How to use the screener

It tests one futures idea, within the day or held for days, against years of recorded tick data and tells you whether it holds up. You describe the idea in plain English, the desk writes it as exact rules and reads them back, and the screen replays it session by session, charges commission and slippage on every fill, and puts the result through seven gates written down before it runs. Measured on the machine that runs the screens, a screen written in the Strategy Language usually takes about 15 to 25 seconds (about 25 on 1-minute bars), and a heavy one, with orders waiting on both sides and thousands of trades, up to about three minutes. On tick bars, one that takes a hundred thousand trades or more can take about five minutes.

A position held overnight or over a weekend can open well beyond its stop. The screen fills that stop at the opening price, and a real order can fill worse, so a loss can be larger than the stop you set.

• Checking a setup before real money rides on it. A verdict takes seconds.
• Testing, in its exact form, a rule you read about, were shown, or have traded on instinct.
• Comparing one idea across markets. The same rules on NQ and on ES are two screens, and the verdicts can differ, because the cost of trading differs.
• Finding out which gate an idea fails, so you know whether it needs more trades, a wider bracket or a different idea.
• Measuring a surviving idea against a funded account’s own rules before you pay for the evaluation.

No. Write the idea the way you would say it to another trader. The desk writes it as rules and lists, under every reply, what it chose where you left something open; a setting you leave out takes the language’s default or the indicator’s usual setting, marked as a default. Before anything runs, the whole specification is read back clause by clause for you to check, and no credit is spent until you press Run.

Yes. When you press Run, the desk asks for an email address and sends a six-digit code; the screen runs once you type it. No password and no card. The free screen shows the verdict, whether each gate passed, and the line written by the gate that decided it. The dollar figures behind a verdict are part of a plan: Pro is $29 a month for 10 screens with the full report on each, and Pro Max is $59 for 40. Compare the plans.

46 U.S. futures markets, each judged on its sessions from the date below except the most recent ones, which are held back from every research path and never scored. There is no date range to choose.

• NQ — E-mini Nasdaq 100, from January 2018.
• ES — E-mini S&P 500, from January 2018.
• CL — Crude Oil, from January 2018.
• YM — E-mini Dow, from September 2021.
• GC — Gold, from March 2022.
• NG — Natural Gas, from February 2023.
• RTY — E-mini Russell 2000, from May 2023.
• SI — Silver, from August 2023.
• ZN — 10-Year T-Note, from January 2018.
• ZB — 30-Year T-Bond, from January 2018.
• 6E — Euro FX, from January 2018.
• 6J — Japanese Yen, from January 2018.
• 6C — Canadian Dollar, from January 2018.
• HG — Copper, from January 2018.
• PL — Platinum, from January 2018.
• HO — Heating Oil, from January 2018.
• ZC — Corn, from January 2018.
• ZS — Soybeans, from January 2018.
• ZW — Wheat, from January 2018.
• ZM — Soybean Meal, from January 2018.
• EMD — E-mini MidCap 400, from January 2018.
• ZF — 5-Year T-Note, from January 2018.
• ZT — 2-Year T-Note, from January 2018.
• TN — Ultra 10-Year Note, from January 2018.
• UB — Ultra T-Bond, from January 2018.
• 6B — British Pound, from January 2018.
• 6A — Australian Dollar, from January 2018.
• 6S — Swiss Franc, from January 2018.
• 6N — New Zealand Dollar, from January 2018.
• 6M — Mexican Peso, from January 2018.
• M6E — Micro Euro FX, from January 2018.
• RB — RBOB Gasoline, from January 2018.
• BZ — Brent Crude, from January 2018.
• MGC — Micro Gold, from January 2018.
• ZL — Soybean Oil, from January 2018.
• KE — KC HRW Wheat, from January 2018.
• LE — Live Cattle, from January 2018.
• HE — Lean Hogs, from January 2018.
• GF — Feeder Cattle, from January 2018.
• MES — Micro E-mini S&P 500, from May 2019.
• MYM — Micro Dow, from May 2019.
• M2K — Micro Russell 2000, from May 2019.
• SIL — Micro Silver, from February 2020.
• MBT — Micro Bitcoin, from May 2021.
• BTC — Bitcoin, from January 2022.
• MNQ — Micro E-mini Nasdaq 100, from December 2022.

Results are quoted per contract of the market a rule trades: a full-size contract, or a micro on MES, MNQ, MYM, M2K, M6E, MGC, SIL and MBT. A trade is flat by the close unless your rule holds it overnight. A position held overnight or over a weekend can open well beyond its stop. The screen fills that stop at the opening price, and a real order can fill worse, so a loss can be larger than the stop you set.

A screen flat by the close gets one of three verdicts:

• Candidate: it survived every gate.
• Inconclusive: there are too few trades to tell the result from noise, so the sample cannot answer the question.
• Not this one: it stopped at a gate. The report names the gate and prints the line it wrote.

Most ideas stop somewhere. Of the 105,204 specifications screened here to date, about 1 in 187 cleared every gate.

