Price Improvement and Effective Spread Explained

Direct Answer

Price improvement occurs when a market order or marketable limit order fills at a price strictly better than the NBBO at the moment of execution, below the National Best Ask for a buy, above the National Best Bid for a sell. Effective spread is the standard measure of actual transaction cost: it equals twice the absolute difference between your fill price and the NBBO midpoint at execution time. A fill at the midpoint produces zero effective spread; a fill at the ask (for a buy) produces an effective spread equal to the full NBBO spread. Brokers and market centers are required under SEC Rule 605 to publish monthly statistics on both metrics so traders can compare execution quality across venues without relying solely on marketing claims.

Key takeaways

• The NBBO is a floor, not a ceiling: Regulation NMS requires your order fill at or better than the NBBO; a good router should consistently beat it, especially in liquid large-cap stocks.
• Effective spread is the honest cost metric: Unlike the quoted bid-ask spread, the effective spread reflects where you actually filled relative to the true mid-market price.
• Price improvement is not guaranteed in fast markets: In volatile conditions the NBBO can move before your order arrives, turning apparent improvement into a race against quote flickering.
• Rule 605 disclosures are publicly available: Every market center that executes retail orders must publish monthly effective spread and price improvement statistics, the data to evaluate your broker exists.
• Payment for order flow (PFOF) and price improvement coexist: Market makers who pay for order flow are still required to deliver the NBBO or better, but "at the NBBO" is only the legal minimum, whether they pass through meaningful improvement is a competition question.
• Effective spread can exceed the NBBO spread: If your fill is worse than the quoted ask (for a buy), your effective spread exceeds the quoted spread, a sign of execution quality below the regulatory floor, which should not happen at a compliant market center.
• Quoted spread vs. effective spread often diverge: A stock may show a narrow one-cent quoted spread, but if your fills consistently land at the ask rather than inside it, the effective spread you pay is the full penny, not the theoretical midpoint.

What this changes for a real user

Most retail traders focus on commission costs. For stocks with active retail participation, effective spread often exceeds commissions as the dominant execution cost, yet it is nearly invisible at the point of trade. Here is how price improvement and effective spread show up in practice:

• Every market buy or sell pays at least half the spread implicitly: Even a commission-free broker cannot eliminate the cost of crossing the bid-ask spread. If you buy at the ask and immediately sell at the bid, you lose the spread. Effective spread measures exactly this round-trip cost as it actually occurred, not as it appeared on the quote screen.
• Active traders pay the effective spread repeatedly: A scalper executing 50 round trips per day on a stock with a $0.01 NBBO spread and zero price improvement pays $0.50 per 100 shares in spread cost each round trip, $25.00 total per day per 100 shares, purely from crossing the market. If the broker's routing delivers consistent price improvement of $0.002 per share. That is $0.20 saved per round trip and $10.00 per day, material over time.
• The difference shows up in Rule 605 data: An active trader who compares brokers using publicly available Rule 605 statistics can quantify whether internalization through a market maker or direct exchange routing produces better effective spreads for their typical order size and stock universe.
• Price improvement rates vary sharply by stock: In large-cap, high-volume stocks (S&P 500 names, mega-cap tech), competition among market makers is intense and price improvement rates tend to be high, sub-penny fills inside the one-cent minimum tick are common. In small-cap or thinly traded stocks, price improvement may be near-zero because the quoted spread is already wide and the market maker keeps most of it.
• Limit orders that rest at the midpoint create their own improvement: A limit buy order placed at the NBBO midpoint, rather than at the bid, may fill if the market maker crosses the spread to execute against it. In that case, you capture the side of the spread that would otherwise go to the market maker. This is sometimes called "posting at the mid" and is a technique experienced traders use to reduce effective spread on passive orders.

Mechanics and definitions

Quoted spread

The quoted spread is the simplest measure of transaction cost. It is the difference between the best ask (National Best Offer) and the best bid (National Best Bid) at a given moment:

Quoted Spread = Best Ask − Best Bid

The NBBO midpoint is exactly halfway between:

Midpoint = (Best Ask + Best Bid) / 2

For example: Best Bid = $50.00, Best Ask = $50.02. Quoted Spread = $0.02. Midpoint = $50.01.

The quoted spread represents the cost a trader who immediately buys at the ask and sells at the bid would incur in a single round trip. It is a theoretical maximum round-trip cost when no price improvement exists.

