Grid Range

Every other measurement tool answers a question about what price did. The Grid Range answers a question about what a strategy would cost. You drag a box over the price band you expect the market to oscillate in, and the tool lays a full grid ladder inside it: the base at the near bound, a rung at every level, take-profit targets stepping away from each entry, the break-even the filled ladder settles at, and the stop if you arm one.

The numbers are not illustrative. Order sizes come from the same lot-scaling mode catalog the trading engines run, so a mode picked here scales exactly the way it will scale live. Rung prices, capital totals, average entry, take-profit ladders and stop-loss losses are all computed with the same formulas the Grid Designer uses. What the tool reports on the chart is what a bot configured the same way would actually commit.

That makes this the fastest path from a chart observation to a funded plan. Spotting a range is the easy part; knowing that a 20-level martingale grid across it needs six times the capital of a fixed one, and that the drawdown at the far bound would be larger than your account, is the part that normally requires leaving the chart. Here it is one drag and a glance.

1. 1
   Select Grid Range

   Pick the tool from the Measurement group in the drawing toolbar.
2. 2
   Drag across the range

   Click at one bound and drag to the other. Dragging downward builds a long grid; dragging upward builds a short one. Release to place.
3. 3
   Choose the lot scaling mode

   Open the first panel and pick a mode. Start with Fixed to see the plan at its simplest, then compare against Capped Martingale or Square-Root Martingale.
4. 4
   Fix the level count or the spread

   Open the Grid panel and choose which of the two you want to pin, then enter it. The other is solved from the range and shown directly beneath as you type. Leave spacing on Equal price unless you specifically want constant-percentage steps.
5. 5
   Enter the capital per lot

   Open the Capital panel and set what one base rung commits. The panel immediately reports the lot total, the capital the whole ladder needs, and the profit per level.
6. 6
   Tell it what you have

   Still in the Capital panel, fill in "I have". If the plan needs more, the box gets a red badge and the figures report how many levels your capital actually covers.
7. 7
   Check the downside

   Turn on the Risk group in the Figures panel, then read the drawdown at the far bound. Arm a stop if you want one and check how many rungs fill before it fires.
8. 8
   Read the execution

   The Execution figures show what the ladder would actually have done over the bars in the box: rungs filled, targets hit, round trips turned, and what is still open. Widen the box horizontally to judge it over a longer period.
9. 9
   Adjust the range

   Drag any corner or edge. Every number recomputes live, so you can widen or tighten the range until the capital, the drawdown and the fill rate all sit where you want them.

Three things are inputs: the range you drag, the capital per lot, and one of either the level count or the spread. Everything else is computed — including the total capital the ladder needs, which is the figure most of this tool exists to produce.

Level count or spread

The range fixes the relationship between the two: a 10% range holds ten 1% rungs, or twenty 0.5% ones. Pin either and the other follows. Both directions describe identical ladder maths, so a grid designed one way can be reproduced exactly by feeding the other way its answer.

Capital per lot in, capital needed out

Capital is always asked for per lot and always reported as a total. That direction matters: per-lot is the number you can actually decide, because it is the order size the venue will accept, while the total is a consequence of it and of how aggressively the grid mode scales. Asking for the total instead would hide the thing that makes grids expensive — the size curve.

Spacing: equal price vs equal percentage

Equal price spacing puts the rungs a constant price distance apart. This is what the engines do, so it is the default and the only choice that reproduces a live ladder exactly. Its consequence is that deep rungs represent larger percentage moves than shallow ones: on a range from 100 down to 80 with ten levels, the first step is 2% of the entry and the last is 2.5%.

Equal percentage spacing puts every rung a constant ratio from the previous one, so each step is the same percentage move. That distributes risk evenly across depth and is the right frame for comparing grids on assets at very different price levels. Treat it as a design lens rather than a deployable configuration.

The mode is the single biggest determinant of what a grid costs. Across the same range with the same level count, a fixed grid commits one base lot per rung while an exponential martingale commits 1024 at level ten. The size profile on the left edge of the box exists precisely so this is visible rather than buried in a total, and the risk dot in the picker exists because the label alone does not convey it.

Notional is the total the plan commits: the opening position plus every rung, each sized by the mode. Margin is that divided by leverage, and it is the number your budget has to cover. Average entry is volume-weighted across the whole filled ladder, which is why a martingale mode pulls it much closer to the far bound than a fixed one does.

The Targets group answers a different question from the Profit group. Profit says what reaching an exit is worth; Targets says where the exits actually are and how far price has to travel to reach them. Keeping them apart matters because a grid can look profitable while its targets sit somewhere the market rarely goes.

Profit per cycle assumes every rung fills and every take-profit is hit once — a complete oscillation through the range. It is a best-case figure by construction, and useful as an upper bound rather than an expectation. The per-grid row divides it by the exit count, which is the number that matters when judging whether individual round trips clear the fees.

The drawdown at the far bound is the honest worst case the range itself describes: every rung filled, price sitting at the edge you drew. It is not the worst case overall — price can leave the range entirely — but it is the loss the plan is explicitly designed to absorb, so if it is unacceptable the design is wrong regardless of anything else.

