Trade Execution: What Decides Live Performance
Trade execution is where paper alpha goes to die. Most managers underestimate how much disappears between the model and the fill.
That is not a marketing line. It is what shows up once model returns are separated from realized execution. A strategy that prints a 2.0 Sharpe in research can settle closer to 1.1 in production, not because the signal stopped working, but because spreads, market impact, missed fills, and routing decisions quietly handed the alpha to someone else. The model did not fail. The execution chain did.
For an algo or quant investor, that distinction matters. Execution is not a back-office function that happens after the interesting work. It is where research becomes investable, or starts leaking away. In the decks we review, the gap between simulated cost assumptions and live transaction-cost analysis is one of the cleanest tells of whether a manager is likely to hit their numbers.
The Algotrader.ch team considers execution the single most consequential operational variable in algorithmic and quantitative trading. Not the signal. Not the data. The execution. It is the dimension where the distance between what managers describe and what careful review reveals is most consistently large. Closing that gap, systematically and publicly, is part of what The Review does.
The useful part: the questions that separate a working execution process from a described one are not especially technical.
How trade execution works once money is live
At portfolio level, a manager decides what should be bought or sold. Trade execution is the discipline of doing it without giving away too much value on the way in or out. That includes order submission, venue selection, routing logic, timing, interaction with market makers or dealers, and post-trade review.
Execution quality directly affects realized returns, especially in strategies with meaningful turnover, smaller-cap exposure, less liquid instruments, or stressed markets. What a strategy actually earns is decided by its live fills.
Polished top-line performance can mislead here. Two managers run similar signals and similar risk controls. The one with better execution often ends up with the more durable live record, because the implementation gives less away.
What drives execution quality
Execution quality is shaped by a handful of operational variables. None is exotic. Together, they decide more than most pitch decks admit.
| Driver | What it means | Why investors should care |
|---|---|---|
| Slippage | The gap between expected price and executed price | Directly erodes returns and exposes weak live implementation |
| Market impact | The price movement caused by the order itself | Becomes critical for larger orders and capacity claims |
| Fill rate | How much of the intended order gets completed | Low fill quality distorts exposure and risk targets |
| Latency and timing | How quickly and when orders reach the market | Matters most for short-horizon or fast-decay strategies |
| Routing quality | Where orders are sent and how venues are selected | Reveals whether best execution is real or merely asserted |
| Liquidity awareness | How execution adapts to volume, spreads, and depth | Important for scalability and for avoiding forced trading at poor prices |
| Opportunity cost | The cost of not getting the trade done | Missed or partial fills leave the portfolio with unintended exposures |
Plenty of managers can talk about execution. Fewer can show that their live fills still hold up when size grows, liquidity thins, or markets stop being accommodating.
Good execution keeps leakage low enough that the live strategy still resembles the one described in research and marketing, rather than just looking clever on a transaction-cost dashboard. That sounds modest, but in real portfolios it is one of the quieter dividing lines between an interesting process and an investable one.
- Managers who describe best execution in adjectives and produce little evidence from the periods that test it
- Slippage reported as a single blended number, which conceals the analysis rather than presenting it
- Backtests using closing prices as if a meaningful position size can be traded there without market impact
- Execution quality documented only in benign periods, with no regime breakdown across stressed sessions
Where execution returns survive and where they die
Execution costs do not hit every strategy equally. A low-turnover global macro portfolio can absorb more friction than a signal set that rebalances aggressively or trades thinner names. That is obvious in theory, yet strangely easy to gloss over in manager presentations.
Take a short-horizon equity strategy with attractive gross returns and frequent turnover. On paper, the edge per trade looks sufficient. Live trading begins, and three things happen: spreads widen during stressed periods, partial fills increase, and the manager has to cross the spread more often to stay aligned with the model. The backtest still held up; the edge did not, because it was too small once implementation costs were paid.
Now compare that with a less glamorous strategy: slower turnover, tighter liquidity filters, position limits that bite, and execution constraints that stop the book from chasing every signal. The expected return may look less exciting in research. Live results can hold up better because the manager can trade the book at scale. That trade-off deserves more attention than it gets.
Be careful with “high Sharpe, low capacity cost” claims. Soft execution assumptions usually mean soft capacity estimates too.
Execution choices matter because the market does not stay still
Most readers know the basic tools. Market orders prioritize immediacy. Limit orders prioritize price control. Routing choices change where liquidity is found. The broader issue is what the portfolio becomes after those choices are made.
Some managers preserve price discipline and miss exposure. Others chase immediacy and give away too much in spread and impact. A third group leans on broker language about smart routing, while showing almost nothing about how venue quality is monitored.
Context decides the answer. Ask whether the execution choices fit the strategy’s horizon, liquidity profile, and turnover. Sophisticated terminology helps only if the fills support it.
How weak execution reporting reveals itself
Execution attracts vague language because it is easy to sound sophisticated without showing much. Terms like smart routing, best execution, low latency, and institutional access can hide thin substance when no evidence sits underneath. Six patterns come up often.
- Gross returns presented without realistic transaction costs. Net returns appear only in footnotes, if at all.
- Turnover disclosed without implementation shortfall or slippage by regime.
- Broker relationships described warmly while venue selection remains opaque.
- Capacity claims based on average volume, with little discussion of stressed liquidity.
- Backtests using closing prices as if size can be traded there without consequence.
- Execution quality measured only in benign periods.
One subtle problem appears when managers report slippage as a single blended number. A single blended slippage number is not transaction-cost analysis. It is a summary that hides the analysis. Good execution review separates normal conditions from difficult ones, then breaks results out by instrument type, order size, time of day, and volatility regime.
