Direct Answer
Direct answer: Stock order routing is the process by which your broker decides where to send your order for execution, to a stock exchange, an alternative trading system, or a market maker that internalizes the trade. The destination affects your fill price, execution speed, and the likelihood of getting price improvement. Retail orders sent via payment for order flow (PFOF) are typically internalized by wholesale market makers, while direct-access brokers route to lit exchanges or use smart order routers that seek the best available price across multiple venues. No single routing path is universally better; the quality of your fill depends on order type, stock liquidity, and market conditions at the moment of execution.
What this changes for a real user
Most retail traders never think about order routing. They click "buy," see a confirmation, and move on. But routing decisions happen between that click and the confirmation, and they can mean the difference between filling at the quoted price, getting a slightly better price through price improvement, or paying a wider effective spread than the displayed quote suggests.
- Price improvement (or the lack of it): Wholesale market makers who receive PFOF orders sometimes fill them at prices better than the national best bid or offer (NBBO). How often and by how much varies by broker, market maker, and stock. SEC Rule 605 reports let you compare execution quality across brokers.
- Speed and certainty: Internalizing market makers can fill orders in microseconds with near-certainty. Exchange routes can be faster for some stocks but slower during volatility spikes when matching engines queue orders.
- Effective spread: The bid-ask spread you see is the quoted spread. The spread you actually pay, the effective spread, can be narrower (price improvement) or wider (adverse fills). For active traders, the difference compounds across many trades.
- Transparency: You can request your broker's routing disclosures under SEC Rule 606, which shows where your orders go by order type and security.
For a buy-and-hold investor placing one or two trades per month, routing is a minor concern. For a trader placing dozens of orders per day in volatile stocks, routing quality can meaningfully affect net returns.
Mechanics and definitions
The National Market System (NMS) and NBBO
U.S. equity markets operate under Regulation NMS (National Market System), enacted by the SEC in 2005. Reg NMS established rules designed to promote fair access to quotes and best execution across fragmented venues. The cornerstone concept is the National Best Bid and Offer (NBBO): the highest bid and lowest offer across all registered exchanges at any moment. Brokers subject to best-execution obligations must consider the NBBO when routing orders, though the specific requirements are more nuanced than simply always executing at the NBBO price.
Where orders can go
Once you submit an order, your broker's routing logic, a smart order router (SOR) or a fixed routing arrangement, evaluates potential destinations:
| Venue type | How it works | Typical use |
|---|---|---|
| National exchanges (NYSE, Nasdaq, CBOE) | Orders enter a central limit order book (CLOB). Matches occur between resting and incoming orders at transparent prices. All trades are publicly reported. | Large-cap, high-volume stocks. Institutions and direct-access retail traders. |
| Wholesale market makers (internalization) | A market maker receives the retail order, fills it from its own inventory at or better than NBBO, and reports the trade. No exchange matching engine is involved. | Most retail market orders routed via PFOF brokers (e.g., Robinhood, E*TRADE, TD Ameritrade). |
| Alternative Trading Systems (ATS / dark pools) | Private matching platforms that don't display quotes publicly. Matches large orders against other participants without moving the displayed market. | Institutional block trades; some retail orders at brokers with ATS relationships. |
| Electronic Communication Networks (ECNs) | Automated systems that match buy and sell orders directly, often posting and filling limit orders without a traditional market maker. | Direct-access retail traders; active traders who pay ECN fees/rebates. |
Payment for order flow (PFOF)
PFOF is an arrangement in which a brokerage sells its customers' order flow to a wholesale market maker in exchange for compensation, typically fractions of a cent per share. The market maker profits by capturing the spread between what retail buyers pay and sellers receive, while the broker uses the revenue to offer commission-free trading. The retail customer theoretically benefits from price improvement (fills better than the NBBO quote), but critics argue that PFOF systematically diverts order flow away from lit exchanges where better prices might be available. The SEC has debated PFOF reform for years; as of 2026, the practice remains legal in the U.S. but is banned in the U.K. and Canada.