A rule that can hold a position overnight is a swing screen. It is graded candidate, promising, doubtful or dead rather than judged gate by gate, and it is never inconclusive.

The grade says how likely its edge over random entries on matched days, held the same way, is to be real and to clear a set hurdle. Its own trades, the same rule on the other markets it runs on, and what the record of swing screens knew before it ran all count as evidence, and the report says how much of the answer each carried.

• Candidate: likely a real edge over holding the market.
• Promising: more likely real than not, not yet enough to call it.
• Doubtful: probably no edge over holding the market.
• Dead: no edge over holding the market, or the causality or cost check failed.

How likely the edge must be for each grade, and the hurdle it must clear, were set by a calibration study on rules with no edge and rules with a planted one.

A position held overnight can open beyond its stop, so a loss can be larger than the stop you set.

Questions fixed before any run, in the order they run, each answered by one or more checks:

• Causality: Could the signal have read a bar it could not have seen live?
• Edge vs. noise: Is the average trade clearly above zero: over enough trades, not carried by a handful of sessions, and after counting every idea screened here?
• Drift null: Does it beat random entries with the same trade count, hold and long/short mix?
• Walk-forward: Does it hold on sessions the selection never saw?
• Mirror test: Does the opposite trade, taken on the same bars, do clearly worse?
• Cost: Does the edge clear real per-fill commission and slippage, with room to spare?
• Era split: Did it also pay in the years before the sample it was judged on? Reported on survivors; it never decides a verdict.

The first check that fails decides the verdict. The checks after it still run and are shown greyed on the report: they decide nothing, so a pass among them is not a pass for the idea. How the gates work explains each one.

Yes. The sample report is a real screen from the record, shown in full: the verdict, the seven gates in the order they ran, and the numbers a plan adds — dollars and drawdown per contract, the year-by-year record and the equity curve. On a paid plan your own screens look the same, with every trade to download; the free screen shows the verdict, whether each gate passed, and the line written by the gate that decided it.

On a paid plan, every screen comes with the full report, whatever its verdict:

• Dollars per year and maximum drawdown per contract, net of costs.
• The year-by-year record.
• The figure behind every gate.
• Every trade it took, as a spreadsheet: entry, exit, reason, R and dollars.

Every report has a permanent link that opens for your account alone, and every plan emails it to you when it finishes.

Read the deciding gate first. It tells you whether the idea is dead or only under-resolved, and the desk will explain it if you ask. Changing one thing for a reason you can state is a new, honest question. Changing things until something passes is the search the gates exist to catch, which is why every report quotes how many specifications have been screened.

Second questions about a screen that has already run, one credit each:

• Leak audit: Is this result an artefact of the data rather than the market?
• Regime read: Which market states does this live in, and which kill it?
• Capacity and impact: How many contracts can this take before the edge is gone?
• Funded-account fit: What does this do to my funded account, under its own rules?

Pro can order the regime read and leak audit; Pro Max can order all four. The leak audit can be ordered on any verdict; the others need a candidate or an inconclusive result.

They are ordered on a screen made from the menu: a screen written in the Strategy Language cannot take them yet.

No. A screen tells you whether the specification survived the gates; it builds nothing for a trading platform, and nothing it produces connects to your broker or your account.

Yes, a free one, and it takes only an email address. Your first message to the desk, or your first Run, asks for it and emails a six-digit code; the message goes, or the screen runs, once you type it. The free plan is 8 desk questions, which do not reset, and one screen, with no password and no card.

A question is any message to the desk: shaping an idea, asking what something means, asking about a result. A credit is one screen or one follow-up report. Only pressing Run or ordering a report spends a credit; the desk never does.

Each month the desk answers 8 questions on Free, 100 on Pro and 300 on Pro Max.

Unused credits roll over for one billing period while you stay subscribed, then lapse. A plan can be cancelled at any time from your account; it runs to the end of the period you have paid for, and its unspent credits expire when it ends.

The credit goes back to your account automatically. A verdict of “not this one” is a delivered answer rather than a failed run, so it is not refunded. The refund policy covers the rest.

You keep ownership of every specification you submit. The terms let Waypoint Quant store and process it to run the service, keep it in the internal research record, and use aggregate, de-identified information about it.

Nothing is published in a form that identifies you or your parameters unless you opt in. The one opt-in is the question under the verdict: an idea that does not pass can be added to the public research notes anonymously, and a candidate never is. Every screen counts toward the published totals, and only the counts are published.

The record is kept by specification, so someone who later arrives at exactly the same specification is told that it was screened here, when, and its verdict (with its figures, on a paid plan), never by whom. The terms and the privacy policy are the full version.

Yes, with a share link. A report's own address opens for your account alone. Under a finished screen, Share a link makes a separate link that anyone can open: it shows the verdict, the rules that were screened and the gate that decided it, never the figures, the trades or your own words, whatever plan the viewer is on. Making a new link stops the old one, and you can stop a link at any time.