Futures grid bot funding fees are payments or receipts tied to contracts still open at each funding settlement, separate from profit on completed grid cycles. A bot can show positive grid profit while funding, trading fees and losses on its remaining position reduce total equity. Leverage changes the collateral supporting that exposure and may also change order size, so review both quantities and margin.
This guide concerns grids trading perpetual futures. It uses an original hypothetical USDT-linear ledger to connect a completed cycle with the position left behind. The examples explain accounting and operating decisions; they do not describe a KLINEO grid-bot feature or establish profitable settings.
Identify what the grid-profit number includes
A completed buy-and-sell pair answers a narrow question: what did that matched cycle earn? It does not explain the full result of a strategy that still holds contracts.
Bybit's documented futures-grid calculation deducts the paired trades' fees from grid profit. Its total P&L adds closed profit and unrealized P&L, with funding included in the closed-profit calculation. These are that venue's definitions, so check the equivalent fields on the platform you use. Bybit futures-grid P&L calculations.
Build your reconciliation from individual fills and cash flows. If a displayed cycle profit already subtracts both trading fees, subtracting those same fees again understates the result. Conversely, the fees on entries that have not completed a cycle still belong somewhere in the full ledger.
Subscription, hosting and other operating expenses can sit outside the exchange's bot dashboard. The complete trading-bot cost worksheet adds those expenses. Keep them separate while you first make the exchange balance, position and funding records agree.
Follow one profitable cycle into an accumulating position
Assume a fictional long grid starts with 600 USDT, no position, and no deposits or withdrawals. Each filled order is 0.01 BTC. Every fill pays an invented 0.05% commission in USDT. The fill prices, marks and rates below are constructed assumptions, not observed trading results or current exchange charges.
The configured range is 58,000–61,000 USDT. This teaching sequence assumes the listed fills occur in order; it does not simulate a venue's initialization process, order queue or matching engine. The first cycle closes completely before the later position builds, avoiding a mismatch between paired-cycle and average-entry accounting in this example.
| Event in sequence | Fill price or settlement mark, USDT | BTC open afterward | Realized cash change, USDT |
|---|---|---|---|
| Buy 0.01 BTC | 59,000 | 0.01 | −0.2950 fee |
| Sell 0.01 BTC; complete one cycle | 60,000 | 0 | +10.0000 price P&L; −0.3000 fee |
| Buy 0.01 BTC | 60,000 | 0.01 | −0.3000 fee |
| Buy another 0.01 BTC | 59,000 | 0.02 | −0.2950 fee |
| Funding settlement A at +0.020% | 59,500 | 0.02 | −0.2380 funding |
| Buy another 0.01 BTC | 58,000 | 0.03 | −0.2900 fee |
| Funding settlement B at +0.040% | 58,000 | 0.03 | −0.6960 funding |
| Review at a lower mark | 57,000 | 0.03 | No new cash flow |
The completed cycle's gross price profit is 0.01 × (60,000 − 59,000) = 10 USDT. Its paired commission is 0.295 + 0.300 = 0.595 USDT, leaving 9.405 USDT of net cycle profit.
The three later entries remain open. Their quantity-weighted average entry is (600 + 590 + 580) ÷ 0.03 = 59,000 USDT/BTC. At the review mark, their unrealized price P&L is 0.03 × (57,000 − 59,000) = −60 USDT.
All five fills cost 1.480 USDT in commission. Funding costs 0.934 USDT. The full result at that snapshot is therefore:
10 − 1.480 − 0.934 − 60 = −52.414 USDT
Equity is 600 − 52.414 = 547.586 USDT. The modeled USDT balance before unrealized P&L is 607.586, which explains why a balance increase can coexist with an equity loss.
You can also reconcile from the net cycle number: 9.405 − 0.885 − 0.934 − 60 = −52.414. Here 0.885 is only the commission on the three unmatched entries. This second route checks that the completed cycle's 0.595 commission was counted once.
Assume the position survives every modeled observation under its applicable maintenance rules. This is a performance ledger, not a liquidation simulation. Closing the remaining contracts would introduce an actual exit price and further commission; the review mark does not promise an executable exit.
Calculate funding from the position at each settlement
For its USDT perpetual contracts, Bybit documents funding as contract quantity multiplied by mark price and the funding rate. A positive rate makes longs pay shorts; a negative rate reverses that direction. Contract specifications matter: this formula should not be copied unchanged into an inverse-contract worksheet. Bybit funding-fee calculation.
