Charm: Delta Decay and What Charm Exposure Measures

Start with delta

Delta is the first thing to understand before charm makes sense. Delta measures how much an option's price changes when the underlying moves by one dollar, holding everything else constant. A call with a delta of 0.50 gains approximately $0.50 in value for every $1 rise in the underlying. Delta is not fixed. It depends on how far the option is from the strike, how much time remains, and what implied volatility the market is pricing.

A trader who wants to be delta-neutral, meaning indifferent to small moves in the underlying, must hold a position in the underlying that offsets the option's delta. Hold 100 shares of stock to offset one short call with a delta of 1.00. Hold 50 shares to offset one short call with a delta of 0.50. This is the logic of delta hedging.

What charm measures: gamma versus charm versus vanna

Three second-order Greeks each describe a different way delta can change. Understanding which dimension each one covers prevents confusion when reading options analytics.

• Gamma measures how much delta changes when the underlying price moves. A $1 rise in the underlying price changes the option's delta by an amount equal to gamma. Gamma is the curvature of the option's price-to-underlying relationship.
• Charm measures how much delta changes as time passes. One calendar day or one trading day advancing changes the option's delta by an amount equal to charm. This is the dimension called delta decay.
• Vanna measures how much delta changes when implied volatility moves. A 1-point rise in implied volatility changes the option's delta by an amount equal to vanna.

A delta-hedged book can drift out of delta-neutral through any of these three channels. Charm specifically captures the time channel.

Charm is not theta

Theta and charm are both time-related Greeks, but they measure different things. Theta measures the change in the option's price per day of time passing. Charm measures the change in the option's delta per day of time passing. They are not the same number, they are not simply proportional, and tracking one does not give you the other.

An option can have a large theta and a small charm, or a small theta and a large charm. The relationship between the two depends on where the option sits relative to the strike and how much time remains. Near-the-money options with little time left tend to have large charm relative to their theta because delta is shifting rapidly even as the dollar-value time decay of price is small.

A simple hedging example

Assume a trader owns options representing 1,000 contracts, each with a multiplier of 100, and a current delta of +0.25. Total delta-equivalent exposure:

1,000 × 100 × 0.25 = +25,000 shares

A delta hedge would be approximately short 25,000 shares. Now assume the underlying price and implied volatility are unchanged, but one trading day passes. The option's delta falls from +0.25 to +0.22 because of time decay in delta. New delta exposure:

1,000 × 100 × 0.22 = +22,000 shares

The existing hedge is still short 25,000 shares. The portfolio is now approximately 3,000 shares net short relative to the new option delta. To restore neutrality, the hedger would buy roughly 3,000 shares. Nothing in this simplified example required the stock price to move. The hedge drift came entirely from time. That is the intuitive meaning of charm.

Why charm grows near expiration

Delta for an out-of-the-money call starts near zero and stays near zero for most of an option's life, then declines further toward zero in the final days before expiry. Delta for an in-the-money call starts high and stays high for most of an option's life, then rises further toward 1.0 near expiry. The rate of this convergence, measured per unit of time, is what charm captures. Because the convergence accelerates in the final weeks and especially the final days, charm typically has its largest magnitude as expiration approaches.

Single-option charm versus aggregate charm exposure

For a single option, charm is a property of that contract. For a book of options, aggregate charm exposure sums the charm of all positions weighted by contract size and shares per contract. The result estimates how much the combined delta of the book will shift per day due to time passage alone, assuming prices and implied volatility stay constant.

Dealers who are net short a large number of call options on an index may carry a significant aggregate negative charm exposure. As each day advances, their effective short delta grows (because short calls losing delta means a shorter net delta for the dealer), which may require them to buy more of the underlying to stay delta-neutral. This is the hedging flow that aggregate charm exposure attempts to quantify.

Dealer charm is an estimate, not a fact

Published dealer charm exposure data is estimated, not directly observed. Options positioning data from CBOE and FINRA captures broker-dealer aggregate positions at a delay and does not separate market-making books from directional books. The estimate requires assumptions about which counterparties are dealers, what positions they carry, and whether they hedge continuously or discretely. Use aggregate charm data as one input in a broader view of likely hedging flows, not as a precise prediction of order flow.