Smart order routing (SOR)
At brokers that don't rely primarily on PFOF, including most direct-access and professional trading platforms, a smart order router evaluates available venues in real time and splits or sequences orders to minimize cost and maximize fill probability. A SOR typically considers quoted price, available size, venue fees or rebates, historical fill rates, and latency. For liquid large-cap stocks, the difference between SOR venues is often sub-penny and sub-millisecond. For thinly traded securities, the routing decision matters far more because displayed quotes may not reflect true available liquidity.
Make/take fee structures
Most U.S. exchanges use a maker-rebate / taker-fee model. A maker places a limit order that rests in the book, providing liquidity; the exchange pays a rebate (typically $0.002-$0.003 per share). A taker places an order that immediately executes against a resting order, consuming liquidity; the exchange charges a fee (typically $0.003 per share). Some brokers route limit orders to venues offering higher maker rebates, which can benefit or harm the customer depending on whether the routing delay results in a worse price. This is known as rebate arbitrage and is a subtle but real distortion in routing incentives.
Worked example: two brokers, one order
Assumptions: Stock XYZ is trading with a displayed NBBO of $50.00 bid / $50.02 ask. You place a market order to buy 100 shares. The order is submitted at 10:15 a.m. ET during a normal session. This is a hypothetical example; actual routing outcomes vary.
Scenario A: PFOF broker (internalized)
- Your broker sends the order to a wholesale market maker (e.g., Citadel Securities or Virtu Financial).
- The market maker fills you at $50.015, a half-cent better than the $50.02 ask.
- Effective spread paid: $0.015 above midpoint ($50.01), or 3 basis points on the $50 stock.
- The market maker earns the other half-cent of spread it didn't give back as price improvement, plus any PFOF payment from your broker.
- Fill time: ~100-300 milliseconds. High certainty of fill at the stated price.
Scenario B: Direct-access broker (exchange-routed via SOR)
- Your broker's SOR evaluates Nasdaq, NYSE Arca, and CBOE BZX in real time.
- It routes to Nasdaq where 300 shares are offered at $50.02; your 100-share order fills at $50.02.
- Effective spread paid: $0.01 above midpoint, or 2 basis points, better than Scenario A in this example.
- Alternatively, if the quote moves away while the SOR is routing (latency), you might fill at $50.03.
- Fill time: 1-10 milliseconds to exchange. Fill certainty depends on whether the resting quotes are still valid.
What the example shows: The PFOF path provided guaranteed price improvement over the ask but may have left money on the table compared to the exchange midpoint or a better limit order. The exchange path was potentially cheaper but carried quote-fade risk. Neither outcome is universally better. The right evaluation metric for your situation depends on how often you trade, how liquid your stocks are, and whether you use market or limit orders.
Important note: Price improvement statistics published by brokers and market makers are averages across millions of orders. A single order's outcome tells you nothing about the quality of the routing arrangement. Evaluate routing quality using Rule 605 reports over statistically meaningful sample sizes.
Failure modes and what can go wrong
Quote fade (latency arbitrage)
By the time your market order reaches an exchange, high-frequency traders (HFTs) may have already pulled or adjusted the quotes that were displayed when you submitted. This is most common in volatile, fast-moving stocks. You see $50.02 on the ask when you click; the order fills at $50.05. The displayed quote was real, but it was no longer available by the time your order arrived. PFOF internalization partly avoids this, the market maker guarantees your fill price at submission time, but at the cost of potentially giving up mid-spread access.
Internalization without adequate price improvement
Not all PFOF arrangements produce meaningful price improvement. If a market maker consistently fills at the NBBO rather than inside the spread, you're getting the same price you'd get on an exchange but your broker is still receiving PFOF. Rule 605 reports show the percentage of orders that received price improvement and the average amount, use these to compare brokers before assuming your fills are competitive.
