context layer — a reason why a market might grind in a direction that fundamentals alone don't explain, and a warning sign for when low-vol complacency can unwind quickly. It does not replace price analysis; it informs it.