Effective spread

The effective spread measures the actual round-trip cost as experienced by the trader, relative to the midpoint at the moment of execution:

Effective Spread (buy) = 2 × (Fill Price − Midpoint)

Effective Spread (sell) = 2 × (Midpoint − Fill Price)

Or equivalently:

Effective Spread = 2 × |Fill Price − Midpoint|

The factor of 2 converts the one-way cost (half-spread) into a round-trip equivalent, making it directly comparable to the quoted spread. Key reference points:

• If the fill is exactly at the midpoint, effective spread = 0 (maximum price improvement).
• If a buy fills at the ask, effective spread = quoted spread (zero price improvement).
• If a buy fills between the midpoint and the ask, effective spread is between 0 and the quoted spread (partial price improvement).
• If a buy fills above the ask, effective spread exceeds the quoted spread, this should not happen at a compliant venue executing at the NBBO or better.

Price improvement, definition and measurement

Price improvement for a buy order occurs when the fill price is strictly less than the National Best Ask at execution time. For a sell order, it occurs when the fill price is strictly greater than the National Best Bid. The amount of price improvement is:

Price Improvement (buy) = Best Ask − Fill Price

Price Improvement (sell) = Fill Price − Best Bid

A fill exactly at the NBBO provides zero price improvement but is still Rule 611-compliant. A fill inside the NBBO spread, between the bid and the ask, provides positive price improvement and reduces the effective spread below the quoted spread.

Realized spread

Researchers and market microstructure analysts also use a related metric called the realized spread, which measures how much of the quoted spread a market maker actually retains as profit after accounting for subsequent price movement:

Realized Spread (buy) = 2 × (Fill Price − Midpoint five minutes later)

The realized spread is lower than the effective spread because it accounts for adverse selection, the information cost embedded in each trade. A market maker who fills a buy order and then watches the price rise has effectively "lost" part of the spread to an informed trader. The difference between the effective spread and the realized spread is called the adverse selection component or price impact. Understanding this decomposition helps explain why market makers in illiquid or volatile stocks quote wider spreads: their realized spread must cover higher adverse selection costs.

Rule 605 reporting

SEC Rule 605 (formerly known as Rule 11Ac1-5) requires all market centers, exchanges, alternative trading systems (ATSs), and broker-dealer internalizers, to publish monthly execution quality statistics for covered orders (market orders and marketable limit orders of 100 to 9,999 shares). Key Rule 605 statistics include:

• Effective spread: Average effective spread in dollars and as a percentage of the quoted spread, broken down by order size bucket and security type.
• Price improvement rate: Percentage of orders receiving price improvement, and average dollar improvement per share.
• Fill rate and execution speed: Percentage of eligible orders filled and average time from order submission to execution.
• Orders receiving the quote: Percentage filled at the NBBO (zero price improvement).
• Orders receiving outside the quote: Percentage filled worse than the NBBO (should be near-zero under normal conditions).

Rule 605 data is published monthly, usually with a one-month lag, and is available on the websites of each market center as well as through aggregators. Comparing Rule 605 statistics across brokers or market centers for the same security and order-size bucket provides objective evidence of fill quality differences.

Worked example

Assumptions: a liquid large-cap stock; NBBO at time of order arrival: Best Bid = $100.00, Best Ask = $100.02; NBBO midpoint = $100.01; a retail market buy order for 200 shares is submitted at 10:30:00 ET; the order is internalized by the broker's affiliated market maker.

What this means in dollars for the 200-share order:

• No improvement: round-trip cost = 200 × $0.02 = $4.00 (twice the effective half-spread for one leg).
• Sub-penny improvement: round-trip cost = 200 × $0.01 = $2.00, the market maker captured $1.00 of the spread and passed $1.00 to you as price improvement.
• Midpoint fill: zero round-trip cost from spread crossing, you paid no spread at all and the market maker captured the full spread from the other side of the trade.

Why markets rarely fill at the midpoint: A midpoint fill only happens when the market maker (or the limit order resting at the mid) decides it is worth crossing the spread to fill your order. In practice, midpoint fills are uncommon for aggressive market orders in fast-moving markets, they are more common for passive limit orders placed at the mid in quieter conditions, or for internalized retail orders in very liquid stocks where competition among market makers is intense.

The PFOF trade-off illustrated: In Scenario 2 (sub-penny improvement), the market maker bought your order flow via PFOF, gave you $0.005/share improvement, and retained $0.005/share of the spread. Absent PFOF, if your broker had routed directly to the exchange where the $100.02 ask was posted, you might have filled at exactly $100.02, no improvement, but without the market maker's intervention. In a competitive market, the sub-penny improvement you received via internalization is better than the exchange fill. In a less competitive market, it may not be.