Sign conventions vary by platform

Charm's sign is one of the most frequently misread elements of second-order Greeks. Two fully correct formulations can produce numbers with opposite signs:

• ∂Δ/∂t: differentiate delta with respect to calendar time moving forward. As calendar time advances, time remaining shrinks. For an out-of-the-money call, delta falls as expiry nears, so this derivative is typically negative.
• ∂Δ/∂τ: differentiate delta with respect to τ (time remaining to expiration), which shrinks as calendar time advances. Because dt = −dτ, this version has the opposite sign.

A platform using ∂Δ/∂t and another using ∂Δ/∂τ will display opposite-signed charm values for the same option. Neither is wrong. They are using different sign conventions. A Swoopr charm chart must therefore label which convention it uses and state in plain English what positive or negative displayed charm means. Common practical conventions:

• Long call charm is typically negative in the most common convention (∂Δ/∂t): the call's delta declines toward zero as time passes for out-of-the-money calls.
• Long put charm is typically positive in the same convention: the put's delta rises toward zero from negative values as time passes for out-of-the-money puts.

Always verify the exact convention used by any platform or data source before interpreting a reading. The time unit (calendar days vs. trading days) creates an additional ~44% scaling difference.

The Black-Scholes expression for charm

In the Black-Scholes framework, charm for a European call can be expressed as a function of the standard inputs: underlying price, strike, time to expiration, risk-free rate, dividend yield, and implied volatility. The expression involves the standard normal density function and depends on both d1 (the standardized distance between the current price and the break-even boundary) and d2. The formula is derivable from the partial derivative of delta with respect to time, but the exact form varies with whether dividends are continuous or discrete and how the calendar convention is applied.

What charm can tell an investor

• How much a delta-hedge will need to be adjusted if nothing else changes but one day elapses.
• Which contracts in a portfolio are contributing most to time-driven delta drift.
• Whether a portfolio's aggregate delta exposure is likely to grow or shrink passively over the next week.
• Roughly how much dealer re-hedging activity might be expected from expiring open interest.

What charm cannot tell an investor

• Whether a trade will be profitable. Charm describes delta dynamics, not direction or expected return.
• What implied volatility will do. Volatility changes are captured by vega (first-order) and vanna (second-order), not charm.
• What the underlying price will do. That is simply not what charm measures.
• The precise hedging flows that will occur. Actual dealer behavior depends on cost of hedging, inventory, spread income, and risk limits, not on the mechanical charm calculation alone.

Charm and 0DTE options

Zero days to expiration (0DTE) options have their entire remaining delta convergence compressed into a single trading session. A 0DTE call that is five percent out of the money may begin the day with a small delta and end the day with a delta of zero, or spike toward 1.0 if the underlying crosses the strike intraday. The charm over that one-day window can be enormous compared with a contract with thirty days of life. This is part of what makes 0DTE flow analysis different from standard options flow analysis: the hedging re-adjustment cycle is compressed to hours rather than days.

Charm and gamma interaction

Charm and gamma are both second-order Greeks but they operate in different dimensions. Near expiration, both gamma and charm tend to be large in magnitude for near-the-money options. A large gamma means delta moves quickly with the underlying price. A large charm means delta moves quickly with time. The two combine to make near-expiry, near-the-money options extremely sensitive to both dimensions simultaneously, which is why managing such positions requires frequent re-hedging along both axes.

A first-approximation decomposition of how delta changes over a short interval:

Change in delta ≈ Gamma × underlying-price change + Charm × time change + Vanna × volatility change + residual higher-order effects

This decomposition is not exact for large moves, and the Greek values themselves change during the interval. Its value is explanatory: attributing all estimated dealer hedge to gamma is incomplete. Even with no price movement, the charm term can change delta; with volatility movement, vanna can contribute too.