If the data is always smooth, be suspicious, because markets never are.
Capacity often breaks in execution before it breaks anywhere else
Managers often discuss capacity as if it were mostly a portfolio-construction issue, when it usually becomes visible first in execution: wider impact, slower fills, more leakage around crowded windows, weaker exits when liquidity thins.
Average daily volume is only a rough starting point. Rigorous capacity analysis asks harder questions. How much of the position has to be established quickly? How much trading is concentrated at the open or close? What happens when the strategy grows and the market becomes less cooperative?
A strategy loses capacity in its live fills, where orders actually meet the market.
Judging trade execution: three things to look for
A compact way to assess execution looks at cost, control, and credibility together rather than as separate boxes to tick.
On cost: how much return is lost between decision and fill? Look for implementation shortfall, slippage, spread paid, and how those numbers behave under pressure, not just in average periods. A manager who can only show benign-period data has either not stress-tested the methodology or has chosen not to share the results.
On control: does the manager own the execution process, or mostly outsource it and rely on the broker to care? Real control shows up in routing rules, venue oversight, escalation procedures, and post-trade review, all of which sit on the manager’s trading infrastructure. Outsourcing is workable on its own; the real problem is a lack of oversight over outsourced execution.
On credibility: do the execution claims fit the strategy’s turnover, horizon, instruments, and scale? If the setup sounds demanding but the reported drag is negligible, caution is sensible. There is a record of every order sent, every price achieved or missed, every partial fill. The question is whether that record is being examined or just stored.
Three execution profiles that separate theory from investability
Strong model, weak execution
A quantitative equity manager trades mean-reversion signals with daily rebalancing. Backtests show strong gross alpha. Live results land well below expectation. Post-trade analysis reveals persistent adverse selection: the fund buys just before liquidity disappears and sells into shallow bids. The model signal still has merit. Execution timing and order slicing are the problem. The result is tradable in research and falls short of investable in production.
Modest signal, disciplined implementation
Another manager runs a slower multi-factor book with explicit liquidity thresholds, position caps tied to average daily volume, and tighter rules for volatile opens and closes. The backtest is less dramatic. Live performance is steadier. The manager accepted a smaller theoretical edge in exchange for a higher probability of realizing it. That is not a compromise. It is often the point.
Capacity breaks before the marketing deck says it does
A strategy works well at smaller asset levels. As assets grow, orders become a larger share of daily volume. Market impact rises non-linearly. Execution quality deteriorates first in stressed sessions, then more broadly. Reported capacity was based on average conditions. Real capacity was lower. Familiar failure mode.
Knight Capital is the canonical example of execution-chain failure compounded by absent governance. On August 1, 2012, a deployment error caused the firm’s algorithms to flood the market with unintended orders for 45 minutes. The loss was $440 million. The cause was the operational chain around the algorithm rather than the trading algorithm itself. Pre-trade controls were inadequate, the kill procedure was not tested, and nobody had clear authority to switch off live capital fast enough. Most execution failures are quieter. They look like a 30 basis point drag nobody flagged because nobody measured it properly.
What to verify before trusting a manager’s execution process
For due diligence, broad questions are not enough. Ask for proof that links process to outcomes.
Execution evidence
- Show slippage relative to decision price, not just end-of-day marks.
- Break out results by instrument, order size, volatility regime, and trading horizon.
- Show how partial fills, missed trades, and delayed exits are tracked.
- Show execution quality in difficult periods, not just average periods.
Routing and venue oversight
- Who controls routing logic: the manager, broker, or platform?
- How is venue quality reviewed over time?
- What evidence shows that “best execution” is measured rather than merely stated?
Capacity discipline
- What assumptions sit behind capacity estimates?
- How did execution quality change as assets grew?
- What happened to fills and market impact in stressed markets?
Governance and reporting
- Is there a written trade policy and escalation process?
- How are trading errors handled and disclosed?
- Is transaction cost analysis routine, or produced only when investors ask?
A capable manager answers these cleanly. A weak one retreats into abstractions.
The execution question is not whether the process sounds sophisticated. It is whether it still works when liquidity worsens, size grows, and markets become less forgiving.
Trade execution sets the standard for whether a strategy is real
Trade execution gets underestimated because it sits between idea and outcome. Many investors spend more time on signals than on fills. That is backwards. If a manager cannot turn portfolio intent into disciplined, repeatable market outcomes, the edge is weaker than it looks.
The Algotrader.ch team evaluates the strategies featured in The Review against exactly this standard. Execution evidence has to be specific, regime-aware, and verifiable. Most managers do not pass. The live observations informing that standard are tracked in Research Notes as The Review’s profiles develop.
FAQs
Is trade execution mainly a concern for high-frequency or very active strategies?
No. It matters most there, but even lower-turnover strategies can suffer from poor execution in less liquid instruments, larger orders, or volatile markets. The impact appears less often. It can still be material.
What is the simplest way to spot weak execution reporting?
Ask for transaction cost analysis, implementation-shortfall reporting, stress-period evidence, and an explanation of how execution quality changed as assets grew. You will not see every trade, but you can judge whether the manager actually measures execution or only talks about it.
What is the difference between slippage and market impact?
Slippage is the gap between the expected price and the executed price. Market impact is one cause of slippage: your own order moves the price against you. Slippage also comes from delay, volatility, poor routing, or thin liquidity.
Can a manager outsource execution and still have strong execution quality?
Yes, but only if execution is monitored properly. Outsourcing does not remove responsibility. A diligent manager should still be able to show routing oversight, venue review, transaction-cost analysis, and evidence that delegated execution is being measured rather than assumed.