Skip to main content

How Market Makers Set Spreads

The bid-ask spread—the difference between what a market maker will pay to buy and what they'll charge to sell—is not arbitrary. It's not pulled from thin air. Instead, it emerges from a rigorous, often automated economic calculation that weighs multiple competing forces: the risk of holding inventory, the probability that the next trader knows something the market maker doesn't, the cost of operating, and the competitive pressure from other market makers. Understanding how spreads are set reveals the hidden architecture of market liquidity and explains why spreads widen when volatility spikes and tighten when markets calm.

Quick definition: A market maker's spread is the width between their bid (buy) and ask (sell) prices, calculated to compensate them for inventory risk, adverse selection risk, and operational costs while remaining narrow enough to attract volume and remain competitive.

Key Takeaways​

  • Spreads are set dynamically based on real-time inventory position, volatility, and bid-ask imbalance
  • Inventory risk compensation increases when a market maker holds large long or short positions
  • Adverse selection costs increase when the market maker faces informed traders or large orders
  • Spreads scale with volatility—high volatility justifies wider spreads due to greater price movement risk
  • Competition from other market makers and electronic systems puts downward pressure on spreads
  • Regulatory quoting obligations often force spreads tighter than pure inventory economics would dictate
  • Modern algorithmic systems set spreads continuously, updating dozens of times per second

The Economics of Spread-Setting​

At the most basic level, a market maker's spread must cover three categories of costs and risks. First, operational costs: processing trades, managing systems, employing staff, and maintaining technology. These are fixed and variable costs that occur regardless of volatility. A market maker needs enough spread revenue to cover these costs.

Second, inventory risk: when a market maker buys 10,000 shares, they're exposed to the risk that the stock price drops before they can resell those shares. The wider the spread, the more profit they capture per trade, which helps offset this risk.

Third, adverse selection risk: the risk that the next trader has better information than the market maker. If a trader suddenly shows up wanting to buy 100,000 shares, it might be because they know good news is coming. If a market maker sells them shares at the current ask price, they may be leaving money on the table. Spreads must be wide enough to compensate for the statistical probability of dealing with informed traders.

The simplest model for spread-setting, developed by market microstructure theorists like Bid-Ask spread researcher William Silber and others, expresses this in mathematical form:

Spread = 2 × (Inventory Risk Cost + Adverse Selection Cost) + Operational Cost

The factor of 2 appears because the spread applies to both the buy and sell side. In practice, spreads are set asymmetrically—a market maker might have a tighter bid than ask if they're holding excess inventory and want to sell.

Inventory Risk and Position Management​

Inventory risk is perhaps the most direct driver of spreads. When a market maker accumulates a large long position (too many shares), they face downside risk. To incentivize someone to buy from them (absorb their inventory), they lower their ask price. But they simultaneously raise their bid price (to discourage selling to them), protecting their inventory.

Conversely, when a market maker is short too many shares, they raise their ask price (discouraging buying) and lower their bid (encouraging selling).

The magnitude of this adjustment depends on several factors:

Position size: The further the market maker is from their target inventory level, the larger the adjustment. If their target is zero, and they hold 50,000 shares of a 1-million-share daily-volume stock, they're heavily imbalanced and will adjust spreads aggressively.

Volatility: High volatility makes holding inventory more dangerous. A 1% intraday move in a volatile stock translates to a larger loss than a 1% move in a stable stock. So inventory-induced spread adjustments are larger during volatile periods.

Time to rebalance: If the market maker can liquidate their position quickly in a liquid market, they can tolerate a larger position imbalance. If it will take hours to unwind, they need wider spreads.

Correlation with broader market: If the excess inventory is in a stock that's highly correlated with the broader market (like a large-cap index constituent), the market maker can hedge using index futures or ETFs, reducing effective inventory risk. Conversely, if it's an uncorrelated, idiosyncratic stock, hedging is costly, so spreads widen.

Modern market makers use sophisticated algorithms to calculate target inventory levels continuously. These algorithms factor in historical volatility, recent order flow, and broader market conditions. When actual inventory drifts from target, the algorithm automatically adjusts the spread in real time.

Adverse Selection Risk and Information Asymmetry​

Adverse selection is trickier to model but often more important than inventory risk. The question is: what's the probability that the next order I receive is from someone with better information than me?