Failure modes and what can go wrong

• Quote flickering erases apparent improvement: The NBBO changes in microseconds. A market maker may quote an improvement at the moment your order arrives at the broker, but by the time the order reaches the execution venue the quote has moved. You receive a fill at the new NBBO, which may be worse than the pre-trade quote you observed, without any Rule 611 violation, because the NBBO at time of execution is what matters.
• Price improvement is reported on executed shares only: Rule 605 statistics cover orders that were filled. Partially filled or cancelled orders may not be captured. A broker with high cancellation rates on marketable orders may show good effective spreads for the shares that did fill while obscuring the quality of the complete order interaction.
• Effective spread statistics aggregate across order sizes: Rule 605 breaks down statistics by order-size buckets (100-499 shares, 500-1,999 shares, etc.), but averages within a bucket can obscure wide variation. A single large order filled at an unfavorable price can drag down the average for all orders in its bucket, or conversely, many small orders filled at the midpoint can make the bucket look better than the typical experience of a user with a specific order size.
• Realized spread vs. effective spread confusion: Some brokers or market makers cite realized spread in their marketing because it is smaller than effective spread (it adjusts for subsequent adverse selection). Realized spread is a valid research metric but is not the right measure of what a trader actually paid on entry. That is the effective spread. Mixing the two inflates the apparent quality of execution.
• Improvement in illiquid stocks can be meaningless: If the quoted spread on a thinly traded stock is $0.20, a $0.001 improvement is statistically measurable but practically irrelevant. The trader is still paying an enormous round-trip cost. Percent-of-spread metrics can look similar across liquid and illiquid stocks while masking the absolute dollar cost difference.
• Sub-penny rule and minimum tick constraints: For stocks priced above $1.00, U.S. exchanges must quote in increments of $0.01. Market makers internalizing orders can fill at sub-penny increments (e.g., $100.015), but exchange orders cannot be posted at sub-penny prices. This means sub-penny improvement is only possible via internalization, not via exchange execution, a structural feature of how the U.S. market is organized rather than a sign of superior execution by the market maker.
• Stale Rule 605 data for fast-moving stocks: Rule 605 reports cover the prior calendar month. For stocks that have recently experienced major changes in volatility, spreads, or trading activity, last month's effective spread data may not reflect current conditions. Always treat historical 605 data as directional context, not a precise forecast of your next fill.

Risk, limitations, and when effective spread is not enough

Effective spread and price improvement statistics are useful benchmarks, but they have important boundaries:

• Effective spread only captures one leg of the trade: The standard effective spread formula measures the entry half-spread relative to the midpoint. It does not automatically account for market impact, the fact that your buy order itself may push the ask higher before or during your fill, raising the midpoint against you. For small retail orders in liquid stocks this impact is minimal; for orders larger than the displayed size at the ask, it can be material.
• The metric does not capture fill rate and opportunity cost: A limit order placed at the midpoint may never fill in a fast-moving market. The effective spread on executed trades looks excellent (zero) but the trader missed the move entirely. Evaluating execution quality requires balancing fill rate against effective spread, a tension that passive orders magnify.
• Effective spread does not include commissions or exchange fees: The all-in transaction cost includes commissions (even if zero for most retail platforms), exchange fees or rebates (which the broker may keep), and the effective spread. A broker routing to a maker-taker venue may earn a rebate on a limit order fill that the trader never sees; a broker routing to an inverted exchange pays a fee. These fee structures affect routing behavior and do not appear in effective spread statistics.
• Price improvement statistics are averages, not guarantees: Historical average price improvement per share tells you nothing about the distribution, some orders may receive $0.05 improvement, others $0.00. For strategies where consistent near-zero effective spread is critical (scalping, market-neutral pairs), average statistics may mask unacceptable variance in individual fills.
• Different regimes require different benchmarks: In very illiquid stocks, the effective spread relative to the quoted spread may look similar to a liquid stock, but the liquidity needed to exit at a reasonable price may not exist. VWAP and implementation shortfall benchmarks are more relevant than effective spread for these situations.
• Effective spread does not capture delayed execution risk: A market maker who internalizes your order can delay execution for a brief period (within regulatory limits) to observe how the market moves. If they see an unfavorable move, they fill you quickly; if favorable, they may wait slightly longer. This "internalization option" is captured in realized spread studies but not in the simple effective spread formula.

Connection to orders, routing, and fill quality

Price improvement and effective spread sit at the intersection of every topic in the Orders, Routing & Fill Quality cluster. They are the output metrics that reveal whether the routing, order type, and execution venue decisions made by your broker are actually working in your favor.

NBBO and the Order Protection Rule: The NBBO is the benchmark that effective spread is measured against. The Order Protection Rule (Reg NMS Rule 611) sets the legal floor, fills must be at or better than the NBBO, but effective spread tells you how far above or below that floor your actual fills land. Reading the NBBO and Order Protection Rule article before this one provides the regulatory context for why the NBBO midpoint is used as the reference point.

Smart order routing: A smart order router's job is to identify the execution venue most likely to deliver price improvement or the fastest fill at the NBBO. Whether it routes to an exchange, an ATS, or an internalizer determines both the fill price and whether sub-penny improvement is possible. Effective spread is the primary metric for evaluating whether a smart router's routing decisions are actually improving outcomes.

