Bid-Ask Spread Explained: 5 Key Drivers of Intraday Costs

The Anatomy of Intraday Transaction Costs: Beyond the Broker Fee

You press the bid on AAPL at 9:31 AM. The Level 2 shows clean size at 185.20, but the offer ticks up to 185.23 before your limit fills. You exit at 185.28, net 5 cents — except the round-trip spread pulled 3 cents out of the trade. That is 60% of your gross profit, vanished on a name that just opened at the most volatile moment of the session.

The bid-ask spread is not a fee a broker imposes. It is a dynamic, market-driven cost of execution, and it is the first line item you should measure before sizing any scalp. Five core variables force spreads to widen or contract during the trading day:

1. Adverse selection — protection against informed flow

2. Inventory holding costs — dealer risk of holding unwanted positions

3. Order processing costs — operational expenses and tick size constraints

4. Return volatility — premium for price uncertainty

5. Trading volume and liquidity — competition among market makers

Each driver responds to a separate signal. If you cannot identify which variable is currently expanding the spread on your chart, you cannot stack a valid entry.

The bid-ask spread is your real entry tax. Quoted size on the tape is theoretical liquidity until the fill prints.

Adverse Selection: Protecting Against Informed Market Participants

Market makers widen their quotes when they suspect they are about to trade against someone with superior information. That is adverse selection — the information asymmetry baked into the quote.

When a large institutional desk or a trader acting on unreleased news crosses the spread aggressively, the market maker loses. To compensate, the dealer pads the quote before the print. Empirical estimates place the adverse selection component of the quoted spread between 9.6% and 43%, depending on the market and model. Stoll's 1989 study on NASDAQ put the figure at 43%. Huang and Stoll's 1997 model compressed that range to 9.6%–21.5%. The variance is the point — the component is latent, not fixed, and it shifts with the information environment.

For a scalper, this means the spread widens when the tape is about to move. If you see the Level 2 thinning a second before a headline prints, you are watching the information asymmetry in real time. The market maker is not scared of your 500-share limit order. They are scared of the institutional flow they expect to follow.

The mechanical trigger: heavy size disappears on one side of the book. The bid or offer becomes thin. The opposite side holds firm. That is absorption and information asymmetry working together. If you are not already in a position, do not chase the side that just got absorbed. Wait for the spread to normalize after the move completes. The informed flow has already crossed.

Inventory Management and the U-Shaped Intraday Spread Pattern

Market makers carry inventory on their books. If they accumulate a long position during a buy imbalance, they are forced sellers at some point. To avoid that risk, the dealer widens the spread in the direction that would force them to add to that inventory.

The result is a U-shaped intraday spread pattern. The spread peaks at the open and at the close. It tightens during the midday session.

• Open (9:30–10:00 AM ET): Overnight news accumulates. Dealers return to their books uncertain about fair value. Spreads widen as a defensive measure.
• Mid-morning (10:00 AM–12:00 PM ET): Information is digested. Market makers establish inventory. Spreads compress to the tightest levels of the day.
• Lunch (12:00–2:00 PM ET): Volume drops. Fewer market makers compete. Spreads drift wider despite low volatility.
• Close (3:00–4:00 PM ET): Dealers offload inventory to avoid overnight risk. Spreads widen into the closing auction.

Timing implication: if you are scalping a $0.50–$1.00 target, the spread will eat a larger percentage of your trade at the open and close. Either adjust your position size down or skip the setup entirely during those windows. Liquidity is not equal across the session. A $0.01 spread at 11:00 AM on NVDA is not the same execution cost as a $0.03 spread at 9:45 AM on NVDA, even on identical price action.

Trade the tightest window, not the widest. Mid-morning on high-cap names is where the spread costs you the least as a percentage of the move.

Operational Efficiency: Order Processing and the Impact of Tick Sizes

Order processing costs are the operational expenses of running a market-making book — clearing, exchange fees, technology, and staff. In electronic markets, this component is structurally low. A study on the HUF/EUR interbank FX market attributed 47.09% of the spread to order processing costs. In US equities, the figure is compressed by scale and automation.

The more important constraint is tick size. Under SEC Regulation NMS Rule 612, the historical minimum pricing increment for US stocks priced at $1.00 or more is $0.01. That is a structural floor. If the fundamental cost of market-making is $0.005, the spread cannot compress to $0.005 — it cannot print below $0.01.

