The method

How the desk trades.

Five rules, run the same way on every position — tested on 1,198 positions over 9.8 years, with real costs charged, before a single trade was published.

The rules

Five rules. One sequence. No exceptions.

The same sequence on every position. No averaging down, and nothing opened to fill a quiet day.

  1. 01

    Wait for expansion

    Nothing opens until a market breaks out of its recent range. A quiet market is left alone, however long that takes.

  2. 02

    Fix the risk

    The stop is set before the entry, at a distance the market's own volatility decides. Every position then risks the same.

  3. 03

    No take-profit

    Winners are never sold at a target. Every target tested on this method cost money; the few positions that run far carry the record.

  4. 04

    Trail the exit

    As a position moves the desk's way, the stop follows it — one way, never back. The trailing stop decides every exit.

  5. 05

    Cap the book

    Markets that move together count as one bet, and the book has a hard ceiling. When it is full, nothing new is taken.

Measured across 1,198 positions over 9.8 years.Every market the desk scans, with its grade →

Risk management

The loss is decided before the trade is

The four controls this desk runs, and what they are worth in money.
  1. The stop exists before the entry Worst case known at entry, not discovered after.
  2. Volatility sets the distance The market decides the width. The desk decides the risk.
  3. Things that move together are one bet The desk caps how many it holds in the same direction.
  4. There is a hard ceiling on the book When it is full, nothing new is taken.
15Positions max
4Per correlated group
0.50%Risk per position
1Way a stop can move

The desk's own limits on its own book — not a recommendation of what anyone else should risk. No account here is ever seen or assessed.

Free-Roll

Four steps. Every lot paid for by profit already locked.

Most desks take one position and hold it. When a trend really runs, this one buys more of it — but only after the stop has already locked in enough profit to pay for the next lot.

A real one: the desk's GBP/JPY short built to five lots — one lot alone would have returned +5.96R, the stack returned +14.36R. See it lot by lot →
Members see every lot as the desk buys it — the market and the level, on Telegram, the moment it happens. Profit factor measured over 9.8 years, September 2016 to July 2026, after costs; past performance is not a guide to future results. No method wins every position, and every one carries its stop before it opens.

lot 1 add add add exit together
Size grows. Risk does not. Every lot exits at the same stop, and each one was only added once locked profit covered it.
Profit factor Same entries · same stop · same risk per trade
One lot 1.69
With Free-Roll 2.00

The only thing that changes is when the desk adds.

9.8 years tested

1,198 positions. Real costs. Same rules.

1.99Profit factor
2.77:1Payoff, win vs loss
42%Positions profitable
13.7%Deepest drawdown
Oct 2016backtested equity, after costsJul 2026
No take-profit. Ever.
1 in 20positions run past +3R
99%of all profit comes from those
3.7×longer a winner is held than a loser
+47.9Rthe single best position in 9.8 years

That is the whole reason nothing is ever sold at a target. A rule that caps a winner deletes the handful of positions the method exists to catch — every target ever tested on this method cost money, and the best fifty positions alone produced 37% of the entire profit.

A backtest is not a record and is not presented as one — it is what the rules did on history that had already happened.

9.8 yearsSep 2016 – Jul 2026
15 of 24markets opened, the rest only watched
Real spreadcharged on every fill
0.5% riskand the same caps the desk runs
Your account

The same 1,198 positions, on your money.

$
—Equivalent yearly rate
—Ending balance
—Profit over 9.8 years
—Deepest fall from a peak
—…of the account at the time
—1R — most one position risks
—Typical win / typical loss

The backtest's own deepest fall is 13.7%, measured with size growing as the account grows. This holds risk flat, so the same dollar fall is a smaller share of a bigger balance — which is why the figure above is lower.

Sep 2016your account, replayedJul 2026

These figures are mathematical translations of the desk's published R results using the assumptions you enter. They are not forecasts, projections or guarantees of future returns. Actual execution can differ because of spread, slippage, commissions, financing and instrument-specific contract specifications.

Your selected account size and risk percentage determine your own dollar outcome. EDGEwise does not determine or recommend your personal position size.

Lot size is not calculated here — it changes with every position's stop distance and with your broker's contract spec. Risk percentage is the input that travels.

Learn the craft

Three lessons, free. Read a market in plain words.

The ideas behind every number on this site, written for traders — with what each one can and cannot tell you. Education, never advice.

Free lesson

R: one unit of risk

What does +2R mean?

Read the lesson

What it shows

R is the distance from where a position starts to its first stop. Everything the desk reports is counted in it: a position closed at its first stop is −1R, one that made twice that distance is +2R. It lets a small account and a large one read the same record, and it puts every market on one scale.

What it cannot tell you

R is not money until you choose a size: 1R is whatever you decided to risk on that position. A result in R is before your own costs and your own fills, which differ from the desk’s. And a stop can be passed through in a fast market, so a loss can be larger than 1R.

+2R
A gain of twice the distance to the first stop.
Stop moved up
A stop past the entry holds a result even if price turns back.
Free lesson

The trend on five timeframes

Why can a trend read flat while price is rising?

Read the lesson

What it shows

One rule, applied the same way to every market and every timeframe. The trend is up when price is above its 50-bar average and the 20-bar average is above the 50-bar one. It is down when both are below. Anything else is flat.

What it cannot tell you

It describes what has already happened, so it turns late: a trend is called only after part of the move is behind it. The five timeframes often disagree, and that is information, not a fault. It does not say the trend will continue.

20-bar average
The average close of the last 20 bars, weighted towards the latest.
50-bar average
The same over 50 bars, so it moves more slowly.
Flat
Price and the 20-bar average are not both on the same side of the 50-bar average.
Free lesson

Markets that move together

When are two markets one bet?

Read the lesson

What it shows

Two markets are compared on the days both traded. Two things are counted: on how many of those days they moved the same way, and how closely their daily changes tracked each other — the correlation, from −1 to 1. Markets are called one bet when every two of them have a correlation of 0.6 or more. Holding several of them has been close to holding one of them several times over.

What it cannot tell you

It is the past. How markets move together changes, and most of all in a sell-off, when markets that usually differ fall as one. A low figure is not protection, and a high one is not a forecast. It says nothing about which of two markets will move more.

Correlation
1 is always together, 0 no relation, −1 always opposite.
One bet
A group in which every two markets reach 0.6.

See it live

The same rules, on the desk's live book.

Every position as it opens, every result as it closes — losses included.

Follow the desk See the live record