Crypto futures risk limit tiers are venue-defined exposure bands that can change maintenance margin requirements and maximum permitted leverage as a position grows. The relevant calculation uses the contract's current tier, valuation rules and any maintenance-margin deduction, rather than one fixed percentage. Check positions, qualifying open orders and account mode before assuming that an unchanged leverage setting means unchanged margin requirements.
A useful tier review asks two separate questions: which band applies, and how much maintenance margin that band requires. Confusing the tier's quoted rate with the final amount can produce a large calculation error at a boundary.
This article uses an original hypothetical schedule to show the difference. It is a reading method for exchange rules, not a live parameter table, liquidation calculator or recommended position size.
Read the whole row, including the deduction
A tier row may contain a position-value limit, maintenance margin rate, initial margin rate and maximum leverage. Bybit's risk-limit API documentation exposes those fields, plus a maintenance-margin deduction. They describe different constraints; a maximum leverage value is not the maintenance rate.
For Bybit USDT perpetual and expiry contracts, the documented core position calculation is:
Maintenance margin = position value × tier maintenance rate − tier deduction
Its USDT maintenance-margin guide explains the deduction through a tier-by-tier calculation. It also notes that the position-tab amount includes an estimated closing fee. A worksheet that excludes that fee should not be expected to equal the displayed total.
The deduction is part of the schedule, not a rebate paid to the trader. In a graduated schedule it prevents applying the higher marginal rate to exposure already covered by lower bands. Do not discard it because its name sounds optional.
Work across a boundary with a fictional schedule
Assume a hypothetical USDT-linear contract whose position value is quantity multiplied by mark price. There are no open orders. The example excludes fees, funding, collateral adjustments and all other positions.
These parameters are invented for teaching and are not current terms for any exchange or symbol:
| Tier | Position value in USDT | Maintenance rate | Maintenance deduction in USDT | Maximum leverage |
|---|---|---|---|---|
| 1 | Up to 100,000 | 0.50% | 0 | 50x |
| 2 | Above 100,000 to 250,000 | 1.00% | 500 | 25x |
| 3 | Above 250,000 to 500,000 | 2.00% | 3,000 | 10x |
The second deduction is 100,000 × (1.00% − 0.50%) = 500. The third is 250,000 × (2.00% − 1.00%) + 500 = 3,000. Within these fictional bands, the formula therefore matches adding each slice at its own rate.
| Hypothetical observation | Position value | Applicable tier | Core maintenance calculation | Core requirement |
|---|---|---|---|---|
| Current holding | 90,000 | 1 | 90,000 × 0.005 | 450 USDT |
| Exactly at first cap | 100,000 | 1 | 100,000 × 0.005 | 500 USDT |
| Just above first cap | 100,001 | 2 | 100,001 × 0.01 − 500 | 500.01 USDT |
| Proposed larger holding | 120,000 | 2 | 120,000 × 0.01 − 500 | 700 USDT |
| Exactly at second cap | 250,000 | 2 | 250,000 × 0.01 − 500 | 2,000 USDT |
| Further expansion | 260,000 | 3 | 260,000 × 0.02 − 3,000 | 2,200 USDT |
The first boundary raises the marginal rate, but the requirement does not suddenly double. At 100,001 USDT, ignoring the deduction produces 1,000.01 USDT instead of 500.01. That is a 500 USDT overstatement.
For the proposed increase from 90,000 to 120,000, the core requirement rises by 700 − 450 = 250 USDT. Applying the old 0.50% rate to the whole new position gives 600 and understates the requirement by 100. Applying 1.00% without the deduction gives 1,200 and overstates it by 500. Both shortcuts fail differently.
These amounts are maintenance requirements, not the cash needed to open the increase or a maximum tolerable loss. Initial margin and available collateral need their own checks. The leverage-versus-position-size guide separates those quantities.
An unchanged quantity can enter another tier
In the same fictional contract, 1.8 units at a 50,000 USDT mark represent 90,000 USDT. At a 60,000 mark they represent 108,000, even if the trader submits no new order. The worksheet moves from tier 1 to tier 2, and core maintenance becomes 108,000 × 0.01 − 500 = 580 USDT.
This observation alone does not establish that account risk increased: equity, unrealized PnL and collateral may also have changed. It establishes that freezing the tier while updating prices is inconsistent with this example's valuation rule.
Bybit documents real-time tier changes as mark-valued exposure changes, alongside automatic tier adjustments based on positions and active orders. Its risk-limit rules also distinguish isolated and cross modes from portfolio margin. The latter uses portfolio-level risk, so a simple single-position tier worksheet is not a substitute for its calculation.
Open orders need a separate exposure check
A common failure is to inspect the current position, see room below a cap, and submit several additions that all assume the same remaining room. Before any fill, those instructions can affect the venue's exposure calculation or margin reservations.
Do not automatically add every visible order. Bybit's cited risk-limit rules define effective exposure by position mode and exclude Reduce Only and Close on Trigger orders from that calculation. Whether an order can add exposure depends on its actual exchange configuration and state.
There is also a distinction between determining a tier and calculating maintenance for an unfilled order. Keep a position-only illustration like the table above separate from the venue's order-margin calculation. Reconcile accepted orders and fills before interpreting the next account snapshot.
Keep a reusable tier-change worksheet
Use this record before an exposure increase and refresh it afterward:
| Record | Question it resolves |
|---|---|
| Venue, account mode, exact contract and timestamp | Which rules and observation apply? |
| Current quantity and valuation price | What is the position value now? |
| Qualifying open orders and position mode | What counts toward effective exposure? |
| Applicable floor, cap, rate and deduction | Which complete tier row applies? |
| Proposed post-fill and stressed-price values | Could execution or repricing change that row? |
| Initial margin, maintenance and fee components | Which requirement is each number measuring? |
| Exchange-confirmed account state after action | Did the observed result match the assumption? |
Obtain parameters for the relevant account where required. Binance's signed USDⓈ-M notional and leverage brackets endpoint returns bracket floors, caps, maintenance ratios and maximum initial leverage. It also documents a user-specific symbol bracket multiplier when adjusted. A generic screenshot may therefore be insufficient evidence for an account's bracket.
Retain the retrieval time and full row, not just the rate. If the response is stale or unavailable, flag the calculation as unresolved. Reusing a previously valid table without its date conceals the missing evidence.
A tier calculation does not prove an exit will work
Maintenance rules describe support requirements. The mark-price and liquidation guide covers the relevant risk reference and account context. The futures trigger-price comparison separately explains why a stop watching another series may activate at a different moment.
Even a matching source does not guarantee an exit. A stop order's price-protection condition can add another guard to activation, and the resulting order still needs to execute. Include the exit's actual state in a review of remaining exposure.
For a shared collateral pool, use the cross-versus-isolated margin guide to identify what supports the obligation. A correct tier calculation for one position cannot establish the health of that entire pool.
Editorial method: Prepared with AI assistance on 23 October 2026 from the official sources linked above. The tier schedule and calculations are original hypothetical teaching examples. No live exchange account or KLINEO margin calculation was tested.