Market Making - How Liquidity Providers Work

TL;DR

Market makers are traders who provide liquidity by continuously posting both buy (bid) and sell (ask) orders. They profit from the spread (difference between bid and ask) while taking on inventory risk.

• Maker = Adds liquidity (posts limit orders that rest on book)
• Taker = Removes liquidity (executes against resting orders)
• MMs earn rebates for providing liquidity, takers pay fees
• MMs must manage inventory risk - holding assets that move against them
• Exchanges need MMs for tight spreads and deep liquidity

---

What Does a Market Maker Actually Do?

A market maker's job is simple in concept: always be willing to buy and sell.

Without Market Maker:
  Order Book:
    Asks: $105 (10 units) ← Seller wants $105
    Bids: $95 (10 units)  ← Buyer wants $95

  Spread: $10 (10%!)

  You want to buy NOW? Pay $105.
  You want to sell NOW? Get $95.
  Terrible prices for everyone.

With Market Maker:
  Order Book:
    Asks: $100.05 (1000 units) ← MM willing to sell
    Bids: $99.95 (1000 units)  ← MM willing to buy

  Spread: $0.10 (0.1%)

  You want to buy NOW? Pay $100.05 (much better!)
  You want to sell NOW? Get $99.95 (much better!)

The market maker provides a service: immediate execution at reasonable prices. In exchange, they capture the spread.

---

Who Are the Major Crypto Market Makers?

The crypto market making space is dominated by a mix of TradFi veterans (firms from traditional finance) and crypto-native shops.

The Big Players

Firm	Profile
Wintermute	~$15B daily volume, 65+ venues, crypto-native (2017). Hybrid culture: TradFi execution speed + crypto startup agility.
Cumberland DRW	Subsidiary of DRW (TradFi trading giant). Institutional focus since 2014. Known for block trades, TWAP execution, and deep bank relationships.
Jump Crypto	Arm of Jump Trading. Quantitative HFT, ultra-low latency, research-heavy. One of the most sophisticated technical operations.
GSR Markets	OTC + exchange MM. Strong in token launches and derivatives.
DWF Labs	MM + venture investor. Provides liquidity as part of token ecosystem deals.
Flow Traders	European firm, ETF specialists expanding into crypto.
Virtu Financial	TradFi HFT giant. Handles significant retail equity flow via PFOF. Expanding in crypto.

TradFi vs Crypto-Native

Aspect	TradFi Firms (Jump, DRW, Virtu)	Crypto-Native (Wintermute, GSR)
Experience	Decades in equities/futures	Built for crypto from day 1
Infrastructure	Massive, battle-tested	Purpose-built for 24/7 markets
Regulatory	Deep compliance expertise	More agile, less legacy process
DeFi	Slower adoption	Native integration
Culture	Traditional finance	Startup / crypto community

Both types compete on the same exchanges, often in the same order books.

---

How Market Makers Profit

The Spread

The spread is the difference between the best bid and best ask. Market makers profit by:

1. Buying at the bid price
2. Selling at the ask price
3. Pocketing the difference

MM posts:
  BID: Buy 100 BTC @ $99,950
  ASK: Sell 100 BTC @ $100,050

Scenario: Both sides fill
  - Bought 100 BTC for $9,995,000
  - Sold 100 BTC for $10,005,000
  - Profit: $10,000 (the spread × quantity)

Spread capture = (Ask - Bid) × Quantity filled on both sides

The Reality: It's Not That Simple

In practice, both sides rarely fill equally. The MM ends up with inventory:

MM posts:
  BID: Buy 100 @ $99,950
  ASK: Sell 100 @ $100,050

What actually happens:
  Hour 1: 80 people buy from MM (MM sells 80)
  Hour 2: 20 people sell to MM (MM buys 20)

  Net: MM is now SHORT 60 BTC

  If price goes UP: MM loses money on short position
  If price goes DOWN: MM makes money on short position + spread

This is inventory risk - the core challenge of market making.

