How Options Market Makers Hedge: Quoting, Inventory, Delta and Vega Hedging, and Risk Limits

An options market maker hedges by turning each customer trade into a position it can live with: it buys or sells the underlying stock to neutralize delta, trades other options to neutralize vega, and keeps what is left (mostly gamma, theta, and the spread it earned) inside risk limits it can measure across a grid of price and volatility shocks. The goal is not to predict where the stock goes. The goal is to earn the bid-ask spread on a large number of trades while keeping the inventory that those trades leave behind small enough to survive a bad day.

In short: market makers quote both sides of thousands of option series because exchange rules require it, price those quotes off a volatility surface, and adjust them according to the inventory they already hold. After a fill, they hedge delta with stock or futures almost immediately, hedge vega with other options when the position is large enough, and accept residual gamma and theta. Academic evidence (Garleanu, Pedersen and Poteshman, 2009; Muravyev, 2016) shows that the inventory risk they cannot hedge away is priced into options. Exchange obligations, the Regulation SHO market-maker exception, Regulation T broker-dealer credit, and SEC net capital rules shape how the business is run.

This piece assumes you know what delta, gamma, and vega measure (see option greeks explained) and how listed options trade and clear (see how options markets work). Everything here describes how professional firms operate. It is not a recommendation, and none of it is reproducible without the capital, technology, and exemptions a registered market maker has.

The misconception: “the market maker is betting against you”

When a retail trader buys a call and the market maker sells it, it is tempting to think the firm now wants the stock to fall. In practice, a market maker that sold 100 calls buys shares within seconds so that a small move in the stock does not change its profit and loss. It has no directional view to defend. What it does have is a book of thousands of option positions whose combined risks (gamma, vega, skew, dividends, early exercise) it must manage, and a spread it earned on the way in. Whether the customer later makes or loses money on the call barely matters to a hedged market maker. What matters is whether the edge it charged was larger than the cost of hedging the position it was left holding.

What exchanges require of market makers

Being a registered market maker on a US options exchange is a set of obligations that come with privileges. Using Cboe Exchange, Inc. as the example (its rulebook as updated November 9, 2026):

Rule 5.51 (Market-Maker Obligations) says a market maker’s transactions “must constitute a course of dealings reasonably calculated to contribute to the maintenance of a fair and orderly market.” Ordinarily it must maintain a continuous two-sided market in its appointed classes, deal for its own account when there is “a temporary disparity between the supply of (or demand for) a particular option contract,” compete with other market makers, and update quotes when conditions change.
Rule 5.52(d) (Continuous Electronic Quotes) applies once a market maker trades more than 20% of its contract volume electronically in an appointed class in a calendar quarter. From the next quarter it must quote continuously for 90% of the time it is required to quote, in 60% of the series of its appointed classes, excluding adjusted series, intraday add-on series on the day they are added, Quarterly Option series, and series expiring in more than 270 days. Compliance is measured monthly across all appointed classes collectively.
Rule 5.55 (LMMs) requires a Lead Market-Maker, during regular trading hours, to quote continuously 90% of the time in at least the lesser of 99% of the non-adjusted series or all but one call-put pair, and to enter opening quotes promptly.
Rule 5.52(c) (Two-Sided Quotes) requires that a bid be matched by an offer. Cboe filed in February 2026 (SR-CBOE-2026-013) to add a maximum bid-ask differential of $5, and in August 2026 (SR-CBOE-2026-074, SEC Release No. 34-106208) to set the limit at $5 for series expiring in 270 days or less and $15 beyond that. The rulebook states the SR-CBOE-2026-074 amendments become operative on December 7, 2026.

Quotes are also firm under Rule 5.52(a) and Rule 602 of Regulation NMS: a market maker must honor its displayed price for its displayed size, subject to narrow exceptions such as system malfunctions. That is the core business risk. A firm that shows a quote in 60% of series all day is offering free options to anyone who can trade faster than it can update.

The flow they face

A market maker’s book is shaped by who trades against it.

Index options. Garleanu, Pedersen and Poteshman (“Demand-Based Option Pricing,” Review of Financial Studies, 2009) used data identifying the positions of dealers and end users and found that end users are net long index options, especially out-of-the-money puts, which leaves dealers net short them. Their model shows that when intermediaries cannot hedge perfectly, demand for an option raises its price in proportion to the variance of the part that cannot be hedged, which helps explain why index puts look expensive and why the skew is steep.
Single-stock options. The same paper finds end users are net short single-stock options on average, the reverse of the index pattern.
Retail flow. Bryzgalova, Pavlova and Sikorskaya (“Retail Trading in Options and the Rise of the Big Three Wholesalers,” Journal of Finance, 2023) report that retail trading recently reached over 60% of total options market volume, that nearly 90% of payment for order flow in options comes from three wholesalers, and that retail investors favor cheap, short-dated options with an average quoted bid-ask spread of 12.6%, and lose money on average. The mechanics of routing and payment for order flow are covered in options order flow and PFOF.

