Choosing a Forex Broker for Expert Advisors

Why is the broker part of the EA’s result?

An expert advisor does not trade in a vacuum. Every order it places passes through a broker, and the broker determines the spread the advisor pays, the price it receives, the stop distances it may use, the lot sizes it may trade and whether automated trading is permitted at all. Two brokers can offer the same platform and the same instrument, and an advisor can perform very differently on each because of those conditions.

Many traders choose a broker by the headline figures on its homepage: a tight spread, a high leverage, a bonus. Those figures matter, but they describe the broker’s marketing, not the conditions the advisor will meet. The spread on the pair and at the hours the advisor trades, the minimum stop, the execution on fast moves and the terms for withdrawals are the figures that decide whether the advisor’s results can be trusted.

This guide sets out the checks to run on any broker before an advisor is used on its account. It does not rank brokers. A fair ranking would need the same tests run on each broker, over the same period, with the same costs and conditions, and that is not a test this site has run. The checks below are the method, and the record of their results is the basis for any choice.

Who regulates the broker, and which entity will you deal with?

Start with the regulator. A broker that is regulated in your region must follow its rules on client funds, on complaints and on the conduct of its business. Regulators publish public registers that list the firms they license, and those registers are the first place to check.

The name on the website is not always the entity you will contract with. A group may have several legal entities, each under a different regulator and with different rules. The terms of service name the entity, and that is the one to check on the register. An entity in one jurisdiction may offer different protections from one in another, and the difference can matter a great deal if a dispute arises.

Confirm that the licence covers the services you will use, such as trading in the instrument and account type you plan. Some licences cover only certain clients or products. Record the entity name, the licence number, the regulator and the date you checked, so that the check can be repeated later.

How do you check the spread on the pair your advisor trades?

The headline spread is usually the figure for the most liquid moment of the day, and it can be much tighter than the spread at the hours your advisor trades. Check the spread for the pair at the sessions your advisor uses, and at the rollover time. A broker that shows a tight spread in the middle of the European session may charge far more at the start of the Asian session.

Record the typical spread at each session, and the spread during a major release. A large widening around news can change the result of an advisor that trades those times, and that cost belongs in the test. The spread data should come from the broker’s own platform, observed over several days, not from a single quote.

Compare the spread with the commission, where the account charges one. The total cost per lot is the figure that matters for an advisor, and the choice between a spread-only account and a commission account should be made on the total cost for the advisor’s expected trade size.

What minimum stops and lot rules apply?

Every broker sets a minimum distance for stops and targets, and an advisor that places an order inside that distance will be refused. The minimum is published in the contract specification for each instrument, and it differs between brokers and pairs. Check it for each pair the advisor trades, and confirm that the advisor’s stops are comfortably outside it.

Check the minimum lot, the maximum lot and the lot step. An advisor that calculates a lot size outside these rules will be refused, and a rounding error in the advisor can produce this problem only on certain account balances. Test the advisor’s sizing on demo with the broker’s exact specification, so that these errors are found before a live account is used.

Is automated trading permitted on the account?

Confirm in writing that automated trading is allowed on the account type you choose. Some brokers allow expert advisors only on certain accounts, some restrict specific strategies such as scalping, hedging or martingale-style grids, and some limit the number of orders per second. These restrictions are part of the terms, and they can change, so check the current version.

A restriction that is discovered after the advisor has been running can lead to closed positions, reduced access or a dispute about the trades already made. Keep a copy of the terms in the journal, with the date, and check them again when the broker sends a notice of change.

How do execution and slippage differ between brokers?

Execution is the process by which an order is filled. A broker may fill at the requested price, at a slightly worse price or with a requote, and the behaviour can differ in fast markets. An advisor that relies on a precise entry will be affected by these differences, and the difference can be larger than the spread.

Test execution on the broker’s demo account, and then with the smallest live size once funded. Record the requested price, the fill price and any requote for each trade, and calculate the average slippage. Compare the slippage with the allowance in the advisor’s test, and adjust the test if the broker’s execution is worse than assumed.

What do the withdrawal and deposit terms say?

Read the terms for deposits and withdrawals before funding the account. Check the methods available, any fees, the time a withdrawal takes and any conditions that apply, such as a requirement to withdraw to the original source. A broker that makes withdrawals difficult or slow creates a risk that no trading test can measure.

Keep a copy of the withdrawal terms with the journal, and test a small withdrawal early in the account’s life, so that the process is known before a large sum depends on it. A broker that handles a small request promptly and clearly has shown something useful about its service.

How do you check the broker’s complaint process?

Find the broker’s complaint procedure, and note the escalation route, including the regulator and any independent dispute service. A broker that has a clear process and a regulator to escalate to offers more protection than one that does not. Check whether the regulator publishes complaint data for the firm, since that information can show a pattern.