Rungs before stop is the check most grid designs fail. A stop measured from the base at a distance shorter than the range width sits inside the ladder, so it fires before the deep rungs ever fill — the grid takes the drawdown without getting the recovery. Either widen the stop past the far bound or narrow the range.

The box has a width as well as a height, and the width is not decoration. The price span says what the grid trades; the time span says over what period it is being judged. With replay on, the tool walks every bar inside that window and executes the ladder against it: entries fill when a bar reaches them, targets fill when a bar reaches the target, and each rung is closed slice by slice exactly as the engine ladders its exits. Drag the right edge forward and the same grid is re-evaluated over more history.

This turns a design tool into a first-pass audit. A grid that looks well-placed can still turn out to have filled only its top two rungs in six months, or to be sitting on eight open rungs with the market below the whole ladder. The Execution figures say which of those it is.

What the model does and does not know

Only bar extremes are available, never the path within a bar. When one bar both reaches a rung and reaches that rung's target, the round trip is counted — the standard convention for grid backtests, and optimistic by exactly the cases where price gapped past the entry and back without ever resting there. The effect is negligible on low timeframes and material on high ones, which is why the bar count is printed alongside every execution figure.

The opening position is a market order, so it is entered once at the start of the window and never re-enters after it closes. Ladder rungs are resting orders and do re-arm, which is what produces multiple round trips on a single rung. Slippage, partial fills, funding and liquidation are not modelled. Fees are charged per actual traded notional at half the round-trip rate on each side, so they track real activity rather than the hypothetical full cycle the Profit figures use.

The tool can report over fifty figures across six groups. Showing all of them at once would produce a panel taller than most charts, so the Figures panel gives two levels of control: a master switch per group and a switch per figure. The group switches exist because the tool gets used two different ways — designing a grid needs Range, Capital and Profit; auditing one needs Execution and Risk — and flipping between those should not mean re-selecting a dozen rows.

• Sizing a grid to a range — find the consolidation band you expect price to oscillate in, drag the tool across it, and read off what a grid there would cost. This is the tool's core job: turning a chart observation into a funded plan.
• Comparing lot-scaling modes — with the box in place, cycle through modes and watch the notional and the size profile change. A fixed grid and an exponential martingale over the same range differ by orders of magnitude at depth, and the profile makes that visible instantly.
• Checking a budget covers the ladder — set the budget to the capital you actually have. If the plan overcommits, a red badge appears on the box and the Affordable row tells you how many levels the budget really funds.
• Stress-testing the downside — turn on the Risk section and read the drawdown at the far bound. If losing that much would be unacceptable, the range is too wide, the mode too aggressive, or the order size too large — before any capital is committed.
• Placing the stop honestly — arm the stop and watch Rungs before stop. A stop inside the range fires mid-ladder, which means the grid never gets the chance to work; a stop below the range only triggers when the thesis is already broken.
• Tuning level count against profit per grid — more levels book more round trips but each is smaller and the whole thing costs more. The Per grid row makes the trade-off explicit rather than a guess.
• Fitting a grid to the account you have — fix the spread you want, set the per-lot size the venue allows, fill in what you actually have, and read the capital needed against it. This answers “can I even run this?” before any of the strategy questions.
• Finding the densest grid a capital base supports — with per-lot and capital fixed, walk the spread down until the capital needed crosses what you have. The spread just before that crossing is the tightest grid the account can fund.
• Pricing a scaling curve — hold the range, spread and per-lot fixed and change only the grid mode. The capital needed moves by a factor of five between a fixed grid and a linear martingale over ten levels, which is the honest cost of faster averaging.
• Validating fees are survivable — enter the venue's round-trip fee. On a tight grid the fee can consume most of the step, and the net-per-cycle row shows exactly how much is left. The Fee share figure flags it directly once fees exceed a quarter of gross.
• Checking a range was actually traded — drag the box across past price action and read the fill rate. A grid whose deep rungs never filled was wasted capital; one whose deepest rung filled repeatedly was too shallow. This is the fastest way to tell the two apart.
• Counting the round trips a range produced — the Round trips and Targets hit figures say how many times the grid would have cycled over the window. Divide the realised profit by the period and you have a grounded estimate rather than the best-case per-cycle number.
• Finding where a grid gets stuck — turn on rung markers and look for rungs with fills but no round trips. Those are the levels that absorbed inventory and never got it back, and they are usually the reason a grid that looks profitable on paper is not.
• Handing a plan to the wizard — once the range and settings read the way you want, the same numbers configure a real bot in the Grid Designer. The bounds give the base price and the spread, the mode and sizes carry over directly.

• 1.Grid mode catalog — Quatick engine — grid_modes.hpp / lot_calculator.py
• 2.Ladder plan math — Quatick Grid Designer — ladderPlan.ts
• 3.Grid Trading Strategy — Quantitative trading reference
• 4.Martingale and Anti-Martingale Position Sizing — Risk management literature
• 5.Dollar-Cost Averaging — Investment research