Large orders are a red flag for adverse selection. If a trader suddenly appears wanting to buy 50,000 shares of a stock with normal daily volume of 100,000, the market maker's alarm bells ring. This might be someone who knows the company is being acquired. If the market maker sells them shares at the current ask price, they may regret it when the acquisition news hits.

To protect against this risk, market makers widen their spreads in the presence of large orders. An order for 50,000 shares might face a spread twice as wide as the standing quote for 1,000 shares.

The adverse selection model can be formalized through the adverse selection cost formula:

Adverse Selection Cost = (Probability of Information Leakage) × (Expected Impact on Price)

This probability depends on several factors:

Order size relative to typical: Large orders relative to normal trading volume suggest informed trading. Tiny orders suggest noise traders.

Timing and pattern: A sudden order after the stock has spiked suggests informed trading. Orders that arrive in a typical pattern suggest routine trading.

Bid-ask imbalance: If far more buyers than sellers are showing up, that could suggest a positive catalyst is known to insiders. Market makers widen spreads when the imbalance is extreme.

Recent volatility: High recent volatility suggests event-driven uncertainty, increasing the likelihood of informed traders.

Stock-specific factors: Stocks with upcoming earnings announcements, regulatory decisions, or other binary events face much wider spreads as adverse selection risk peaks.

The SEC and academic researchers have documented that adverse selection risk explains a substantial portion of spreads in equity markets. During earnings season, spreads widen dramatically. On the day before a major economic announcement that could move the entire market, spreads across equities tend to widen.

Volatility and Spread Widening​

Volatility is one of the most visible drivers of spread widening. When realized volatility (actual price movement) or implied volatility (market expectation of future movement) increases, spreads expand.

The logic is straightforward: if a stock's price is moving 2% per day, a market maker's downside risk from holding inventory is proportionally higher than if the stock moves 0.5% per day. To compensate, they widen their spread.

Volatility's impact on spreads is well-documented. Research by Hasbrouck and Seppi (1992) and numerous subsequent studies shows a strong positive correlation between volatility and spreads. During the COVID-19 crash in March 2020, for example, spreads in major equity indices widened from typical levels of 0.01% of stock price to as much as 0.2-0.5% in extreme cases—a 20-50x expansion.

Market makers adjust spreads in real-time based on volatility indicators. Most major market makers monitor:

  • Historical volatility: The realized standard deviation of recent returns
  • Implied volatility: Market expectations of future volatility, extracted from options prices (the VIX for broad market volatility)
  • Intraday volatility: Real-time measures of how much the stock has moved today
  • Predictive volatility models: Forecasts of future volatility based on recent patterns and broader market conditions

When volatility spikes, spreads widen almost immediately, sometimes within seconds of the volatility shock. This is why you'll often see large spreads in the minutes following major economic announcements or earnings surprises.

Competition and Spread Compression​

If only inventory risk and adverse selection existed, market-maker spreads might be much wider than we observe. The reason they're narrow is competition. When multiple market makers quote the same stock, they compete on spreads. The most aggressive market maker—the one posting the tightest spread—captures the most volume.

This creates competitive pressure to narrow spreads, constrained only by the need to remain profitable. Market makers engage in a continuous quasi-auction to be the "best bid" or "best ask"—the tightest quote that other traders will hit.

This competition has intensified dramatically over the past two decades. The rise of electronic communication networks (ECNs), high-frequency trading (HFT), and the shift toward decimalization (moving from 1/16-inch quotes to penny quotes) have all compressed spreads. Retail investors benefit enormously from this compression—the average quoted spread in major equity indices has shrunk from 2-5 cents in the 1990s to 1 cent or less today.

The competitive pressure works like this: Suppose there are three market makers in a stock. They all post a 2-cent spread. If Market Maker A widens to 3 cents, they lose volume because traders will prefer to hit the tighter quotes from B and C. Market Maker A loses trading volume and with it, their chance to profit from the spread. So they're forced to narrow back to 2 cents (or better) to remain competitive. However, if all three narrow to 1 cent, they're still competing hard but all earning narrower profits per trade. If one tries to go to 0.5 cents, they might be competing so tightly that they're losing money. So there's an equilibrium where spreads settle.

Technology has also amplified competition. High-frequency trading firms can process information and adjust quotes microseconds faster than traditional market makers. This speed advantage allows them to quote tighter spreads and still manage their risks. The presence of HFT and algorithmic trading has compressed spreads further, though this is a controversial topic in market microstructure research.