---

Maker vs Taker: The Fee Model

Why Exchanges Incentivize Makers

Exchanges want liquidity. More liquidity = tighter spreads = better experience for everyone. So they:

• Reward makers (rebates) - "Thanks for adding liquidity"
• Charge takers (fees) - "You're consuming liquidity"

Typical Fee Structure:

| Role  | Action                      | Fee        |
|-------|----------------------------|------------|
| Maker | Post limit order on book   | -2 bps (rebate) |
| Taker | Execute against resting    | +5 bps (fee)    |

bps = basis points (1 bps = 0.01%)

Example:
  You BUY 1 BTC @ $100,000 as TAKER
  Fee: $100,000 × 0.0005 = $50

  MM SOLD 1 BTC @ $100,000 as MAKER
  Rebate: $100,000 × 0.0002 = $20

  Exchange keeps: $50 - $20 = $30

How to Be a Maker vs Taker

Order Book: Best ask $100.10, Best bid $100.00

MAKER orders (add liquidity):
  Limit BUY @ $99.90  → Goes on book (below best ask)
  Limit SELL @ $100.20 → Goes on book (above best bid)

TAKER orders (remove liquidity):
  Limit BUY @ $100.20  → Executes immediately (crosses spread)
  Limit SELL @ $99.90  → Executes immediately (crosses spread)
  Market BUY/SELL      → Always taker

Post-Only modifier:
  Forces order to be maker-only
  Rejected if it would cross the spread

---

Fee Tiers and Volume Rebates

Exchanges reward high-volume traders and market makers with better rates:

Volume-Based Fee Tiers

| Tier | 30-Day Volume  | Maker Fee | Taker Fee |
|------|----------------|-----------|-----------|
| 1    | < $1M          | 2 bps     | 5 bps     |
| 2    | $1M - $10M     | 1.5 bps   | 4 bps     |
| 3    | $10M - $50M    | 1 bps     | 3 bps     |
| 4    | $50M - $100M   | 0 bps     | 2.5 bps   |
| 5    | > $100M        | -1 bps    | 2 bps     |

At Tier 5: Makers get PAID to provide liquidity

Market Share Rebates

Beyond volume, some exchanges reward market share - what percentage of all maker volume you provide:

| Market Share | Additional Rebate |
|--------------|-------------------|
| > 1%         | +0.5 bps          |
| > 5%         | +1.0 bps          |
| > 10%        | +1.5 bps          |

If you're 10% of all maker volume:
  Base maker fee: 0 bps
  Market share rebate: +1.5 bps
  Net: You receive 1.5 bps on every fill

This incentivizes MMs to provide consistent, deep liquidity.

---

Exchange-Market Maker Relationships

Exchanges and market makers have a symbiotic relationship. Understanding this deal helps explain why liquidity exists.

The Symbiotic Deal

What Exchange Offers MM:              What MM Provides Exchange:
├─ Fee rebates (negative fees)        ├─ Tight spreads (better UX)
├─ Designated MM programs             ├─ Deep liquidity (less slippage)
├─ Colocation / low-latency access    ├─ 24/7 continuous quoting
├─ Early listing access               ├─ Backstop during liquidations
├─ Higher API rate limits             ├─ Volume (attracts more traders)
└─ Revenue sharing deals              └─ Market stability

Why this works: Exchange makes money on taker fees. More liquidity → tighter spreads → more takers → more revenue. Paying MMs to provide liquidity is profitable for the exchange.

Designated Market Maker (DMM) Programs

Many exchanges have formal DMM programs with specific obligations:

Requirement	Typical Value	Why It Matters
Max spread	10-50 bps from mid	Ensures tight quotes
Min size	$50K-$500K notional per side	Ensures depth
Uptime	90%+ of market hours	Ensures reliability
Markets covered	Multiple symbols	Ensures breadth

In return, DMMs receive:

• Negative maker fees (get paid to trade)
• Cash rebates based on performance
• Direct relationship with exchange
• Priority support and API access

Example (BitMEX DMM Program):

• Regular tier: Basic rebates for meeting quote requirements
• Senior tier: Larger MMs get better rebates + dedicated support
• Daily cash payments calculated from quoting metrics
• 90%+ uptime requirement across 24-hour markets

Why Exchanges Want Multiple MMs

Having several competing MMs is better than one:

Single MM:                    Multiple MMs:
├─ Can widen spreads          ├─ Competition keeps spreads tight
├─ Single point of failure    ├─ Redundancy if one goes down
├─ Price manipulation risk    ├─ Better price discovery
└─ Dependency                 └─ Negotiating leverage for exchange

Most liquid markets have 5-20+ active market makers competing.