Fragmented liquidity and odd-lot quotes
Regulation NMS's Order Protection Rule bars trading through a protected quotation, and odd-lot interest is not a protected quotation. Nor is an odd lot always what you would guess: SEC Rule 600(b)(93) now sets each stock's round lot from its own price, so an order can be an odd lot in one name and several round lots in another at the same share count. The displayed NBBO can therefore coexist with a better-priced odd-lot order your routing never reaches. Since the first business day of May 2026 that odd-lot interest has been consolidated and disseminated by the SIPs, which makes it visible, but visibility is not trade-through protection.
Dark pool information leakage
Large institutions routing to dark pools to avoid market impact can still experience information leakage, other participants infer the order's direction from observable clues and adjust prices before the full order is filled. Retail traders rarely use dark pools directly, but understanding this failure mode matters when reading about institutional execution quality benchmarks.
SOR routing errors and venue outages
Smart order routers are software systems and can route to a venue experiencing technical issues. During Nasdaq outages or exchange halts on specific securities, a SOR that fails to reroute quickly can leave orders unexecuted while prices move. Direct-access platforms typically surface these failures more visibly than PFOF brokers, where the market maker absorbs the venue risk.
Conflicts of interest in routing decisions
A broker that receives PFOF has a financial incentive to route orders to the highest-paying market maker rather than the one offering the best execution quality. Although FINRA and the SEC require brokers to maintain policies that prioritize best execution, the definition of "best" allows significant leeway. Rule 606 disclosures require brokers to report PFOF amounts and routing destinations, reading these reports is the primary way a retail trader can identify potential conflicts.
Risk, limitations, and when routing matters most
When routing matters least
For highly liquid large-cap stocks (think S&P 500 names with millions of shares traded daily), the NBBO spread is often just one penny and price improvement, even when it occurs, is measured in fractions of a cent. A buy-and-hold investor buying 50 shares of a $200 stock every few months will pay a routing-quality difference of perhaps a dollar or two over a year. This is not a meaningful driver of long-term investment returns.
When routing matters most
- High-frequency or active trading: The more orders you place, the more routing costs compound. Even a 1-basis-point difference in effective spread, applied across 500 trades per year, becomes a measurable drag.
- Volatile or thinly traded stocks: Spreads widen, quote depth is shallow, and the gap between the best possible fill and the worst realistic fill grows. Routing to a venue with stale quotes or inadequate depth can produce significantly worse fills.
- Large orders relative to displayed size: If you're buying 5,000 shares and the ask shows only 500 shares at the best price, routing matters enormously. A SOR that slices and routes intelligently will produce a far better average fill than a single exchange destination.
- Market orders during volatile sessions: Wide bid-ask spreads, quote instability, and competing order flow make routing decisions disproportionately impactful during earnings releases, macro data prints, or flash crashes.
Limits of public data
Rule 605 and 606 reports provide aggregate statistics, not individual order analysis. They cover a trailing month and may not reflect current market maker behavior after fee changes or technology updates. Third-party analytics providers (e.g., Bloomberg's TCA tools, S3 Partners) offer more granular execution quality analysis, but these are typically institutional-grade products not available to retail traders at no cost.
Connection to Orders, Routing & Fill Quality
Order routing is the mechanism that connects the orders you learn about, market, limit, stop, stop-limit, to the price you actually pay. Understanding routing without understanding order types produces an incomplete picture: a limit order routed to a dark pool where it never fills is technically well-routed but practically useless. Conversely, knowing that you should use a limit order doesn't help if you don't understand how limit orders interact with exchange matching engines and maker/taker fee schedules.
Fill quality, the set of outcomes produced by the combination of your order type and your broker's routing, ties these concepts together. The subcategory covers bid-ask spread mechanics, price improvement measurement, and the regulatory framework (Regulation NMS, Rule 605, Rule 606) that governs how routing decisions are disclosed and evaluated. This article provides the routing-mechanism foundation; the adjacent articles on fill quality measurement and execution cost analysis build on it.