How order routing works: The routing chain from broker to market center involves decisions that directly determine whether a retail order competes in an exchange auction (where the limit order book may offer midpoint fills from resting orders) or is internalized by a market maker (where sub-penny improvement is discretionary). The How Stock Order Routing Works article explains this chain in detail.

Market orders vs. limit orders: A market order is the natural subject of price improvement and effective spread measurement, it trades immediately against whatever liquidity exists. A passive limit order's effective spread can be zero (midpoint fill) but only if it fills at all. The Market vs. Limit Orders: The Execution Tradeoff article examines how order type choice affects both fill certainty and effective spread.

Checklist: evaluating your broker's fill quality using effective spread data

1. Find your broker's Rule 605 report. Most brokers link to execution quality statistics in their regulatory disclosures or on a dedicated "execution quality" page. If your broker only routes to one or two market centers, those centers' 605 reports contain the relevant data for your orders.
2. Locate the effective spread column. Rule 605 reports show effective spread as a dollar amount and as a ratio to the quoted spread. Look for the order size bucket that matches your typical trade size and the security category (large-cap, small-cap, ETF) that matches your trading universe.
3. Calculate effective spread as a percentage of the quoted spread. An effective spread ratio above 100% is a red flag. It means fills are, on average, worse than the quoted bid-ask. A ratio of 50% means your average fill is halfway between the NBBO and the midpoint. A ratio near 0% means you are consistently receiving near-midpoint fills.
4. Check the price improvement rate. Rule 605 shows the percentage of orders receiving at least some price improvement. In liquid large-cap stocks, a retail-oriented broker should report improvement rates above 80-90% for small market orders. Rates below 50% for liquid stocks warrant scrutiny.
5. Compare across brokers using the same stock universe and order size. Effective spread data is most useful as a relative comparison. Choose a security that trades with similar characteristics across both brokers (a major ETF or large-cap stock) and compare effective spreads for the same order size bucket in the same month.
6. Look at Rule 606 reports for routing transparency. Rule 606 shows which market makers or exchanges receive your order flow and what payment for order flow your broker receives. Combining Rule 606 routing data with Rule 605 effective spread data gives you the most complete picture of how your broker's routing incentives affect your fills.
7. Evaluate separately for different order types. Rule 605 data covers marketable orders (market orders and marketable limit orders). Your passive limit orders that rest on exchange books are subject to different dynamics, their fill quality depends more on queue position, order type, and tick size than on internalization. Do not use effective spread statistics for marketable orders to draw conclusions about your passive limit order experience.
8. Set a review trigger for any change in broker routing policy. Brokers periodically change their market maker relationships, routing algorithms, or PFOF arrangements. When those changes occur, re-examine Rule 605 and 606 data for the subsequent month to verify that effective spreads have not deteriorated.

Which of These Two Numbers Answers Your Question

These measures get quoted together and they answer different questions. One counts how often fills landed better than the quote. The other measures how far the fill sat from the midpoint, which is much closer to what the trade actually cost. If the question is how a broker performs, the cost measure is the more informative of the two, because a high frequency of tiny improvements can sit alongside a poor average cost.

The presentational trap is a statistic expressed as a rate. A percentage of orders improved says nothing about the amount involved, and an amount averaged across orders says nothing about the distribution behind it. Both deserve to be seen before either is trusted.

There is a selection effect built into any aggregate too. Reported figures cover defined order categories and size ranges, so an account whose activity sits outside those ranges is not described by them. Comparing personal experience against a published average compares two different populations.

Neither measure evaluates the decision to trade. They describe the cost of carrying out an instruction, and that cost can look excellent on a trade that should never have been placed.

Frequently Asked Questions

References

• SEC: Disclosure of Order Execution and Routing Practices (Release No. 34-43590, 2000)
• SEC: FAQs: Rule 605 Execution Quality Statistics
• SEC: Responses to Frequently Asked Questions Concerning Rule 606 of Regulation NMS
• FINRA Rule 5310: Best Execution and Interpositioning
• Investor.gov: Executing an Order
• SEC: Responses to Frequently Asked Questions Concerning Rule 605 of Regulation NMS

Next lesson

Next in this cluster: Orders, Routing & Fill Quality hub: explore how smart order routing decisions, alternative trading systems, and the structure of the NBBO all interact to shape the effective spread you ultimately pay.

Educational disclaimer

For education only; not personalized investment, tax, or legal advice. Trading can result in substantial losses.

Regulatory rules, broker practices, exchange mechanics, and market structure requirements change. Verify current requirements with your broker, the SEC, FINRA, or a qualified professional before acting on any information here.