How to Choose a Broker: Costs, Execution, and Safety

Commission-free is not free. This guide shows how to compare brokers on the costs that are hidden and on the protections that matter when things go wrong.

5 min readBeginnerUpdated September 16, 2026

At a glance

First filter
Regulation and client asset protection in your jurisdiction
Second filter
Total cost: commission plus spread plus financing
Third filter
Execution quality and order type support
Often decisive
Data quality, API access, and support responsiveness

Key takeaways

  • Total trading cost is commission plus effective spread plus financing plus data fees. Advertised commission is usually the smallest component.
  • Regulation and segregation of client assets matter most in the rare event that matters most: the broker failing.
  • Execution quality can be measured: compare your fills against the midpoint at order time across dozens of trades.
  • Margin terms, borrow availability, and liquidation policy determine what happens on your worst day, so read them before your first trade.
  • For systematic traders, stable API access and historical data quality outweigh small differences in commission.

Start with safety, not with fees

The cheapest broker is worthless if your assets are not there when you need them. Before comparing costs, confirm the regulatory and protection framework.

  • Regulator and licence. Confirm the licence number directly on the regulator’s register rather than on the broker’s website. Offshore entities of well-known brands often operate under much weaker rules.
  • Client asset segregation. Your cash and securities should be held separately from the firm’s own funds, so they are not part of the estate if it fails.
  • Investor compensation scheme. Coverage limits and what they actually cover (typically broker failure, never trading losses) vary widely by country.
  • Which entity you contract with. Many global brokers route clients to different subsidiaries depending on residence, with materially different protections and leverage limits.
  • Company track record. Look for regulatory actions, history of outages during volatile sessions, and how the firm handled past liquidation events.

Computing the true cost of trading

Cost componentWhere it hidesHow to measure it
CommissionPublished schedulePer share, per contract, or per trade, plus minimums
Effective spreadThe fill price itselfCompare fills to the midpoint at order time across 30+ trades
Payment for order flow effectsRouting decisionsCheck price improvement statistics and Rule 605 style reports
Financing / margin interestMonthly statementAnnual rate on debit balances, tiered by size
Borrow fees for shortsDaily accrualAsk for the indicative rate on names you intend to short
Currency conversionFX conversion on non-base tradesOften 0.2% to 1% per conversion, frequently overlooked
Market dataMonthly subscriptionsReal-time depth, Level 2, and exchange fees
Inactivity and withdrawal feesTerms and conditionsRead the schedule; common at retail CFD and forex brokers
Advertised commission is often the smallest line in this table.

Convert everything into a single figure: estimated round-trip cost as a percentage of a typical position. Then multiply by your expected trades per year. For an active swing trader that number is often between 1 and 4 percent annually, which is directly subtracted from strategy returns.

Execution quality, and how to test it yourself

  1. 1

    Record the midpoint at order time

    Before sending each order, note the bid and ask. The midpoint is your benchmark.

  2. 2

    Record your fill

    Log the executed price and any partial fills. Do this for at least 30 trades across different times of day.

  3. 3

    Compute effective spread

    Twice the absolute difference between your fill and the midpoint. Compare that figure between brokers and against the quoted spread.

  4. 4

    Check behaviour in fast markets

    Note fills around the open, around economic releases, and in the final minutes of the session. Differences between brokers are largest precisely when it matters.

  5. 5

    Test order type support

    Confirm that bracket orders, trailing stops, good-till-cancelled orders, and extended-hours routing behave as documented. Undocumented behaviour is common and costly.

What matters most, by type of trader

Trader typeHighest priorityFrequently underestimated
Long-term investorSafety, low ongoing fees, tax reportingCurrency conversion charges
Swing traderTotal cost, reliable stop orders, data qualityBorrow availability for shorts
Day traderExecution speed, routing control, platform stabilityMargin call and liquidation policy
Options traderPer-contract fees, spread order support, assignment handlingEarly assignment and exercise fees
Futures traderMargin rates, session coverage, rollover handlingIntraday margin changes during volatility
Systematic traderStable API, historical data, rate limitsSandbox environment quality and outage history

Read the margin and liquidation policy before you need it

Every broker reserves the right to liquidate positions when margin falls short, and most do so automatically, at market, without waiting for you. The specifics determine your worst-case outcome.

  • Maintenance margin levels and how quickly they change during volatility. Brokers frequently raise requirements on volatile names with little notice.
  • Liquidation order. Some close the largest position, some the most liquid, some proportionally. This decides which of your trades survives a margin event.
  • Notice period. Automatic liquidation with no call is standard in futures and crypto, and increasingly common in equities.
  • Negative balance protection. Available to retail clients in some jurisdictions, absent in others. Without it you can owe money after a gap.
  • Concentration rules. Many brokers apply special margin to positions that are large relative to the account or to the instrument’s average volume.

Frequently asked questions

Are commission-free brokers actually free?

No. Revenue comes from elsewhere: payment for order flow, wider effective spreads, margin interest, securities lending on your holdings, and currency conversion. Those costs are real but invisible on the trade confirmation. Measure your effective spread against the midpoint to see what you are actually paying.

Should I use more than one broker?

Many active traders do, for redundancy during outages, for access to different markets, and to compare execution. The cost is added complexity in record keeping and tax reporting. Having a second funded account you can trade from during an outage is a reasonable form of operational insurance.

How important is the trading platform itself?

Less than most beginners assume, more than most professionals admit. What matters is reliable order entry, accurate position and risk display, and the order types your strategy needs. Charting and indicators are widely available elsewhere. A platform that hides your true exposure is genuinely dangerous.

What is payment for order flow and should I avoid it?

It is compensation a broker receives for routing your orders to a wholesaler who executes them. Retail orders often receive price improvement over the public quote, so it is not automatically worse, but the incentive is not aligned with you. Compare price improvement statistics and your own measured fills rather than relying on the label.

Do I need direct market access?

Only if you trade actively enough that routing control and queue position affect your results, which generally means intraday strategies. For swing and position trading, the difference between direct access and standard retail routing is small compared with strategy quality and risk management.

Test this idea before you trade it

Describe the rules in plain language and AlgoTrader AI turns them into a structured strategy blueprint with a configurable historical backtest, cost assumptions, and exportable code.

Build a backtest

Keep reading

Referenced by

Educational use only. This guide explains how a strategy works. It is not investment advice, not a recommendation, and no result described here is a forecast. Test any approach on historical and out-of-sample data, size positions conservatively, and never risk money you cannot afford to lose.