Checklist: evaluating your broker's routing
- Request your broker's Rule 606 report. Most brokers publish these on their websites. Look at where your order types (market vs. limit, by size) are routed and whether PFOF payments are disclosed.
- Check Rule 605 statistics for your broker's clearing firm or the market makers receiving your flow. Look at effective spread, price improvement rate, and fill rate.
- Compare brokers using third-party routing quality analyses. Academic researchers and financial journalists periodically publish routing quality studies comparing retail brokers. These are more informative than broker-published marketing claims.
- Know your order type defaults. Understand whether your broker sends market orders to wholesalers by default and limit orders to exchanges, or vice versa. Many brokers use different routing paths for different order types.
- Consider direct-access alternatives for active trading. If you trade frequently in volatile or thinly traded stocks, a direct-access broker with a configurable SOR may produce meaningfully better fills than a PFOF-dependent platform.
- Use limit orders when execution price matters more than certainty of fill. Limit orders route differently than market orders and provide explicit price protection. A limit order at the midpoint will not fill in a fast market but avoids paying the full spread when conditions allow.
- Review your fill prices against the NBBO at time of submission. Your broker's order confirmation typically shows fill price but not the NBBO at submission time. Some direct-access platforms show this; otherwise, you need Level 2 data or time-and-sales records to reconstruct it.
- Set a dated reminder to re-evaluate. PFOF arrangements, fee structures, and routing technology change. A routing arrangement that produced good results two years ago may have changed.
Acting on a Decision You Do Not Make
The uncomfortable fact in this material is that the routing decision belongs to the broker on the great majority of retail orders. That leaves two levers rather than none. The first is the choice of firm, made once and reviewed occasionally. The second is the choice of order type and size, made on every trade and frequently more consequential.
The misreading to avoid is treating internalisation as inherently worse than an exchange fill. Orders handled by a wholesaler often execute inside the quoted spread, and orders sent to an exchange come with no assurance of anything better. What determines the outcome is the fill price relative to the quote at the time, and that is measurable without knowing where the trade printed.
Routing analysis has a natural ceiling. It explains the venue and the terms of execution. It does not explain the market conditions that set the quote in the first place, and a wide spread caused by thin trading stays wide wherever the order is sent.
Directed orders are offered by some firms for anyone who wants the decision back, usually with a fee attached and no assurance that the result improves.
Frequently asked questions
What is the NBBO and why does it matter for order routing?
The National Best Bid and Offer (NBBO) is the highest published bid price and lowest published ask price across all U.S. stock exchanges at any given moment. It represents the best publicly quoted price available nationwide. Brokers are required under Regulation NMS to consider the NBBO when routing orders, and market makers who internalize retail orders must fill them at or better than the NBBO. The NBBO is the baseline against which execution quality is measured, a fill inside the NBBO spread is price improvement; a fill outside it (worse than the ask for a buy, or worse than the bid for a sell) signals a quality problem.
Is payment for order flow bad for retail traders?
It depends on what you compare it against. PFOF enables commission-free trading, which has saved retail investors billions in explicit commissions over the past decade. Market makers who receive PFOF do provide price improvement on average, fills inside the NBBO spread. However, critics argue that PFOF systematically keeps retail orders away from lit exchanges, narrows the pool of price discovery liquidity, and may produce worse fills than a well-designed exchange-based routing system would. Academic research is mixed: some studies find retail traders do better under PFOF, others find exchange routing superior for active traders in volatile names. The answer also depends heavily on your trading frequency, stock selection, and order types.
How do I read a Rule 606 routing disclosure?
Rule 606 requires brokers to publish quarterly reports showing, by order type and stock tier, the venues or market makers that received their order flow and whether any PFOF was received. Look for: (1) which market makers received your market orders vs. limit orders; (2) whether the broker received PFOF or paid for access; (3) how the routing changed over time. Note that Rule 606 shows destination and compensation, not execution quality, use Rule 605 reports for quality metrics like effective spread and price improvement rate. Not all brokers make 606 reports easily accessible; some require you to contact them directly.
