DCA safety-order volume rounding

DCA safety-order volume rounding changes the submitted quantities after a size multiplier is applied. Calculating every intended order from the original unrounded amount and rounding once per order preserves a different schedule from multiplying the previous rounded quantity. Check the rule, units and accepted ladder before treating a geometric preview as executable.

A multiplier such as 1.2 does not uniquely describe the orders an exchange receives. The missing detail is which number the bot multiplies next: the mathematical target, the submitted quantity or the actual fill. Small differences can accumulate across a ladder.

The multiplier needs a rounding policy

The 3Commas TradingView parameter guide describes its safety-order size multiplier as the relationship between successive order sizes. That establishes the intended growth rule for that script, but does not establish how any live bot propagates exchange rounding.

The distinction matters because permitted quantities live on a grid. Binance Spot's filter documentation defines a LOT_SIZE step alongside minimum and maximum quantities; it separately documents MARKET_LOT_SIZE for market orders. A strategy's unrounded number can fall between accepted increments.

For a positive quantity and a hypothetical increment s, define downward quantization as:

F(q) = s × floor(q / s)

This means discard the fractional step. It is a chosen budget-preserving policy for the examples, not a universal exchange instruction to round down. Rounding up or to nearest produces other schedules and may increase the purchase commitment.

Two implementations can then interpret the same first target q1 and multiplier m differently:

• Independent calculation: submitted_n = F(q1 × m^(n−1)). Each order starts from the original mathematical schedule.
• Recursive calculation: submitted_1 = F(q1); later orders use submitted_n = F(submitted_(n−1) × m). Each rounded result becomes the next seed.

The volume-scale capital guide explains why larger multipliers concentrate funding in later orders. Here the multiplier stays fixed; only the propagation of rounding changes.

Six additions from the same intended first size

Consider an invented unleveraged spot ladder. The first safety-order target is 1.37 base units, the multiplier is 1.2 and the quantity increment is 0.1 units. There are six additions. Prices below are hypothetical limit prices, assumed valid on a separate price grid. Ignore fees, assume the quantities meet all other instrument filters, and assume every order fills completely at its stated price. The base order is excluded from both totals.

The second independent order starts from 1.37 × 1.2 = 1.644, then becomes 1.6. The second recursive order starts from the already rounded 1.3, giving 1.3 × 1.2 = 1.56, then becomes 1.5. By the sixth order, the submitted quantities differ by 0.4 units.

Summing quantity times the corresponding price gives 124.90 quote units for the independent ladder and 114.72 for the recursive ladder. The latter buys 1.1 fewer base units and spends 10.18 fewer quote units under these assumptions. Identical displayed settings therefore need not produce identical deployment.

This does not make the larger ladder better. It creates more exposure. The comparison identifies a configuration difference; it does not estimate profit or recommend either policy.

Why the small loss can grow

Independent downward quantization loses less than one quantity increment on each positive target. With six targets and a 0.1-unit grid, its total discarded quantity is less than 0.6 units. In the example it is 0.3039904 units.

Recursive quantization also discards less than one increment at each rounding operation, but it multiplies earlier losses into later targets. The first discarded 0.07 units would itself have grown with the multiplier. Comparing the recursive sequence with the original unrounded sequence therefore gives a larger total shortfall: 1.4039904 units.

A particularly revealing edge case is a plateau. Start at an already permitted 1.0 units, use a 1.05 multiplier and round down to 0.1. Recursive sizing calculates 1.05, submits 1.0 and repeats forever. Independent targets reach 1.1025 on the third addition and submit 1.1. A multiplier above one can therefore leave a recursively rounded ladder flat.

If the first target rounds to zero, a recursive sequence seeded from that zero remains zero. Zero is not a useful substitute for a valid order. The bot needs an explicit invalid-size outcome rather than a silent success or an automatic increase in capital.

Quote budgets require another calculation

The worked table scales base quantity. Many DCA configurations instead specify a quote-currency amount. For an independently calculated quote budget B_n and submitted limit price P_n, a simple positive spot conversion is:

quantity_n = F(B_n / P_n)

The budget sequence and the base-quantity sequence now have different growth. A falling price can increase units even if the quote budget stays constant. Do not take yesterday's rounded base quantity, multiply it and assume that implements a quote-budget multiplier.

Keep the price used for planning beside the submitted price. A candle close or signal price does not establish a fill, as the signal-close versus execution-fill guide explains. Some APIs let spot market buys specify quote units directly: Bybit's place-order reference documents its marketUnit choice. Use the exact account and order-path semantics rather than applying this limit-order conversion everywhere.

The decimal-precision guide covers exact representation and serialization. Decimal arithmetic prevents accidental numerical artifacts; it cannot choose the intended schedule for you.

Audit the ladder with a short worksheet

Export or reconstruct one row per intended safety order before trusting the total:

Carrying residual quote value forward is a third policy. It can bring later spending closer to a cumulative budget, but increases some later orders and needs a defined cap. It should appear in the worksheet rather than being hidden inside a multiplier.

Recheck minimum notional, maximum size and available funds after quantization. A valid increment alone does not establish an executable order; the exchange-minimums worksheet handles that separate feasibility problem. Do not silently round upward to repair a minimum failure.

Finally, reconcile actual fills before changing the exit. A partially filled addition must not seed the next schedule merely because its fill quantity is convenient, unless that behavior is the stated design. Use actual holdings for the take-profit recalculation after fills. This article demonstrates alternative sizing policies; it does not claim that KLINEO exposes a particular rounding selector or residual-carry feature.

Editorial method: This AI-assisted explanation uses synthetic spot quantities, independently checked decimal arithmetic and primary documentation reviewed on 23 October 2026. No trading outcomes were observed.