Settlement A uses the 0.02 BTC then open, not the final 0.03 BTC and not the total quantity traded during the run:
0.02 × 59,500 × 0.00020 = 0.238 USDT paid
Settlement B uses the larger remaining position:
0.03 × 58,000 × 0.00040 = 0.696 USDT paid
The decimal conversion matters: 0.040% becomes 0.00040. Neither payment is calculated from the 600-USDT starting contribution. Funding follows exposure, while the contribution determines what resources initially support it. For a fixed contract quantity, the leverage setting is not another multiplier in this funding formula.
Carry can continue when completed cycles stop arriving. In a separate flat-price extension, hold 0.03 BTC at a 58,000 mark and assume a constant +0.040% rate. Each further settlement costs 0.696 USDT. Fourteen would cost 9.744 USDT, exceeding the earlier 9.405 net cycle profit without another price loss or completed cycle. This is conditional arithmetic, not an expectation that those rates or holdings persist.
The perpetual funding settlement ledger develops changing rates and contract units further. Record actual settlement cash flows separately from the estimated next charge; an estimate does not establish what was ultimately debited.
Separate a leverage setting from filled quantity
Changing the collateral behind an unchanged position does not change its dollar price P&L or the funding arithmetic above. With the same fills, costs and marks, the 52.414-USDT loss is 8.74% of a 600-USDT starting allocation. Against a hypothetical 300-USDT allocation, it is 17.47%.
That comparison assumes both runs remain open and satisfy maintenance requirements throughout. It does not mean a smaller allocation would actually survive the path, nor does it calculate either run's liquidation price. Comparing percentages without checking survival can conceal the very event that invalidates the comparison.
There is another complication in a bot setup screen: selecting higher leverage can change contract size. OKX's futures-grid documentation says per-grid quantity depends on investment excluding reserved margin and on selected leverage. Its reserved margin is excluded from per-grid sizing. OKX futures-grid sizing and reserved margin.
Consequently, compare the actual order ladder after a settings change. If each order becomes larger, price exposure and settlement funding can both increase. A higher displayed return on allocated margin may reflect larger exposure rather than a better grid process.
Use the cross-versus-isolated collateral worksheet to identify which resources can support losses. Record later margin contributions as transfers. Adding support does not erase the price loss, and it changes the amount of capital committed to the run.
A price range does not define the liquidation boundary
In the example, the review mark is below the configured lower grid level. That fact alone does not establish that the contracts closed. Bybit explicitly says its futures-grid positions remain open outside the configured range while new grid orders stop. Its bot FAQ separately identifies liquidation at a bot Maintenance Margin Rate of 100% or above. Those are Bybit implementation rules. Bybit futures-grid range and liquidation behavior.
A range answers where the strategy places orders. A maintenance calculation answers whether the current exposure has sufficient support. Treat them as separate observations, with timestamps and the relevant account scope.
Pending orders matter too. OKX describes its estimated liquidation price as a projection after pending orders fill. Check which holdings a displayed estimate assumes before comparing it with a stop or grid boundary. OKX futures-grid terminology.
The practical stress question is what happens if more entry orders fill before a recovery. Recalculate quantity, average entry, funding notional and margin requirements for that state. A price move can simultaneously add exposure and reduce equity; inspecting only the currently filled position misses that possible next state.
Complete a reconciliation before the next operating decision
Use one consistent snapshot time for balances, positions and pending orders. Preserve the venue's native records, then build the following worksheet alongside them.
| Evidence to record | Validation or decision it supports |
|---|---|
| Contract, settlement asset and quantity unit | Confirm that the funding formula uses the right notional and currency |
| Filled quantity, price, fee and trade ID | Rebuild cycle profit and unmatched-entry costs without duplicate fees |
| Position size at each settlement and actual funding cash flow | Separate changing exposure from a changing funding rate |
| Remaining order ladder and projected filled position | Check exposure after further entries, not only exposure now |
| Equity, margin mode and venue risk metric with timestamp | Compare the actual account state with its maintenance rules |
| Contributions, withdrawals and closure instructions | Distinguish transferred cash from performance and intended exits from fills |
Two operating traps deserve explicit checks. First, a pause can leave contracts open: OKX's documented pause action cancels pending orders but retains the position. Verify the resulting position before assuming price exposure or future funding has ended. OKX futures-grid pause behavior.
Second, avoid treating a local countdown as settlement evidence. Bybit cautions that opening or closing within five seconds of funding time does not guarantee inclusion or exclusion. Use the exchange's actual funding record and confirmed fills to resolve the result. Bybit settlement-timing caveat.
If promotional claims present frequent winning cycles as proof of guaranteed returns, use the trading-bot claim-verification checklist before relying on the claim or granting credentials.
Before continuing a futures grid, make its completed cycles, remaining contracts, funding debits and current equity reconcile. Then inspect the margin state after plausible pending fills. If either calculation is unresolved, a positive grid-profit number does not supply the missing evidence.