Exchanges tag each trade with a capacity. The Options Clearing Corporation’s origin codes, as described by Ni, Pearson and Poteshman (Journal of Financial Economics, 2005), are C for public customers, F for firm proprietary traders, and M for market makers. That tagging is what lets researchers, and the exchanges themselves, reconstruct how much inventory market makers are carrying.

Quoting: fair value, edge, and inventory skew

A market maker does not quote a price first. It quotes a volatility. Its pricing engine holds a fitted implied volatility surface for each underlying (by strike and expiry), converts the fitted volatility for each series into a theoretical value, and then sets a bid below and an offer above that value.

Three things move the quote around the theoretical value:

Edge. The half-spread compensates for adverse selection (some counterparties know more) and hedging cost. On a 30-day at-the-money call on a $50 stock at 30% implied volatility, the theoretical value is $1.7956. Selling at 30.5 volatility instead (a price of $1.8241) earns $2.85 per contract. Selling at 31 volatility ($1.8526) earns $5.70.
Inventory. If the firm is already short a lot of vega in a name, it raises both its bid and its offer in volatility terms so it is more likely to buy and less likely to sell. Muravyev (“Order Flow and Expected Option Returns,” Journal of Finance, 2016) shows that the inventory risk faced by market makers has a first-order effect on option prices, and that the inventory component of price impact is larger than the information component.
Correlated risk. If every customer is buying upside calls across one sector at once, the firm’s exposures add up across names. Pricing engines aggregate risk to the sector or index level and skew quotes for the whole group.

Worked example: hedging a customer’s call purchase

Maple Widgets Inc. trades at $50. A customer buys 100 contracts of the 30-day $50 call from a market maker at the theoretical value. We use Black-Scholes with a 4% interest rate and no dividends; each contract covers 100 shares, so the position is 10,000 option-shares.

Option	Implied vol	Price	Delta	Gamma (per $1)	Vega (per vol point)	Theta (per day)
30-day $50 call	30%	$1.7956	0.5324	0.0925	$0.0570	-$0.0315
91-day $50 call	29%	$3.1264	0.5562	0.0546	$0.0986	-$0.0185

Step 1: what the sale leaves behind

Short 10,000 option-shares of the 30-day call, the market maker collected $17,956 and now has:

Delta: -5,324 shares (it loses about $5,324 per $1 rise in the stock)
Gamma: -925 shares per $1 (its delta gets more negative as the stock rises)
Vega: -$570 per volatility point
Theta: +$315 per day (time decay works in its favor)

Step 2: delta hedge with stock

It buys 5,324 shares. Now a small move does little, but a larger one hurts in either direction because gamma is negative:

Instant stock move	-$2	-$1	+$1	+$2
P&L, delta hedged only	-$1,864	-$466	-$454	-$1,772

That is the short-gamma shape: about one half times 925 times the move squared. The theta of +$315 a day pays for it only if moves stay small. The break-even daily move is the square root of (2 x 315 / 925), or about $0.83, which is 1.65% of the stock price. The 30% implied volatility the call was sold at corresponds to a daily move of 30% divided by the square root of 365, about 1.57%. The two match closely, as they should: selling an option at a given implied volatility is a bet that realized volatility will come in below it. The P&L of this trade over time is the subject of delta hedging and gamma scalping.

Step 3: vega hedge with a longer-dated option

A $570-per-point vega exposure means a 5-point jump in implied volatility costs $2,850 even with delta hedged. If the firm’s limits do not allow that, it buys vega elsewhere on the curve. Buying 58 contracts of the 91-day $50 call (5,800 option-shares, costing $18,133) brings vega to about +$2 per point. The extra calls carry delta of their own, so the share hedge falls from 5,324 to 2,098 shares.

The book after both hedges:

Risk	Delta hedged only	Delta and vega hedged
Vega per vol point	-$570	+$2
Gamma per $1	-925	-608
Theta per day	+$315	+$208
P&L if all vols rise 1 point	-$570	+$2
P&L if all vols rise 5 points	-$2,850	+$10
P&L on an instant $2 drop	-$1,864	-$1,218

Two lessons come out of that table. First, a vega hedge with a longer option also cuts gamma (from -925 to -608) but cannot remove it, because a 91-day option has less gamma per unit of vega than a 30-day option. Gamma is a short-dated risk, and only short-dated options hedge it well. Second, the vega hedge works only for parallel shifts. If 30-day volatility rises 3 points while 91-day volatility rises only 1, the hedged book loses $1,138. If the shift goes the other way (30-day up 1, 91-day up 3), it gains $1,146. The firm has swapped outright vega risk for term-structure risk, which is smaller but real, and has to be limited separately. The same logic applies across strikes, where a vega hedge at one strike leaves exposure to changes in skew (see volatility surface and skew).