Keep a record of all communications with the broker, including emails, chat transcripts and support tickets. A written record is the basis for any complaint, and it is far harder to assemble after a dispute has begun.

What should you record for each broker you check?

For each broker, record the legal entity, the regulator and licence number, the date of the check, the spread for your pair at your sessions, the minimum stop, the lot rules, the automated trading terms, the execution results from a live test, the withdrawal terms and the complaint route. Keep the record in one place, and date each entry.

Compare the records across the brokers you consider, using the same checks for each. A comparison built on the same checks is far more useful than a list of impressions, and it allows the choice to be explained to someone else.

What mistakes do traders make when choosing a broker?

The first mistake is to choose by the headline spread or the leverage offer, without checking the spread at the sessions the advisor trades. The second is to assume that an advisor is permitted, without reading the terms. The third is to open an account with an entity that the regulator does not cover for the services used.

A fourth mistake is to ignore the withdrawal terms until a withdrawal is needed. A fifth is to fund the account before testing execution with a small live size. A sixth is to change brokers mid-test without treating the change as a new sample, which mixes the conditions and makes the results impossible to compare.

Does the broker change the advisor’s test?

Yes, and the test must reflect it. An advisor tested on one broker’s feed and spread is a test of that broker’s conditions. Moving to another broker is a new test, with new spreads, new minimum stops and possibly new execution. Record the broker name in every test, and treat a change of broker as the start of a new sample.

This discipline keeps the results honest. A result from one broker cannot be quoted as a result for another, however similar the platforms look.

How do you read a broker’s contract specification?

The contract specification is the document that defines how an instrument trades on a given account. It lists the contract size, the point value, the tick size, the minimum and maximum lot, the lot step, the minimum stop distance, the margin requirement, the trading hours and the swap rates. For an advisor, almost every number in it affects a trade, and reading it carefully is one of the most useful checks a tester can make.

Start with the contract size and the point value. They determine the money risked per point and per lot, and an advisor’s sizing depends on them. Then check the minimum stop, the lot step and the margin requirement. A wrong assumption about any of these can make an advisor look profitable on paper and fail in practice. Save a dated copy of the specification with the journal, because the broker may change it, and a change can alter the risk of a trade without any change to the advisor.

Check the trading hours as well. An instrument may close for a period each day or each weekend, and an advisor that expects continuous trading may fail to manage a position through a closure. The specification is the source for those hours, and the advisor’s session settings should match it.

How do you compare account types?

Brokers often offer several account types, distinguished by the spread, the commission, the minimum deposit and the permitted trading methods. An advisor may be allowed on one type and not another, and the difference can be decisive. Compare the account types on the total cost for the advisor’s expected trade size, not on the headline figure alone.

A spread-only account may look cheaper, but its spread can widen at the times an advisor trades most. A commission account may have a tighter spread and a fixed charge per lot, which can be easier to model. Run the advisor’s test on each account type you consider, and compare the results on the same bars. The account type that fits the advisor’s trade frequency and size is the one that performs better in practice.

Check whether the account type limits the number of open trades, the number of orders per second or the use of hedging. Those limits can change how an advisor behaves, and they must be reflected in the test.

What should you do before the first live trade?

Before the first live trade, confirm that the advisor’s settings match the broker’s specification for each pair. Check the minimum stop, the lot step, the magic number, the symbol name and the server time offset. A mismatch in any of them can cause errors that are easy to miss on demo.

Place a small test order manually, if the broker allows it, to confirm that the account, the platform and the funding behave as expected. Record the fill, the spread and any delay. Then start the advisor at the smallest size, and watch the first sessions closely, recording every trade and every error in the journal.

How do you handle a broker that changes its terms?

Brokers change their terms from time to time, and the change may affect an advisor directly. A new minimum stop, a new restriction on automated trading or a change in the spread on a key pair can alter the result of a test that was valid before. Keep a dated copy of each version of the terms, and check for changes at the start of each month.

When a change affects the advisor, treat the new conditions as a new sample. Record the change, the date and the reason, and test the advisor under the new terms before continuing. A test that spans a change without a note is a test whose results cannot be trusted.

What does a complete broker record contain?

A complete broker record contains the legal entity and the regulator, the licence number and the date of the check, the account type and its cost, the spread for each pair at each session, the minimum stop and lot rules, the terms on automated trading, the execution results from a live test, the deposit and withdrawal terms, the complaint route and the date of each review. It is a single document that a reader can check without contacting the broker.

Keep the record current. A record that was accurate a year ago may now be misleading, and each entry should show when it was last checked. A review of the record at a fixed interval, such as each quarter, keeps it reliable.

How do you handle an account restriction after it is opened?

An account can be restricted after it is opened, for reasons ranging from a change in the broker’s policy to a review of the trading pattern. A restriction can prevent automated trading, close positions or limit access. Read the notice carefully, check whether it applies to your account and note the date on which it takes effect.