Regulatory Constraints on Spreads​

In practice, regulatory quoting obligations set a maximum spread that market makers are allowed to post. In many cases, the actual market spread is tighter than the regulatory maximum, driven by competition. But during stress, market makers will often widen to the regulatory maximum when conditions are difficult.

For example, if NASDAQ rules allow a maximum spread of 5 cents for a particular stock, but competition normally pushes the spread to 1 cent, a market maker will post 1 cent when trading is normal. But if volatility spikes and inventory imbalances become severe, they may widen to 3 or 4 cents—still within the regulatory limit but much wider than normal.

This creates a two-step adjustment process:

  1. Normal adjustment: Market makers adjust spreads within competitive constraints, often posting tighter spreads than the regulatory maximum.
  2. Stressed adjustment: During extreme conditions, spreads widen toward the regulatory maximum as market makers face severe inventory risk or adverse selection risk.

Regulatory spreads are set with the intention that they be tight enough to remain competitive but loose enough to allow market makers to survive stress. If spreads are too tight, market makers will exit the business, harming liquidity. If spreads are too loose, they fail to protect investor interests.

The Role of Order Flow and Price Pressure​

A subtle but important factor in spread-setting is order flow imbalance. When there's an imbalance between buy and sell orders (buy pressure exceeds sell pressure, for example), market makers interpret this as a signal that the price may move. To protect against this, they adjust their spreads.

If there are far more buyers than sellers in a stock, market makers widen their ask price (the price at which they'll sell) to lock in a profit if the stock indeed rises. They may also tighten their bid to reduce the incentive for more buying. Conversely, when there's heavy selling pressure, they widen the bid and tighten the ask.

This is not price manipulation—it's rational risk management. The order flow imbalance is a signal that something has changed about the stock, and the market maker adjusts their quotes to protect against the risk that they've mispriced the stock.

Asymmetric Spreads: When Bid and Ask Are Not Equal​

In the models above, we've treated spreads as symmetric—the distance from mid-price to bid equals the distance from mid-price to ask. In reality, spreads are often asymmetric. A market maker might quote a bid of $99.50 and an ask of $99.51, which is symmetric, but they might quote a bid of $99.49 and an ask of $99.51, which is asymmetric.

Asymmetry reveals market makers' inventory positions and their views on the stock:

  • Tight bid, wide ask: Market maker holds excess long inventory (too many shares) and wants to sell. They tighten their bid to discourage selling to them and widen their ask to encourage buying from them.
  • Wide bid, tight ask: Market maker holds excess short inventory (short too many shares) and wants to buy. They widen their bid to encourage sellers and tighten their ask to discourage buyers.

This asymmetry is also driven by adverse selection. If a market maker believes an informed trader is more likely to be a buyer (say, based on the pattern of order flow), they'll widen their ask (tighten the spread on the sell side) to discourage buying and tighten their bid to the mid-price or beyond, offering a wider spread on the buy side.

Real-World Examples of Spread Changes​

Apple (AAPL) on a Quiet Day: On a normal trading day with steady volume and low volatility, Apple might trade with a typical spread of 1 penny. Market makers quote AAPL at bid $150.00 / ask $150.01, capturing the penny spread on thousands of trades.

Apple (AAPL) After an Earnings Miss: On the day Apple reports earnings that miss expectations, volatility spikes. The stock gaps down 3% in pre-market. When the market opens, the spread widens to 5 cents or more. Market makers quote AAPL at bid $149.50 / ask $149.55. Why? Because the adverse selection risk is extreme (is more bad news coming?) and inventory imbalances are severe (many more sellers than buyers). The wider spread compensates for this risk.

Highly Liquid Index ETF (SPY) During Crisis: The S&P 500 ETF (SPY) is one of the most liquid securities ever. On a calm day, the spread is 1 penny. But on March 16, 2020 (in the COVID crash), spreads in SPY briefly widened to 10-20 cents as volatility exploded, inventory imbalances reached extreme levels, and market makers reassessed their risk. The spread then tightened again as electronic systems rebalanced and volatility stabilized.

Small-Cap Stock During Thin Trading: A small-cap stock with lower daily volume might normally trade with a 10-cent spread. But when volume dries up (like in the final minutes of the trading day or during a holiday), spreads widen to 25-50 cents, or even dollars. The market maker is compensating for illiquidity and the risk that they won't be able to quickly lay off their position.