Risk limits: grids, not single numbers

Delta, gamma, and vega are local measures. Professional risk management revalues the whole book across a grid of large moves, because that is where losses come from.

The SEC’s own net capital rule works the same way. Appendix A to Rule 15c3-1 (17 CFR 240.15c3-1a) computes capital charges on options positions in a market maker’s account from theoretical gains and losses at 10 equidistant valuation points across a range of underlying moves: plus or minus 15% for equity securities with a ready market, narrow-based indexes, and non-high-capitalization diversified indexes, and plus or minus 10% for high-capitalization diversified indexes. The theoretical prices must come from an approved options pricing model rather than a firm’s proprietary one.

Revaluing our Maple Widgets book (holding volatility fixed) across that kind of grid:

Stock move	-15%	-9%	-3%	+3%	+9%	+15%
Delta hedged only	-$22,468	-$9,049	-$1,050	-$1,011	-$8,184	-$19,678
Delta and vega hedged	-$13,517	-$5,759	-$688	-$663	-$5,244	-$12,135

The worst point on the grid, not the delta, is the number that consumes capital. A desk’s internal limits typically add volatility shocks (with short-dated volatility moving more than long-dated), skew shocks, and stress scenarios replayed from history, and they cap each trader’s book by the worst loss across the grid.

The exemptions that make the business work

Market makers carry obligations other traders do not, and the regulatory system gives them a handful of specific accommodations in return. Three matter most for hedging:

Short sale locates. Regulation SHO Rule 203(b)(1) requires a broker-dealer to locate shares before a short sale. Rule 203(b)(2)(iii) exempts “short sales effected by a market maker in connection with bona-fide market making activities.” A market maker that sells puts to customers and must short stock to hedge can do so without first arranging a borrow.
Close-out timing. Rule 204(a)(3) gives a fail to deliver “attributable to bona fide market making activities by a registered market maker, options market maker, or other market maker obligated to quote” until the beginning of regular trading hours on the third consecutive settlement day after the settlement date to be closed out, rather than the standard next-day close-out.
Financing. Regulation T, 12 CFR 220.7(g)(5), lets a creditor extend credit “with good faith margin” to a member of a national securities exchange or registered broker or dealer “to finance its activities as a market maker or specialist.” Hedging stock positions are therefore financed on terms negotiated with the clearing firm, not at the 50% initial margin a retail account faces.

These are narrow. The locate exception covers bona fide market making, not directional short selling dressed up as hedging, and the close-out accommodation is tied to the same test. The capital rule ensures that the leverage the exemptions allow is backed by capital sized to the grid losses above.

What a hedged book is still exposed to

Gaps. Delta hedges are rebalanced in continuous trading. An earnings announcement, a trading halt, or an overnight event moves the stock before the hedge can be adjusted, and the short-gamma loss is realized on the whole gap.
Volatility regime changes. A book that is vega neutral on the day can be badly exposed when short-dated volatility jumps far more than long-dated, as the term-structure example shows. In a sell-off, skew also steepens, so a book hedged with at-the-money options loses on its short out-of-the-money puts.
One-sided flow. When customers all want the same thing (upside calls in a popular stock, or index puts in a sell-off), the market maker cannot find offsetting customer trades and must hedge in the stock or with other dealers, at a cost it passes on through wider quotes. Garleanu, Pedersen and Poteshman’s evidence is that this shows up in prices.
Expiration and assignment. Near expiry, gamma on at-the-money options becomes very large, and a market maker short calls on a dividend-paying stock can be assigned early. Pin risk (not knowing whether a short option will be assigned when the stock closes at the strike) can leave an unhedged stock position over a weekend. See what happens at expiration.
Technology and stale quotes. Because quotes are firm and obligations are continuous, a pricing error or a delayed volatility update is filled by faster traders at the firm’s expense, across every series it quotes.

How professionals summarize the job

Inside a market-making firm the job is usually described in three layers. The pricing layer keeps the volatility surface fitted and the quotes competitive. The hedging layer removes delta continuously and vega when it crosses limits, using whatever instrument is cheapest at the moment (stock, futures, ETFs, other options). The risk layer sets limits by grid loss, watches concentrations across correlated names, and decides how much inventory the firm will warehouse in exchange for edge. The firm makes money if the spread it collects across millions of trades exceeds what it loses hedging the residual gamma and the moves it could not hedge in time.

The short version

Market makers are obligated by exchange rules to quote continuously and honor their quotes, which means they absorb whatever customers want to trade. They neutralize delta with stock right away, neutralize vega with other options when the exposure is large, and carry the residual gamma, theta, and term-structure risk inside capital-based limits that look much like the SEC’s 10-point valuation grid. The research record shows that the risk they cannot hedge is priced into options, which is one reason index puts are expensive and retail traders pay wide spreads on short-dated contracts. The privileges they get in return (locate and close-out exceptions under Regulation SHO, good faith financing under Regulation T) are tied to bona fide market making, and are the reason this is a business for registered firms rather than for individuals.