How to Vet a Strategy Before You Trust It

Most people vet a strategy as a checklist and run it in whatever order the marketing material presents. That is backwards. Vetting works as a funnel (cheap disqualifiers first, expensive judgement last), and the value is in the sequence and in knowing exactly when to stop.

There is no shortage of advice on what to look at before you put money behind a trading strategy. Verification. Length of record. Drawdown. Trade count. The logic behind the entries. All of it matters. What almost nobody specifies is the order, and the order is where the time goes. Run the list wrong and you will spend an hour modelling the risk profile of a record that was never independently verified, a check that would have taken ten seconds and settled the matter. Run it right and most candidates are gone before you open a spreadsheet.

This article is the sequence and the stopping rules. It is deliberately not a repeat of the individual checks; each stage links out to the piece that covers its contents properly.

Why order of operations beats the list

A vetting checklist treats every item as equal. A funnel does not. Two properties decide where a check belongs in the sequence:

• How cheap it is. Some checks cost seconds. Others cost an evening and some judgement.
• How binary it is. Some checks return a clean yes or no. Others return a grey you then have to weigh against your own circumstances.

Cheap and binary goes first. Expensive and subjective goes last. Ordered that way, the funnel protects your attention, and the failure mode of retail due diligence is not laziness but diligence spent on the wrong candidate.

The second reason is psychological. The longer you study something, the more invested in it you become; anyone who has spent two hours building a case for a strategy finds it harder to walk away at hour three. Front-loading the disqualifiers means the walk-away decision is made while it is still cheap to make.

Stage 1: Verification, the ten-second gate

The first question is not how good the results are. It is whether the results are recorded by someone other than the person selling them. A third-party track record (a Myfxbook account linked to a live trading account, for instance) is produced by pulling the trade history directly from the broker’s server. Screenshots, spreadsheets and PDF statements are produced by whoever sends them to you.

This check is first because it is binary and free. Either a verified record exists and you can open it yourself, or it does not. Our red-flags checklist for verified versus unverified results covers what counts as verification and the ways presentation can imply it without providing it.

Walk-away trigger: the No-Record Stop. If there is no independently verified record you can open and read yourself, stop. Not “proceed cautiously with a smaller allocation”; stop. Everything downstream is analysis performed on numbers, and unverified numbers do not support analysis. A promised return, rather than a recorded one, ends the conversation here too: no trading operation controls the market well enough to guarantee anything.

Stage 2: Sample adequacy, before you read a single result

A verified record can still be too small to tell you anything. Stage 2 asks one question: is there enough here to draw a conclusion from? You are not yet asking whether the results are good, only whether they are readable. Three independent dimensions decide that:

• Elapsed time. How long the account has been running live.
• Trade count. How many closed positions the record contains. A slow strategy can run for two years and still produce a thin sample.
• Regime coverage. Whether the period spans more than one kind of market: trending and ranging, calm and volatile, at least one genuine shock. A record that covers a single regime tells you how the strategy behaves in that regime and nothing more.

Regime coverage is the one most often skipped, and it is the one that matters most. A strategy that has only ever traded a quiet trending market has not yet been asked the question you actually care about.

Set your thresholds before you look. Writing them down in advance is what stops a promising equity curve from talking you into an exception. As an illustrative starting point only (not a recommendation and not a standard), some allocators will not read a record at all below twelve months live and a few hundred closed trades. Your own numbers should reflect the strategy’s trading frequency and your tolerance for uncertainty.

Reading an adequate sample is a separate skill with its own piece: how to read a Myfxbook track record walks through the columns and what each one is telling you.

Walk-away trigger: the Thin-Sample Stop. If the record falls below your pre-written thresholds on time, trade count, or regime coverage, stop and note a date to revisit. A thin sample is not a red flag; it is simply not yet evidence. The error to avoid is reading a short, flattering record as a strong one because the line points upward.

Stage 3: Risk shape, measured against you

Only now do you read the results, and you read the risk before the return. Two figures do most of the work: how deep the worst decline was, and how long it lasted. Duration is the one people underweight, and it is what they actually abandon strategies over.

These figures are covered in full elsewhere: how to evaluate a track record before you allocate for the numbers themselves, and why drawdown, not return, defines a strategy for why they carry more information than the headline percentage.

The stage-specific point is that the measurement is against you, not an abstract standard. There is no universally acceptable drawdown, only the decline you would hold through without interfering, at the size you intend to use. Two people can read one record and reach opposite, equally correct conclusions.

So make it concrete. Take the worst decline in the record, apply it to the amount you actually intend to allocate, and read the result as a cash amount over the months the recovery took. If that number makes you flinch, you have your answer, and it is better found here than live.

Walk-away trigger: the Intolerable-Shape Stop. If the historical worst case, in your own money over its real duration, is one you would not sit through, stop. Assume the future holds a worse one, because a record contains only the declines that have happened so far. A strategy you would abandon at the bottom is worse for you than one you never started.

Stage 4: Logic and operational fit

The last stage is the most expensive, which is why it is last. By now you have a verified record, of adequate length, with a risk shape you can live with. The remaining questions are qualitative, and you put them to a human being. Two clusters:

• Logic. Can the edge be explained in plain language? What market behaviour does it depend on, and what would break it? Is risk per position fixed by a rule or set by discretion? You are not asking for proprietary detail, only whether an explanation exists at all, given without reaching for adjectives.
• Operational fit. Where does your money sit, and in whose name? Who executes, and on what platform? How do you stop, and how quickly? What are the total costs, all in?

Our questions to ask before you follow any signal or strategy is the script for this conversation. Ask them at stage 4, not stage 1; a provider’s answers are worth your time only once the record has earned it.

Walk-away trigger: the No-Explanation Stop. If the edge cannot be described in a way you can restate to someone else, stop. Secrecy is not evidence of sophistication. A matching operational stop sits alongside it: if your capital would leave your own name and control at any point, that ends the process regardless of how the record looked at stages 1 to 3.

Running the funnel

Written out, the sequence is short enough to keep in your head.

• Stage 1, verification. Ten seconds, binary, free. Stop if no independent record exists.
• Stage 2, sample adequacy. A few minutes. Stop if it falls below thresholds you wrote down beforehand.
• Stage 3, risk shape. Half an hour. Stop if the worst case in your own money is one you would not hold through.
• Stage 4, logic and fit. A conversation. Stop if the edge cannot be explained, or if custody or control of your money would leave you.

Notice that nothing in the funnel asks how much the strategy returned. Return is what you assess after a candidate survives all four stages, not what you use to decide whether to assess it; a return figure read early has a way of quietly adjusting your thresholds.

The discipline is in the stopping, not the analysing. Every stage carries an explicit trigger for the same reason a strategy carries an exit rule written before entry: the decision made in advance beats the one made while looking at an equity curve you have started to like.

Key takeaways

• Vetting is a funnel, not a checklist. The order is the method.
• Cheap, binary checks first; expensive, subjective judgement last. That is what protects your attention.
• Sample adequacy is a separate question from result quality, and regime coverage is the part most often skipped.
• Risk shape is measured against your own tolerance and your own capital, not an abstract standard.
• Write the thresholds and the walk-away triggers down before you look.

This article is education, not financial advice, and it is not a recommendation to allocate to any strategy. Any figures mentioned are illustrative only. Trading foreign exchange and CFDs carries risk, including the risk of losing your entire investment. Past performance is not a guarantee of future results, even when it has been independently verified. Consider your own circumstances and seek independent professional advice where appropriate.