A worked scenario

An investor holds a 30-day, at-the-money call on a stock trading at $100, with a delta of 0.52 and a charm of -0.018 per calendar day. Ten calendar days pass with the stock still at $100. Approximate effect from charm alone: delta changes by 10 times -0.018, or -0.18. New estimated delta: roughly 0.34. If the investor held 52 shares to hedge the original delta of 0.52, the hedge has drifted by 18 shares. They now hold too many shares relative to the option's current delta.

Note that this approximation holds only if the stock stays exactly flat. If the stock moves, gamma will additionally change delta. The charm calculation assumes the stock is frozen; it isolates the time-passage dimension.

How to design a Charm Explorer tool

A Charm Explorer built for investor education should allow users to specify: underlying price, strike, days to expiration, implied volatility, risk-free rate, and dividend yield. The tool should display both single-option charm and the chart of projected delta over time, holding inputs fixed. An aggregate exposure tab should allow entry of multiple positions with their contract sizes to sum to a portfolio-level charm figure. Users should be told explicitly what convention the tool uses and what assumptions it holds constant.

Time scaling: calendar days versus trading days

Charm is typically quoted per calendar day in academic literature but per trading day in many trading platforms. The distinction matters because there are roughly 252 trading days per year but 365 calendar days, and implied volatility itself is usually annualized using a trading-day calendar. Using the wrong basis creates a roughly 44 percent error in the daily charm estimate. Verify the convention whenever comparing charm figures across platforms.

Charm around events

Implied volatility tends to be elevated before known events such as earnings announcements or Federal Reserve decisions and collapses immediately after. Because charm operates in the time dimension and volatility changes are captured by vanna, charm in isolation does not describe what happens to delta around an event. The post-event volatility collapse may change delta by more than charm would predict from time passage alone. In practice, both time (charm) and volatility (vanna) need to be accounted for when managing delta around events.

Limitations of charm as a hedging tool

• Charm is a local estimate. It assumes small movements in all dimensions. For large time periods or large price moves, higher-order effects make the charm approximation inaccurate.
• Charm assumes continuous hedging. Actual hedging is discrete, adding transaction costs.
• Charm is computed from a model (typically Black-Scholes). Real options markets price skew and term structure in ways the model does not fully capture, so realized charm can differ from model charm.
• Platform implementations differ. Some platforms compute charm from the options chain's observed delta changes rather than from a model, which can produce different numbers for the same option.

What positive charm should mean on a display

Any chart or tool displaying charm should adopt one convention and explain it plainly. A workable user-facing statement is: Positive displayed charm means delta is expected to increase as calendar time moves forward, assuming the modeled underlying price, volatility, rates, and dividends remain unchanged. Negative displayed charm means delta is expected to decrease.

If the underlying analytics library returns the opposite time derivative, the output should be normalized at the presentation layer while preserving the raw value in diagnostic metadata. The methodology panel should show both the internal mathematical convention and the user-facing interpretation, making the value reproducible without requiring users to manage calculus notation.

Common mistakes when reading charm

• Reading sign without knowing the methodology. Two correct implementations can show opposite-signed numbers. Always confirm what positive means before acting on a reading.
• Treating charm as observed dealer flow. Charm is a model sensitivity. Turning it into a prediction of market orders requires additional assumptions about dealer positions, hedging behavior, and other Greeks. Charm itself specifies none of those assumptions.
• Using annual charm as daily charm. A model may report annualized charm. Dividing by 365 or 252 gives a rough daily figure, but charm itself changes as time changes, so the finite-difference approach (revalue, step time forward, revalue again) is more reliable near expiration.
• Ignoring volatility movement. Post-event implied-volatility collapse can move delta by more than charm predicts from time passage alone. For event-sensitive options, show vanna alongside charm.
• Assuming 0DTE behavior is linear. A charm estimate valid at 10 a.m. may not extrapolate to the close. Recompute through time for intraday 0DTE analysis.
• Treating aggregate charm exposure as a trading strategy. Charm exposure estimates probable hedge-rebalancing direction under model assumptions. It is market-structure context, not a buy or sell signal.