What is internalization and how does it differ from exchange execution?
Internalization occurs when a market maker fills your order directly from its own inventory rather than routing it to an exchange for matching. The market maker acts as the counterparty, quoting you a fill price and immediately confirming the trade. The trade is reported to the consolidated tape but was never exposed to the exchange's order book. On an exchange, your order enters a central limit order book and is matched against a resting order from another participant, the exchange's matching engine handles execution, not a single market maker. Internalization can be faster and offer price improvement, but it means your order never interacts with the full pool of displayed liquidity on exchanges.
Do limit orders get routed differently than market orders?
Yes, typically. Market orders are more commonly sent to wholesale market makers (via PFOF arrangements) because the market maker can immediately internalize them. Limit orders, particularly those placed away from the current market price, are more likely to be sent to exchanges where they rest in the order book waiting for a matching order. Limit orders near the best bid or offer may also be internalized by market makers who want to add them to their quote, but this depends on the broker and market maker relationship. Understanding this distinction matters because the venue where your limit order rests determines whether it earns a maker rebate (exchange) or sits in the market maker's proprietary book (internalized).
What is a smart order router and when does it help?
A smart order router (SOR) is software that evaluates multiple trading venues simultaneously and routes (or splits) your order to seek the best available price, size, and certainty of fill. SORs are more commonly used by direct-access brokers, professional platforms, and institutions. For a retail-sized order in a liquid large-cap stock, a SOR may produce only marginally better outcomes than a fixed routing arrangement because the stock's quotes are tight and deep across all venues. For larger orders, thinly traded stocks, or volatile conditions, a good SOR can make a meaningful difference by avoiding stale quotes, accessing multiple levels of the order book, and adjusting routing as conditions change during execution.
Can I control where my orders are routed?
At most retail PFOF brokers, you cannot specify a destination. The broker controls routing entirely. Some platforms offer a "directed order" option that lets you specify a particular exchange, but this is uncommon among mass-market retail brokers. Direct-access platforms (Interactive Brokers, Lightspeed, TradeStation, Cobra Trading) typically offer more routing control, letting you direct orders to specific exchanges or ECNs, configure your SOR's venue priority, or use limit orders with exchange-specific routing instructions. If routing control matters to your strategy, the platform choice is as important as the routing feature itself.
How do maker-taker fees affect my fills?
Most U.S. exchanges pay a rebate to participants who provide liquidity (makers, those who post limit orders that rest in the book) and charge a fee to participants who take liquidity (takers, those who submit orders that immediately execute). A typical structure is $0.002/share rebate for makers and $0.003/share fee for takers. If your broker routes your limit order to a venue offering a higher maker rebate. The broker may keep part of that rebate rather than passing it through to you. If your broker routes market orders, taker orders, to venues with lower taker fees, that can reduce the cost to the broker or market maker, but the benefit may not flow to you directly. Understanding maker-taker dynamics helps explain why different brokers route the same order to different venues.
What happens to an order when the venue it was sent to cannot fill it?
The order is typically returned to the router and sent onward to another destination, or it rests at the original venue if it was a limit order that can wait. Each hop adds time, and prices can move during it, which is one reason a fill arrives at a level different from the quote seen at submission. The sequence of destinations attempted is generally not visible to the customer on an individual order.
References
Next lesson
Next lesson: Orders, Routing & Fill Quality hub: covers fill quality measurement, effective spread analysis, and how to use execution cost data to evaluate your trading.
Educational disclaimer
For education only; not personalized investment, tax, or legal advice. Trading can result in substantial losses.
Broker rules, exchange mechanics, routing arrangements, PFOF regulations, and execution quality standards can change. Verify current requirements and broker practices before acting. Rule 605 and 606 data reflect historical periods and may not represent current routing behavior.