Vanna and Charm: Why Dealer Delta Can Change Without a Rally

A one-percentage-point volatility move is 0.01 in decimal volatility. Confusing it with a 1.00 move creates a hundredfold scaling error. Charm annualized under one convention should not be compared directly with a daily estimate under another.

The sign is not a universal market rule

“Volatility down means dealers buy” is too broad. The result depends on moneyness, option type, position sign, expiry, and the portfolio being aggregated. Both customer ownership and dealer offsets matter. A large exposure in one instrument may be hedged against another.

Near expiration, out-of-the-money delta tends toward zero while in-the-money delta tends toward its limiting value, subject to the product's payoff and model assumptions. The resulting time effect differs across strikes. Aggregating everything into a single charm number can obscure opposing concentrations.

Rather than memorizing a directional slogan, reprice a small hypothetical position under controlled inputs. Hold spot constant and change only volatility. Then restore volatility and advance time. Finally change both. The exercise makes unit conventions and interactions visible before they enter a market narrative.

Use event scenarios instead of flow promises

Before a scheduled announcement, build a grid with spot unchanged, higher, and lower against volatility unchanged, higher, and lower. If a portfolio's estimated hedge response flips across reasonable scenarios, confidence in a single directional forecast should decrease. That uncertainty is useful information.

After the event, replace hypothetical inputs with observed ones and compare the estimated delta change with price and traded volume. Do not use the price move itself to decide retroactively which inventory sign must have been correct. That would make the explanation impossible to falsify.

The OIC discussion of volatility skew and term structure is a useful companion because volatility does not always move uniformly across strikes and maturities. A flat volatility shock is a simplifying scenario, not an empirical law.

A practical research checklist

Save the units, inventory convention, time basis, and volatility surface timestamp alongside every exposure estimate. Ask whether an apparent change came from new trades, repricing existing positions, or changing the expiry filter. Keep those explanations separate in the journal.

Review your scenario after a substantial spot move or volatility repricing. If the original view relied on a small-change approximation, large moves can invalidate it. A complex Greek label does not make an outdated assumption more reliable.

Read gamma exposure fundamentals before combining these sensitivities, then use a structured preparation routine to record the assumptions. OrderMap is a read-only analytics platform; use the data views available to your account and verify provider entitlements before building a routine around them.

Stress the unit convention

Suppose an invented portfolio has vanna of 10,000 share equivalents per 1.00 absolute volatility. A decline from 20% to 19% is a change of negative 0.01, implying a local delta contribution of negative 100 share equivalents. If another report expresses vanna per one volatility percentage point, the same sensitivity would be reported as 100. The exposures agree despite very different displayed numbers.

Perform the same reconciliation for charm. An annualized number cannot simply be compared with a daily one without understanding the model's day count and time direction. Weekend and overnight treatment can differ across systems.

For your journal, write one complete sentence for every Greek: input change, output change, units, position sign, and as-of time. If you cannot complete that sentence, the number is not yet ready to support a directional hedge-flow claim.

Explore an OrderMap workspace