What is Best Execution?

Introduction

Best execution is the obligation to handle a client’s order so the result is as favorable as reasonably possible under the market conditions that actually exist at that moment. That sounds straightforward until you look at how modern markets work. The same security may trade on multiple venues, quoted prices may not reflect the full size available, liquidity can vanish while an order is in flight, and the route that looks cheapest at first glance may be slower, less likely to fill, or more conflicted. Best execution exists because simply taking the displayed price is often not enough to protect the client.

The central idea is easy to miss: best execution is not a promise to achieve the single best price that existed somewhere in hindsight. It is a disciplined process for searching, choosing, and monitoring execution methods so that a customer gets the best reasonably available outcome ex ante; before the result is known. Regulators in the U.S. frame this as a duty of reasonable diligence to ascertain the best market and obtain the most favorable terms under prevailing conditions. In the UK and under MiFID-style language, the standard is often expressed as taking all sufficient steps to obtain the best possible result. The wording differs, but the mechanism is similar: firms must make serious, evidence-based choices rather than treat routing as an afterthought.

This matters because execution is where market structure becomes real money. A few basis points lost to avoidable slippage, poor routing, hidden fees, or unnecessary delay can swamp management fees, erase alpha, or meaningfully worsen retail outcomes. That is why best execution sits at the junction of trading, technology, governance, and conflicts management. It is both a Market Infrastructure problem and a conduct problem.

What problem does best execution solve?

An order is not just an instruction to buy or sell. It is a request to convert investment intent into an actual trade in a market that is fragmented, competitive, and only partly visible. If there were a single venue, with infinite liquidity, zero latency, no fees, and no conflicts, best execution would barely need to exist as a distinct concept. The order would simply trade there.

Real markets are not like that. There are multiple exchanges, alternative trading systems, wholesalers, market makers, broker crossing systems, dark pools, periodic auctions, and dealer markets. In some products, especially fixed income and municipal securities, there may be no central lit order book at all. That means the broker or dealer is not just transmitting an order into a neutral machine. It is making choices about where to seek liquidity, how aggressively to trade, whether to expose the order, when to split it, and sometimes whether to trade against its own inventory.

Once that is true, agency risk appears. The client wants the best attainable outcome. The intermediary may have other incentives: payment for order flow, exchange rebates, internalization economics, lower operational costs, easier settlement, or simple habit. Best execution is the framework that says those incentives cannot be allowed to determine routing unless they genuinely improve the client’s result. Disclosure helps, but disclosure alone is not enough. FINRA has been explicit that a firm is not relieved of best-execution obligations merely because it disclosed routing practices or payment for order flow.

So the problem is not merely technical fragmentation. It is the combination of fragmentation and delegated decision-making. Best execution exists to make that delegation trustworthy.

What does “best” mean in best execution?

A common misunderstanding is to treat best execution as synonymous with best price. Price is central, but it is not the whole result. If a venue displays a slightly better price but is slow, inaccessible, or unlikely to fill, routing there may produce a worse outcome than a faster venue with a slightly inferior quote. That is why SEC guidance highlights factors such as the opportunity for price improvement, speed of execution, and likelihood of execution. FINRA Rule 5310 adds a fuller set of considerations: the character of the market, the size and type of transaction, the number of markets checked, accessibility of quotations, and the terms and conditions of the order.

The deeper point is that execution quality is a joint outcome. The client cares about the price actually obtained on the size desired, net of costs, with a high enough probability of completion, within a time profile that fits the order’s purpose. A small market order in a highly liquid stock and a large block in an illiquid bond do not face the same problem, so they cannot be judged by a single simplistic benchmark.

This is why regulators usually avoid hard numeric formulas. “Reasonable diligence” is a facts-and-circumstances standard. It asks whether the firm used a process appropriate to the market and order, not whether it can point to one metric in isolation. MSRB Rule G-18 makes the same point in municipal securities: failure to obtain the single most favorable price does not automatically mean the dealer failed to use reasonable diligence. That is not because price is unimportant. It is because the market may have been opaque, illiquid, or conditional in ways that make the apparent best price not truly available for that client’s order.

The phrase reasonably available does a lot of work here. It excludes fantasy liquidity. A quote seen too late, inaccessible without unacceptable delay, available only for trivial size, or withdrawn before execution may not be meaningfully available. Best execution therefore asks: what was genuinely reachable with an appropriately diligent process?

How does best execution work in practice?

In practice, best execution is a chain of linked mechanisms rather than a single decision. A firm starts with an order and classifies it: what security is involved, how liquid is the market, how urgent is the trade, what are the client’s instructions, and what execution constraints matter most? From there it decides whether to route immediately, seek price improvement, internalize, expose the order to multiple venues, work it over time, or use an algorithm.

For liquid listed equities, this often means some form of smart order routing. The router compares available venues and chooses where to send all or part of the order based on price, size, fees, speed, and fill probability. Nasdaq Nordic’s description of its SOR is a concrete example: routable orders check the local book, use a best-bid-and-offer style view across away venues, can re-price for better execution, may route to multiple venues in parallel, and post any residual volume back to the home order book. The point is not that one vendor’s logic defines best execution. The point is that the duty becomes real through routing logic like this.