The MM's Perspective

MMs evaluate exchanges on:

1. Volume - More trades = more spread capture opportunity
2. Fee structure - Better rebates = higher margins
3. Latency - Can they colocate? API speed?
4. Counterparty quality - Is the flow toxic or retail?
5. Reliability - Does the exchange go down? Matching engine bugs?
6. Regulatory - Legal risk of operating on this exchange?

A top-tier MM like Wintermute might be active on 50+ venues, allocating capital based on these factors.

---

Market Maker Strategies

1. Simple Spread Capture

Post symmetric orders around mid-price:

Mid price: $100.00

MM posts:
  BID: $99.95 (5 cents below mid)
  ASK: $100.05 (5 cents above mid)

Spread: 10 cents
If both fill: 10 cents profit per unit

2. Skewed Quotes (Inventory Management)

Adjust prices based on current inventory:

MM is LONG 50 BTC (wants to sell)

Instead of symmetric:
  BID: $99.95, ASK: $100.05

Skew towards selling:
  BID: $99.90, ASK: $100.02

- Less aggressive bid (less likely to buy more)
- More aggressive ask (more likely to sell)
- Reduces inventory over time

3. Volatility-Based Spreads

Widen spreads when risk is high:

Normal conditions:
  Spread: 5 cents

High volatility / news event:
  Spread: 50 cents

Why? During volatility:
  - Prices move fast
  - MM can get "picked off" by informed traders
  - Wider spread compensates for higher risk

4. Cross-Exchange Arbitrage

Use prices from other venues to set quotes:

Binance BTC: $100,000
Coinbase BTC: $100,050
Backpack BTC: ???

MM on Backpack:
  - Sees Binance/Coinbase prices
  - Quotes $100,020 bid / $100,030 ask
  - If someone buys at $100,030, MM hedges by buying on Binance at $100,000
  - Locks in $30 profit regardless of where price goes

---

The Risks of Market Making

1. Inventory Risk

The biggest risk. You accumulate positions that move against you:

MM starts flat (no position)

Day 1: Market rallies, everyone buys
  - MM sells 100 BTC at various prices
  - MM is now SHORT 100 BTC

Day 2: Market continues up 5%
  - MM's short position loses 5% = $500,000 loss
  - Spread captured: maybe $10,000

Net: -$490,000

The spread profit is tiny compared to directional risk.

2. Adverse Selection

Getting "picked off" by informed traders:

MM is quoting BTC:
  BID: $99,950
  ASK: $100,050

Breaking news: Major company announces BTC purchase
Price should be $105,000

Informed trader:
  - Sees news before MM can react
  - Buys from MM at $100,050
  - MM sold at $100,050, now worth $105,000
  - MM loses $4,950 per BTC

This is why MMs widen spreads during news events.

3. Technical Risk

System failures are catastrophic:

MM system goes down:
  - Orders still resting on book
  - Price moves 10%
  - Orders get filled at terrible prices
  - By the time system recovers: massive losses

MMs need:
  - Redundant systems
  - Kill switches
  - Automatic order cancellation on disconnect

4. Funding Rate Risk (Perpetuals)

In perpetual futures, holding inventory costs money:

MM is SHORT 100 BTC (inventory from selling to buyers)
Funding rate: +0.01% every 8 hours (longs pay shorts)

Good: MM receives funding while short

But if funding flips negative:
  MM PAYS funding while short
  Holding inventory becomes expensive

---

Backstop Liquidity Providers

Some exchanges have designated market makers who agree to provide liquidity in extreme conditions (liquidations). Backpack calls these Backstop Liquidity Providers.