In September 2024, the SEC adopted amendments to Reg NMS to address this directly. The new rule introduced a second minimum pricing increment of $0.005 for certain "tick-constrained" NMS stocks — names where the $0.01 floor artificially inflated transaction costs. The SEC also amended the access fee cap to $0.0010 per share for protected quotations in stocks priced at $1.00 or more. The compliance date for the amended access fee caps is November 2, 2026.

For stocks priced below $1.00, the minimum pricing increment under Rule 612 is $0.0001. Tick size math governs your execution before liquidity does.

Practical implications for live execution:

• A tick-constrained name quoted at $0.01 / $0.015 has a structural problem. The midpoint is $0.0125, but no order can rest there. The result is a wider effective spread than the fundamentals require.
• If you trade tick-constrained names, expect the spread to compress asymmetrically post-2026 compliance. That changes the structural cost of your entries and the profitability of sub-penny scalping strategies.
• Low-priced stocks (penny to $0.99) operate in $0.0001 increments. The spread is measured in basis points, not pennies. A $0.0010 spread on a $0.50 stock is 20 bps — roughly equivalent to a $0.20 spread on a $100 stock.

Volatility and Liquidity: How Market Dynamics Force Spread Expansion

Return volatility and the bid-ask spread are positively correlated. When price uncertainty rises, market makers demand a higher premium for taking the other side. A VIX spike, an earnings surprise, or a fast gap down widens the spread mechanically. This is not a market maker being greedy — it is a market maker pricing the probability that the next 30 seconds will invalidate their quote.

Trading volume and market capitalization are negatively correlated with the spread. Higher liquidity increases competition among market makers. Tighter competition compresses the quote. AAPL at 11:00 AM with heavy two-sided flow trades at $0.01. A small-cap biotech at 11:00 AM with no volume might trade at $0.05–$0.15 on the same day.

This is why the same setup on two different stocks produces two different P&L curves. The execution cost is not the broker's commission. It is microstructure — the intersection of volatility, liquidity, and inventory pressure.

Trading rules for managing spread-driven volatility:

• Match position size to liquidity. A 10,000-share scalp on a $500M-cap name will move the spread against you. A 10,000-share scalp on SPY will not. Your order is a signal to the market maker regardless of your strategy.
• Skip the first 5 minutes of a macro release. Volatility is elevated, the spread is wide, and fill probability is degraded. Wait for the spread to normalize before pressing the entry.
• Use the spread itself as a signal. If your name is suddenly trading at $0.02 when it normally trades at $0.01, something has changed — volatility, news, or one-sided inventory pressure. Treat the spread as a real-time indicator of risk, not a static line item on a confirmation screen.

Execution Rules: Managing Intraday Transaction Costs

The bid-ask spread is the most underpriced line item on most active traders' P&L. It is not a fee and it is not negotiable. It is a real-time price for liquidity, and it is governed by five measurable drivers. Trade without reading the spread, and you are absorbing a structural cost you cannot quantify. Trade with the spread, and you have a leading indicator that arrived before the chart did.

1. Map the session. Tightest spreads between 10:00 AM and 12:00 PM ET on US equities. Avoid scalping during the first 15 minutes and the final 30 minutes unless your strategy is specifically engineered for the opening range or closing auction.

2. Read the Level 2, not just the price. A clean book with multiple participants refreshing quotes on both sides is the lowest-cost execution environment. A stacked one-sided book signals that inventory is heavy, and the spread will widen on the next imbalance.

3. Size for the spread. If the spread is $0.03 and your target is $0.10, your risk/reward is structurally 1:3.3 before commission. If your target is $0.05, the spread is 60% of your gross. Adjust size accordingly or pass the setup.

4. Watch the calendar. Earnings, CPI prints, FOMC decisions, and major macro releases mechanically widen the spread. Know the schedule before you press the entry — the volatility premium is priced in before you see the chart.

5. Track tick size. Trade names that are tick-constrained at $0.01 with respect to your execution model. The September 2024 SEC amendments compress spreads on certain tick-constrained stocks by 2026 — build that into your slippage projections.

6. Exit when the spread expands. If your scalp is working and the spread suddenly widens mid-trade, take the profit. The market maker is signaling that the next 30 seconds carry risk. Honor the signal rather than hold for the full target.