For a more institutional order, the mechanism changes. Suppose an asset manager needs to buy a large position in a stock without pushing the market up too much. Hitting every displayed offer immediately might maximize speed, but it would also increase price impact and reveal demand. Working the order passively over time may reduce impact, but it creates completion risk: the market may run away before the order is finished. So the broker might use an algorithm tied to a benchmark such as VWAP or TWAP, or a participation strategy that trades only as a fraction of market volume. Here best execution is about balancing two forces that pull against each other: impact and completion.

A simple narrative makes the tradeoff clearer. Imagine a portfolio manager wants to buy a large-cap stock after new information changes the fund’s view. The trader’s first instinct might be speed, because the thesis is time-sensitive. But if the order is large relative to displayed liquidity, rushing into the market will move the price. A more careful approach slices the order and routes pieces across venues, favoring pockets of liquidity where execution is less visible. If volume rises around the open or close, the algorithm can participate more there because more natural liquidity is available. If the market starts drifting away, the strategy can become more aggressive to reduce the risk of underfilling. The “best” result is not defined by any one child order. It is defined by the total execution path.

In dealer or quote-driven markets, the process looks different again. There may be no consolidated visible venue to sweep. Instead, best execution depends on gathering and comparing quotes, consulting broker’s brokers or platforms, evaluating inventory positions, and documenting why the chosen market was the best reasonably available. That is why both FINRA and the MSRB require written policies for securities with limited quotations or sparse pricing information. When transparency is low, process discipline has to rise.

Why isn’t choosing the cheapest venue sufficient for best execution?

It is tempting to reduce best execution to venue comparison: just check enough markets and choose the best quote. But this misses the main difficulty. The execution result depends not just on where the order is sent, but on how the order interacts with liquidity after it arrives.

A displayed quote may only be available for a small size. A venue may have a fast top-of-book update but poor fill quality deeper in the queue. Another venue may offer midpoint or dark liquidity that improves price but with uncertain timing. Internalization may produce a fast fill and lower spread cost in some cases, but if the firm’s economics favor keeping flow in-house, that same model can create a conflict. Best execution therefore cannot be solved by a one-time routing table. It requires continuous review of actual outcomes.

This is why FINRA requires firms that route or internalize order flow to conduct regular and rigorous reviews of execution quality, at least quarterly, on a security-by-security and order-type basis, and to modify routing arrangements if material differences in execution quality appear across markets. Order-by-order review is increasingly possible in some markets and is required where firms execute internally. The principle is straightforward: if your routing logic affects outcomes, you must test whether it is actually producing competitive outcomes.

That review cannot be purely formal. The FCA’s multi-firm review of UK cash-equities execution found that some banks had reasonable monitoring frameworks but weak second-line challenge or management information. That distinction matters. A policy that sounds sophisticated but is not connected to outcome data is not much protection. Best execution is a control system, not a slogan.

How should firms measure execution quality?

Once a firm accepts that best execution is about outcomes under real constraints, measurement becomes unavoidable. But measurement is harder than it first appears, because different metrics answer different questions.

The most basic comparison is against the prevailing quote or benchmark at the time the order arrived. Did the client receive price improvement? Was the execution inside, at, or outside the spread? How quickly did the order fill, and for how much size? For many retail-sized listed-equity orders, those statistics are meaningful because marketable orders are often small relative to displayed liquidity.

For larger or more complex orders, practitioners often use transaction cost analysis, or TCA. A key benchmark is implementation shortfall: the difference between the portfolio manager’s decision price or arrival price and the actual execution result, including the effect of market drift and the trader’s own impact. This metric is powerful because it captures the whole cost of turning a decision into a filled trade. CME’s TCA primer describes implementation shortfall as a standard industry measure precisely because it includes both trend and impact.

VWAP and TWAP are useful too, but for narrower questions. A VWAP comparison asks whether the execution did better or worse than the market’s volume-weighted average over the interval. That can be a sensible benchmark when the order’s goal is to trade in line with market volume. A TWAP comparison fits situations where time, rather than volume, is the governing schedule. Neither metric automatically equals best execution. If the client needed urgency, “beating VWAP” by trading too slowly may actually be a poor outcome. Benchmarks only make sense relative to the order’s objective.

Other measurements matter as well: fill rate, spread capture, realized slippage, market impact, and post-trade reversion. Reversion is especially revealing for institutional orders. If the price jumps while the order is being worked and then drifts back after completion, that suggests temporary impact from the trade itself. If the price does not revert, the move may reflect new information rather than execution pressure. This distinction helps firms decide whether their strategy was too aggressive or whether the market was simply moving.

The important caution is that metrics do not substitute for judgment. A firm can optimize one benchmark while harming the client on another dimension. The benchmark must match the economic purpose of the order.

How do conflicts of interest affect best execution?

Best execution becomes most important exactly when the intermediary has reasons to prefer one outcome over another. Payment for order flow is the best-known example. If a broker is paid to send orders to a wholesaler, the broker may gain from routing choices that are not perfectly aligned with the client’s interest. The conflict does not automatically mean the outcome is bad; some wholesalers provide fast execution and price improvement. The point is that the conflict changes what must be proven.