How Backstop Works

Normal liquidation:
  Account MF drops below MMF
  → Engine places IOC orders on the book
  → Regular MMs (and anyone) can fill them

Backstop liquidation (when book is thin):
  Account MF drops further (below mf_auto_close)
  → Engine forces designated backstop MMs to take the position
  → Guaranteed fill, but at a price favorable to MM

Backstop MM Configuration

BackstopLiquidityMarketMaker {
    account_id: ...,
    symbol: "SOL_USDC_PERP",
    capacity_notional: $500,000,      // Current available capacity
    max_capacity_notional: $1,000,000, // Maximum capacity
    refresh_rate_notional_hour: $100,000,  // Replenishes over time
    refresh_rate_notional_minute: $10,000,
}

Key concepts:

• Capacity - How much notional the MM can absorb right now
• Refresh rate - Capacity replenishes over time (can't be exhausted instantly)
• Max capacity - Upper limit on exposure

Why MMs Agree to Backstop

1. Favorable pricing - Backstop fills happen at a discount (MM gets a good entry)
2. Rebates - Additional compensation for backstop role
3. Relationship - Access to other exchange benefits
4. Market stability - A functioning market benefits everyone

---

Market Making Economics

Revenue Sources

1. Spread capture
   $0.10 spread × 1000 BTC/day = $100/day

2. Maker rebates
   $100M volume × 1 bps = $10,000/day

3. Funding arbitrage (perps)
   Collect funding on inventory when favorable

4. Cross-exchange arb
   Price discrepancies between venues

Cost Sources

1. Inventory losses
   Directional moves against position

2. Adverse selection
   Getting picked off by informed traders

3. Infrastructure
   Servers, connectivity, monitoring

4. Capital cost
   Money tied up in positions and collateral

5. Hedging costs
   Fees on other exchanges when hedging

The Math

Successful MM:
  Daily volume: $100M
  Spread: 2 bps average capture
  Gross spread revenue: $20,000

  Maker rebate: 1 bps
  Rebate revenue: $10,000

  Inventory P&L: -$5,000 (managed well)
  Infrastructure: -$2,000

  Net profit: $23,000/day

Unsuccessful MM:
  Same volume and spread...
  But inventory P&L: -$50,000 (got caught in a move)

  Net profit: -$27,000/day

The difference is risk management.

---

How MMs Interact with the Exchange

Order Flow

1. MM connects via WebSocket (low latency)
2. Subscribes to order book updates
3. Calculates fair price using:
   - Other exchange prices
   - Recent trades
   - Order book imbalance
4. Posts bid/ask orders
5. Monitors fills
6. Adjusts quotes based on:
   - Inventory changes
   - Price movements
   - Volatility changes
7. Repeat continuously (milliseconds)

Key API Features MMs Need

- Fast order placement/cancellation
- Bulk order operations (post many orders at once)
- Post-only order type (guarantee maker)
- Self-trade prevention (don't match own orders)
- WebSocket for real-time updates
- Low-latency connectivity
- Kill switch / cancel-all endpoint

Colocation

Professional MMs often colocate (place servers physically close to exchange):

Retail trader:
  Home → ISP → Internet → Exchange
  Latency: 50-200ms

Colocated MM:
  Server → Same datacenter → Exchange
  Latency: <1ms

Why it matters:
  - First to react to price changes
  - First to cancel stale quotes
  - Less adverse selection

---

Common MM Terminology

Term	Meaning
Spread	Difference between best bid and ask
Mid	(Best bid + Best ask) / 2
Edge	Expected profit per trade
Inventory	Net position from MM activity
Skew	Adjusting quotes based on inventory
Picked off	Filled by informed trader at bad price
Stale quote	Order at outdated price
Toxicity	How informed/adverse the order flow is
Fill rate	What % of posted orders get filled
Queue position	Where your order sits at a price level

---

Summary

Market making is a service business:

1. Service provided: Immediate liquidity at tight spreads
2. Compensation: Spread capture + maker rebates
3. Risk taken: Inventory/directional risk + adverse selection
4. Skill required: Managing inventory, reacting fast, surviving adverse selection

It's not free money. MMs provide real value (liquidity) and take real risk (inventory). The successful ones manage risk better than the spread they capture.

MM Equation:
  Profit = Spread Revenue + Rebates - Inventory Losses - Adverse Selection - Costs

  If Spread + Rebates > Losses + Costs → Profitable
  If Losses > Spread + Rebates → Bankrupt

The exchange wants MMs because tight spreads attract traders. MMs want the exchange because volume means more spread capture. It's symbiotic - when it works.