Market Makers & Inventory Risk

You’ve met the market maker as the entity that earns the spread. Now meet its nightmare. A maker quotes a price to buy and a price to sell, all day, on thousands of names — which means it’s constantly accumulating positions it never wanted. Every fill leaves it longer or shorter, holding risk until it can offload. Managing that inventory — not predicting prices — is the maker’s actual job, and the tools it uses (skewing quotes, setting limits, hedging) explain why the book behaves the way it does. This is the supply side of liquidity, demystified.

The market maker’s business

Analogy. A market maker is a foreign-exchange kiosk at the airport, scaled up and sped up a billion times. The kiosk quotes “I’ll buy euros at 1.05, sell them at 1.12” and earns the spread on travelers passing through in both directions. Its danger isn’t predicting the euro — it’s getting flooded with euros nobody wants (a big long position) right before the euro drops. It survives by keeping its euro pile near zero and adjusting its quotes to attract whichever side it needs.

Definition. A market maker continuously quotes a two-sided market — a bid (price to buy) and an ask (price to sell) — standing ready to trade with anyone, and earning the spread on the round trips where it buys low and sells high to different counterparties. It is not directionally betting; ideally it ends each moment near flat (zero inventory), having simply harvested spread from two-way flow. Its profit per round trip is small; its volume is gigantic; and its survival depends entirely on not getting buried under inventory.

There are two broad flavours:

• Designated market makers (DMMs) / specialists — firms with a formal obligation to quote continuously in specific names (and quoting requirements: maximum spread, minimum size), in exchange for privileges. They provide liquidity even when they’d rather not — a backstop.
• Opportunistic / high-frequency makers — firms with no obligation; they quote when it’s profitable and pull quotes when it’s dangerous. They supply most liquidity in good times but can vanish in a storm (a key fragility — recall flash crashes).

Inventory: the core risk

Definition. Inventory is the net position a maker is holding at any moment: long if its recent buys exceed its sells, short if the reverse. Inventory risk is the danger that the price moves against that inventory before the maker can flatten it. A maker that just bought 10,000 shares (because sellers hit its bid) is long 10,000 and exposed to a price drop; one that just sold is short and exposed to a rally. The spread it earned is a thin cushion against a position that can move far more than the spread.

The animated simulator below makes this visceral. A maker quotes both sides; random order flow pushes its inventory long or short, and it random-walks. The dashed lines are the maker’s inventory limits — hit them and the maker is forced into a costly unwind. Run a session with No skew and watch inventory wander freely, sometimes slamming into a limit. Then crank the skew strength and watch the path get pulled back toward flat.

Skewing quotes: leaning against the position

Here’s the maker’s cleverest trick. It doesn’t just sit and pray its inventory stays balanced — it skews (shades) its quotes to actively pull inventory back toward flat.

Analogy. A shop overstocked on umbrellas doesn’t keep pricing them the same as everyone else — it puts them on sale (lower its ask) to sell them faster and quietly raises what it’d pay for more (lower its bid) so it stops buying umbrellas it doesn’t need. A long maker does exactly this with shares.

Definition. When a maker gets long (too many shares), it skews its quotes down: it lowers both its bid and its ask. The lower ask makes its shares cheaper, attracting buyers to take inventory off its hands; the lower bid makes it less attractive for sellers to dump more shares on it. When it gets short, it skews up to attract sellers and discourage buyers. Skewing trades a little expected spread capture for much less inventory risk — it makes the inventory path mean-revert toward flat instead of wandering freely. That’s the pull you see bending the line back in the simulator.

Worked example — skewing in cents

The fair mid is $100.00 and the maker normally quotes a 10¢ spread: bid $99.95 / ask $100.05. Now it’s gotten long 8,000 shares and wants to shed them. It skews its quotes down by, say, 3¢:

• New quotes: bid $99.92 / ask $100.02 (both shifted down 3¢, spread unchanged at 10¢).
• The $100.02 ask is now below the unshaded $100.05, so buyers find its shares attractive → it sells, reducing its long.
• The $99.92 bid is below the unshaded $99.95, so sellers prefer to hit other makers → it buys less, stopping the long from growing.

The maker is essentially saying “I’ll take a slightly worse price to get flat faster.” Note it didn’t widen the spread — it shifted the whole two-sided quote. The width compensates for risk; the skew steers inventory.

Inventory limits and forced unwinds

Definition. No maker lets inventory grow without bound. It sets inventory limits (position caps), and as inventory approaches a limit the maker skews harder and quotes less aggressively on the dangerous side. If a limit is breached anyway — say a torrent of one-sided flow — the maker must unwind: aggressively trade out of the position by crossing the spread itself (becoming a taker) or hedging immediately, accepting slippage and fees to kill the risk. A forced unwind is expensive: the maker pays the very costs it normally earns. Avoiding forced unwinds is most of the art.

This is also why makers widen spreads or pull quotes in stress. When volatility spikes, inventory becomes far more dangerous to hold (a position can move violently before you flatten it), so the maker demands more compensation (wider spread) or simply stops quoting on the risky side. The liquidity you counted on can evaporate exactly when you need it — the supply-side reason behind the “displayed depth vanishes” lesson from earlier.

Hedging: neutralizing risk without unwinding

Sometimes a maker can’t easily offload the exact shares it’s holding, but it can neutralize the risk with a related instrument.

Definition. Hedging offsets inventory risk by taking an opposing position in a correlated instrument — an index future, an ETF, an option, or a basket — rather than unwinding the original position directly. A maker long a basket of tech stocks might short a tech-index future to cancel the broad-market risk while it works out of the individual names. Hedging is why easily-hedged assets have tight spreads: if the maker can instantly lay off the risk, its inventory cost is low, so it quotes tight. An asset with no good hedge (an illiquid single name, an exotic) carries pure inventory risk and pays for it with a wide spread.

This closes the loop with the spread decomposition: the inventory component of the spread is small precisely when hedging is cheap and the asset is liquid, and large when the maker is stuck bearing un-hedgeable risk.

Putting it together

A market maker rents out immediacy: it quotes both sides, earns the spread on two-way flow, and never wants to bet on direction. Its existential risk is inventory — every fill leaves it long or short, exposed until it flattens. It fights inventory by skewing its quotes (shading both toward the side that flattens it, making the position mean-revert), capping it with inventory limits (a breach forces a costly unwind as the maker crosses the spread itself), and hedging un-offloadable risk with correlated instruments. Easily-hedged, liquid assets carry tiny inventory risk and quote tight; un-hedgeable, illiquid ones pay for the risk with wide spreads. And because inventory is far more dangerous in volatile markets, makers widen or pull quotes in stress — the supply-side reason liquidity vanishes when you need it most.

Key Takeaways