FINRA and the SEC have both emphasized that these arrangements cannot interfere with best execution. FINRA’s reminder on payment for order flow states that disclosure does not absolve firms of the duty. Chair Gensler’s 2022 statement supporting a Commission-level best execution rule made the same concern explicit: conflicted transactions, including payment for order flow, would warrant enhanced policies, procedures, and documentation.

That specific SEC rulemaking path has changed. The SEC formally withdrew the earlier proposed Regulation Best Execution in 2025 and said any future action would require a new proposal. But that withdrawal does not erase the existing duty. In the U.S., best execution remains enforced through the longstanding SEC framework and self-regulatory organization rules such as FINRA Rule 5310 and MSRB Rule G-18. So the current practical lesson is simple: the duty is still real even if one proposed Commission-level overlay did not go forward.

Internalization raises similar issues. If a firm fills client orders against its own book, it may reduce costs and deliver competitive outcomes; or it may steer flow inward because it is profitable for the firm. Regulators therefore look closely at whether the firm’s controls benchmark internalized executions against external markets and whether the firm’s oversight focuses on actual client outcomes rather than just procedural compliance.

How do client routing instructions change a broker’s best-execution duty?

Best execution is not unlimited discretion. If a customer explicitly instructs a broker to route to a specific market, the firm’s duty changes. FINRA states that an unsolicited customer instruction to route to a particular market relieves the firm from making a best-execution determination beyond that instruction, though the firm must still handle the order promptly and according to its terms.

This exception reveals something important about the concept. Best execution governs the intermediary’s delegated choices. When the client meaningfully takes over one of those choices, the intermediary’s responsibility narrows to carrying out that instruction properly. The firm still cannot use the instruction as a pretext for delay or mishandling, but it is no longer responsible for comparing all other venues the client chose not to use.

That said, many institutional mandates are more nuanced than a simple directed order. A client may care about minimizing information leakage, staying within a benchmark, or participating passively. In those cases, the broker still has substantial execution discretion, which means the best-execution obligation remains wide.

Why does best execution vary across asset classes and market structures?

The invariant is the same across asset classes: use a diligent process to obtain the most favorable achievable result for the client. What changes is the market structure.

In highly liquid equities, best execution often centers on venue fragmentation, latency, queue position, and small differences in spread or price improvement. In futures, the same principles show up through algorithm design and impact control. In municipal bonds or other less transparent markets, the challenge is often quote discovery and documentation rather than microsecond routing. MSRB guidance is explicit that there is no categorical requirement to use a fixed number of venues or a particular platform in every case. What counts is whether the dealer’s policies are reasonably designed for the actual market and are followed.

This is why best execution is better understood as a governance framework attached to a trading process than as a single trading tactic. Smart order routers, execution algos, venue scorecards, TCA dashboards, quote logs, and supervisory reviews are all tools. None of them alone is “best execution.” They are the machinery used to pursue it.

What are common misconceptions about best execution?

The first mistake is thinking best execution means achieving the absolute best price visible somewhere after the fact. That standard would be impossible and would confuse luck with diligence.

The second mistake is thinking best execution is satisfied by a policy document. It is not. A firm needs evidence that its routing logic, venue choices, and execution methods are regularly tested against actual outcomes.

The third mistake is treating non-price factors as excuses. Speed, certainty, settlement likelihood, and market impact matter because they affect the client’s result, not because they let a firm ignore price. A proper analysis asks how each factor changes the total expected outcome.

The fourth mistake is assuming disclosure cures conflicts. Regulators have repeatedly rejected that view.

Conclusion

Best execution is the discipline of turning a client order into the best reasonably available outcome in a market that is fragmented, uncertain, and full of incentives that can pull the wrong way. Its core insight is simple: execution quality is not just a price on a screen, but the result of a process; routing, timing, sizing, venue choice, and control of conflicts.

If you remember one thing tomorrow, remember this: **best execution is not a guarantee of perfection; it is a test of whether the intermediary used a process genuinely designed to serve the client rather than itself. **

What should an institutional trader evaluate before executing in this market?

Evaluate a market’s liquidity, costs, and execution risk before you trade, then use Cube Exchange to execute the plan with the right order type. On Cube, fund your account, open the market for the instrument, choose the execution approach that matches your liquidity profile, and send the order through the chosen route.

1. Check current liquidity and immediacy: read the top-of-book spread, available size at best bids/offers, and recent trade sizes vs your target block.
2. Confirm venue access and explicit costs: verify which venues or counterparty options are available on Cube and note taker/maker fees or visible rebates that affect net price.
3. Pick an execution method and limits: select market or limit, or choose a time/participation approach (VWAP/TWAP or a participation cap); set a maximum participation rate and a firm limit price or slippage tolerance.
4. Slice, submit, and monitor: if your size exceeds displayed depth, split into child orders or run the chosen time/participation plan, watch fills and market drift, and increase aggression only if fill probability falls below your threshold.