Charm: The Clock Is a Trader
Charm is delta’s sensitivity to time, holding other inputs constant. It estimates conditional potential hedge pressure - not a scheduled dealer order - when spot, IV, and assumed inventory remain near their current values.
The greek that trades on an empty tape
Most flow needs something to happen. Gamma needs a price move. News needs news. Charm needs nothing but the clock. It is delta's sensitivity to the passage of time - ∂Δ/∂t, holding spot and vol constant - and time passes whether or not the tape does anything. A dealer can sit in front of a market that has not moved a tick for ninety minutes and still find the modeled delta of their book quietly decaying, nudging them to sell stock the whole way to keep the hedge in line.
That is what makes charm strange and, once you see it, obvious. The clock is a trader - a modeled one. It never stops and it never changes its mind, and holding other inputs fixed it points the same way all session. The only questions are which direction it pushes, how big the push gets, and whether the day's actual spot and vol moves leave it intact.
This piece is the mechanical companion to our broader Vanna and Charm explainer. That article frames charm as one input into the end-of-day read; this one goes under the hood - where the drift comes from, why it accelerates, and how you can put a dollar figure and a deadline on it before it happens.
Where the drift comes from
Delta is closely related to - but not exactly - the risk-neutral probability that an option finishes in the money. A 0.30-delta call is roughly the market's way of saying there is something like a 30% chance this expires with value. That implied probability is a live estimate, and as expiry approaches it has to collapse toward a verdict: either the option finishes in the money (delta → 1) or it does not (delta → 0), with only the knife-edge at-the-money case left undecided into the bell.
Charm is the speed of that collapse. Watch a slightly out-of-the-money option through an afternoon with spot pinned:
- This morning it had delta 0.35 - real chance of paying off.
- By lunch, with less time on the clock and spot unchanged, delta 0.28.
- By 3pm, delta 0.18.
- Into the bell, delta sliding toward 0.
Nothing moved. The option's delta fell by half anyway, purely because the runway shortened. Every one of those steps is a change in the hedge ratio, and every change forces the dealer holding that option to adjust their stock. That adjustment is charm flow.
In-the-money options do the mirror image, firming from 0.80 toward 1.00 as their outcome becomes a near-certainty. The book's net charm is modeled as the sum across every strike, weighted by how much open interest sits there and, under the positioning convention, which side the dealer is assumed to be on - actual dealer inventory is not directly observable from public option data.
Why it accelerates into the close
Charm is not constant through the day. The rate of delta decay is small when there is plenty of time left and grows as expiry closes in - it is largest in the final hour and largest of all in the final minutes, for the near-the-money strikes that still have a live verdict pending. On a chain dominated by same-day expiries, which is now the default for SPX, the bulk of the day's charm flow is compressed into the last sixty to ninety minutes.
This is the mechanical reason the "into-the-close drift" is a recurring phenomenon and not a chart superstition. It is not that traders get emotional at 3pm. It is that the mathematics of delta decay puts most of its modeled force there, and dealers hedging that decay tend to do the bulk of their adjusting late - even those working hedge bands rather than trading continuously. The modeled flow ramps because the greek ramps.
The live Charm into Close chart draws exactly this: it holds spot and vol fixed, walks the clock forward to the bell, and plots the cumulative stock the dealer book is modeled to trade at each step. The curve starts at zero at the current moment and bends away from zero as the afternoon runs - steepest at the end, because that is where charm lives.
A forecast with a deadline
Here is the property that makes charm uniquely useful, and it is the thing you will not find in a standard greeks writeup.
Every other dealer flow is contingent. Gamma flow depends on a spot move that may or may not come. Vanna flow depends on a vol shift you cannot schedule. The charm component depends only on time, and time is the one input that is going to do exactly what you expect. At 9:35 in the morning, holding spot and vol at their current levels, you can estimate how much stock time decay alone would have dealers buy or sell by the close. You get a modeled size and direction for a large flow hours before it plays out - conditional on the rest of the world sitting still, which it never quite does.
That is a forecast with a deadline. The forecast has a condition attached - "if spot and vol hold near here" - and they rarely hold perfectly, so the real close blends charm with whatever gamma and vanna the day's moves produce. But the charm component is estimable in advance in a way almost nothing else in markets is. It is the closest thing to a predictable, calendar-driven flow the market offers - a modeled pressure set by the clock, not a scheduled order anyone is obliged to fill.
This is essentially the number the Charm-into-Close bulletin surfaces before the open: time decay alone is modeled to have dealers buy/sell \$X into the close if the underlying holds here. A deadline, a direction, and a dollar figure, all estimable at dawn.
Putting a number on it
Suppose it is a Friday with heavy 0DTE positioning in SPY, spot at 560. The dealer book carries the day's same-day options, and as the clock runs to the bell every one of them has to resolve - finish in the money or expire worthless - so the deltas dealers are hedging swing hard. Reprice the whole book at the close with spot held at 560 and the total modeled time-driven flow on a heavy 0DTE day can run into the billion-dollar range. That is the number the live Charm-into-Close chart plots, and it is what "how much dealers are modeled to trade by the close" means in practice.
Two honest caveats on that headline figure. First, most of it is the same-day options resolving at the bell - a pin effect that hinges on exactly where spot settles, not smooth decay - so it is large and it is sensitive to spot. Second, the pure charm drift, the part that is genuinely time decay of the surviving book rather than the expiry event, is a fraction of it: on the order of a few hundred million, accumulating steadily through the afternoon. The dashboard shows both - the full close flow and the charm-only drift - because they answer different questions, and the smaller charm-only number is the cleaner, less spot-sensitive read.
Flip the book's composition and the same clock points to buying instead. Charm does not have an inherent direction the way gravity has "down"; the direction is set by which strikes the dealers are modeled short and long. What tends to hold is the timing: whatever the sign, the modeled flow concentrates into the close, and that concentration is estimable at 9:35 that morning.
How to actually use it
A short discipline:
- Read the sign at the open. The charm-into-close figure tells you which way the clock is pushing today and roughly how hard. That is regime context, not an entry.
- Watch for confluence. When charm points the same direction as the gamma magnet - the heavy strike price is drifting toward - the two forces stack and the into-close drift is at its cleanest. When they disagree, expect chop, not drift.
- Respect the "if spot holds" condition. Charm is a conditional forecast. A 1% move mid-afternoon hands the wheel to gamma and can swamp the charm read entirely. The time-only estimate is easier to isolate on quiet days, but ZeroGEX has not presented that observation as a calibrated performance result.
- Discount it when vol is expanding. On a genuinely volatile day, gamma reactions dominate and the tidy charm drift becomes noise.
Charm holds spot and IV constant by definition. Real portfolios contain offsets across strikes and expiries; dealers may use hedge bands, futures, or other options, and new trades can change inventory. The estimate therefore represents potential time-only hedge pressure, not proof of when or how much dealers will trade.
For the parent concept, see Delta and Its Three Children; for the vol-driven sibling, see Vanna: How Falling IV Can Change Dealer Hedging; and to watch the into-close curve build in real time, open the live Forced Flow page.
Educational content only - none of the above is a trade recommendation.