Common Mistakes in Understanding Spread-Setting​

Mistake 1: Assuming spreads are random or arbitrary. Spreads are calculated, dynamic, and economically justified. They respond predictably to inventory, volatility, and information asymmetry.

Mistake 2: Believing wider spreads always mean the market maker is greedy. Wider spreads during volatility are compensation for real, quantifiable risk. Market makers are not being generous when they tighten spreads during calm periods—they're responding to lower risk.

Mistake 3: Thinking spreads are the same across all venues. Spreads vary between exchanges, ECNs, and alternative trading systems based on differences in liquidity, regulation, and competition.

Mistake 4: Confusing quoted spreads with effective spreads. The quoted spread is what you see on the screen. The effective spread is what you actually pay when you trade, accounting for order execution at better or worse prices than the quote. These can differ significantly during volatile periods.

Mistake 5: Ignoring the role of technology. Modern spreads are set by algorithms that respond to hundreds of inputs in milliseconds. Human judgment plays a role, but automation is the dominant factor.

FAQ​

Q: If market makers are setting spreads based on inventory risk, why don't they just not hold inventory? A: Because they can't avoid it. Market makers provide continuous liquidity, which means they're always buying from some traders and selling to others. Over any given hour or day, these flows are imbalanced, forcing them to hold inventory. Not holding inventory would mean not providing liquidity, which violates their market-maker obligations and business model.

Q: How quickly do market makers adjust spreads? A: Electronic market makers adjust quotes dozens to hundreds of times per second. A significant change in volatility or a large order might trigger spread adjustments within milliseconds.

Q: Are options spreads set the same way as equity spreads? A: The principles are similar, but options spreads tend to be wider due to the additional complexity of pricing multiple inputs (time decay, volatility, Greek sensitivities). Options market makers also face adverse selection risk from traders who understand options pricing better than the average equity trader.

Q: What happens to spreads in after-hours trading? A: Spreads typically widen dramatically in after-hours markets. Volume is lower, so market makers face more inventory risk. Additionally, after-hours trading has less regulation and lower quoting obligations, so spreads can be much wider. It's common to see spreads of 50 cents to several dollars in after-hours trading, versus 1-2 cents during regular hours.

Q: Do spreads differ between limit orders and market orders? A: Limit orders interact with the posted quotes, so they face the advertised spread. Market orders are effectively buying at the offer or selling at the bid, so they face the quoted spread. However, market makers may execute limit orders at better prices than the quote if they choose, which happens during competitive periods.

Q: How do short-selling restrictions affect spreads? A: Short-selling restrictions make it harder for market makers to establish short positions to hedge their long inventory. This increases inventory risk and often leads to wider spreads, particularly in stocks subject to short-selling halts.

Q: Can spreads be negative? A: Technically no, but in extreme illiquidity (like some penny stocks), the bid can be higher than the ask, which is a quote inversion. This typically occurs when the market breaks down and orderly quoting ceases.

  • Bid-Ask Bounce: The tendency for prices to oscillate between bid and ask, driven by the spread itself
  • Market Impact: The cost incurred by large traders due to spreads and inventory adjustment
  • Information Asymmetry: The gap in information between traders and market makers
  • Volatility Clustering: The empirical phenomenon that high volatility tends to be followed by more high volatility, which impacts spread forecasting
  • Order Book Imbalance: The difference between buy and sell orders, which signals informed trading and affects spreads
  • Effective Spread vs. Quoted Spread: The difference between theoretical and actual transaction costs

External Resources​

Research and guidance on bid-ask spread dynamics:

Summary​

Market makers set their spreads through a dynamic, real-time calculation that balances inventory risk, adverse selection risk, operational costs, and competitive pressure. Inventory risk—the cost of holding undesired positions—drives wider spreads when market makers are imbalanced. Adverse selection risk—the cost of trading with informed traders—widens spreads when order flow suggests information leakage. Volatility directly increases both risks, causing spreads to widen during turbulent markets.

Competition and regulatory requirements push spreads tighter than pure risk calculations would suggest. Electronic trading systems have accelerated the speed at which spreads adjust and compressed spread levels overall. The result is a continuous, automated market-making system that provides remarkably tight spreads during calm periods and adjusts appropriately during stress.

For investors, understanding spread-setting reveals that there's no such thing as a "fair" spread that holds constant. Instead, spreads are the market's natural price for liquidity, adjusted dynamically based on the risk the market maker faces.

Next​

Market-Maker Inventory Risk