If a restriction affects an advisor, stop the advisor before the restriction applies, if you can do so, and check the open positions. Ask the broker in writing to confirm what the restriction covers and how open positions will be treated. Keep the correspondence, because it is the record of the restriction and of what was agreed.

Why should you test the broker’s support before you need it?

A support team is most useful when a problem is urgent, and that is the time when a slow or unclear response does the most damage. Test the support early, with a simple question about the terms, and record the time taken and the quality of the answer. A broker whose support answers clearly and quickly in a quiet period is more likely to help in a crisis.

Ask a specific question about automated trading in the test. A clear, written answer that cites the relevant term is a good sign. A vague or contradictory answer is a warning that the terms may not be applied consistently, and that inconsistency may affect the advisor.

How should a live test be structured?

A live test begins with the smallest lot the broker allows, on a funded account whose balance is an amount you can afford to lose. It runs the advisor unchanged for a fixed period, recording each trade with its spread, fill and slippage. The test ends at the end of the period, and the result is compared with the demo results on the same pair and sessions.

The paper trading guide explains the criteria that should be met before this test begins. Do not increase the size during the test, and do not change the advisor’s settings. Any change starts a new test. Keep the journal with the broker’s name, the account type and the date of each trade, so that the live record can be traced to the conditions under which it was made.

How do you weigh a broker’s reputation against the checks?

Reputation is an input, not a substitute for the checks. Reviews, forum comments and social media posts can describe a broker’s service, but they are often written by people with a stake in the outcome, and they can be manipulated. A review that describes a single bad experience may be accurate, and a long run of praise may be paid for. Read reviews as a source of questions to check, not as evidence in their own right.

The most useful reputational information is the regulator’s record: complaints data where it is published, enforcement actions and public warnings. Those sources are official, and they can be checked. Weigh them alongside the spread, the execution and the terms, and let the sum of the checks, not any one opinion, shape the decision.

Why should you keep an open mind about a broker that scores poorly?

A broker that fails one check may still be acceptable for a particular advisor. A high spread may not matter for a slow advisor that trades rarely. A restriction on hedging may not affect an advisor that never hedges. The purpose of the checks is to match the broker to the advisor, and a poor score on one dimension is a reason to look harder, not a verdict.

Record the reason for every rejection and every acceptance. A list of reasons shows whether the checks are consistent, and it prevents a broker from being rejected on a feeling that cannot be explained later. Review the list once a quarter, and check that the reasons still apply to the brokers as they are today, since terms and regulation can change after a decision was made. A decision that was sound when it was taken may no longer be sound after a change, and the review is the moment to notice. Write the date of each check beside its result, so that the age of every piece of evidence is visible, and a reader can judge how much weight each result deserves before relying on it. Where a check could not be completed, say so in the record rather than leaving the entry blank, because a gap that is named is far more useful than one that is hidden. Carry the open item to the next review, with the reason it is still open and the date by which it should be closed, and a note of who is responsible for closing it, so that open items do not drift from one review to the next without anyone noticing. A short list of open items at the top of the record, updated at each review, makes the state of the checks visible at a glance, and it gives the next reviewer a clear starting point rather than a long document to reread, and it keeps the record useful long after the person who wrote it has moved on to other work, which is the only test that matters for a record meant to outlast its author, and it should be applied before any decision is taken on the strength of the record.

What does this guide not promise?

This guide does not name a best broker, and it does not promise that any broker will suit any advisor. Conditions change, brokers change their terms and regulators change their rules. The checks in this guide must be repeated before each account is opened and reviewed periodically afterwards.

The guide also does not recommend opening an account with any firm because it is discussed here. The purpose is to give you the method to check a broker for yourself, and the choice remains yours.

What is the offer in this guide?

The offer is the EUR/USD EA evaluation checklist, a printable checklist that covers the evaluation of an advisor on a broker’s demo and live account. It includes the spread check, the minimum stop check, the execution record and the review of the results. Use it alongside the broker checks in this guide, once you have chosen a broker to test.

The checklist does not replace the regulator’s register or the broker’s terms. It organises the evidence you gather from them, so that the choice can be reviewed later.

How do you start this week?

Start with the broker you already use, or the one you are considering. Find its legal entity in the terms, search the entity on the regulator’s register and record the licence. Then read the automated trading terms and the withdrawal terms, and note any restriction that affects your advisor.

Next, check the spread for your pair at the sessions the advisor trades, and the minimum stop in the contract specification. Run the advisor on demo with the broker’s exact specification, and record the spreads and fills. Only when those checks are complete should a live account be opened, and then at the smallest size.

For how the lot size interacts with the broker’s margin, read the leverage guide, and for the arithmetic of a stop and its cost, the pip guide.

For more free tools and checklists, visit the resource library.