Cheapest stablecoin swap on Binance
Basis points you eat swapping inventory between stables on Binance. Spread of X/USDT mid-price from 1.0000 USDT, audited every 5 seconds.
This benchmark answers the question every CEX trader has when rotating stablecoin inventory on Binance. how many basis points do I actually lose swapping X to USDT (or USDT to X) on the deepest pair available. We poll Binance's bookTicker for USDC/USDT, FDUSD/USDT and USDE/USDT every 5 seconds, record the mid-price, and measure its distance from 1.0000 USDT. p99 over 24 hours is the headline. the worst-1%-of-minutes number a desk holding inventory has to budget against on a forced rebalance. This is NOT a measurement of which stablecoin is fundamentally safest. Cash-backed USDC, monthly-audited FDUSD, and basis-trade-anchored USDe carry fundamentally different reserve risks that no order-book spread can capture. What this bench DOES tell you is. on the venue that actually carries depth for FDUSD and USDe, which pair trades tightest to USDT right now. USDC/USDT is included as a reference baseline. it's the most liquid stablecoin pair on the venue ($2.8B+ daily) and its spread is mostly USDT's own drift away from USD reaching back, so subtract its number to read each other stable's "true vs-USDT" drift. The October 10 2025 USDe flash to $0.65 on Binance USDEUSDT is the tail event the surface is wired to catch. the depeg event flag fires within 5 minutes of any pair breaking the [0.97, 1.03] band.
Methodology
We measure the live spread between each USDT-quoted stablecoin pair on Binance and the 1.0000 USDT reference. The harness polls `/api/v3/ticker/bookTicker` every 5 seconds per pair (USDCUSDT, FDUSDUSDT, USDEUSDT), records the mid-price (bid + ask) / 2, and emits `peg_deviation_usdt_anchored_bps{stable, venue}` per scrape. p50, p90 and p99 over 24 h are computed via Prometheus `quantile_over_time`. The leaderboard ranks by p99, the worst-1% tail a desk rebalancing inventory has to budget for. Calling this "stability" would be misleading. it is the live swap-cost reality of moving inventory between stables on the deepest non-US venue, not a measurement of underlying peg robustness. USDC/USDT sits on the leaderboard as a reference anchor because USDC's deviation here is mostly USDT's own drift reaching back into the pair, so any stable scoring above USDC is genuinely wider vs USDT, and any stable scoring below is in fact tracking USDT more tightly than USDC is. Polling at 5 second cadence keeps the harness inside Binance's free-tier bookTicker rate budget while preserving fine enough resolution to register a flash-depeg within seconds of the order book printing it.
Frequently asked
What does this benchmark actually measure?
The live spread between each Binance X/USDT stablecoin pair and the 1.0000 USDT reference, in basis points. Concretely. if you swap $1000 of USDC to USDT on Binance and the current USDC/USDT spread is 7 bps, you eat ~$0.70 in pair-drift cost (on top of the 0.1% Binance trading fee). This bench ranks the live swap-cost reality on the deepest non-US venue, NOT the fundamental safety of each stablecoin's reserve structure.
Is this the same as 'which stablecoin is safest to hold'?
No, and it would be misleading to read it that way. USDC is cash-backed and audited monthly. FDUSD is also cash-backed and audited monthly but Hong Kong-clearing. USDe is not USD-backed at all but anchored by a delta-neutral perp basis trade. These carry fundamentally different reserve risks that no order-book spread can capture. A tight calm-minute pair spread on Binance does not imply a robust peg under stress. for fundamental peg signals (reserve attestations, on-chain pool depth, basis-trade health) read the issuer's transparency dashboard, not this bench.
What is USDC doing on a USDT-anchored bench?
USDC serves as a reference anchor. USDC currently shows 0.084% (p99 spread vs USDT, 24 h) on Binance USDCUSDT, the most liquid stablecoin pair on the venue ($2.8B daily). USDC is the cleanest cash-backed stablecoin in the cohort, so any drift it shows here is mostly USDT's own drift away from USD reaching back into the pair. Subtract USDC's p99 from any other stable's p99 to get its true vs-USDT drift, controlling for USDT's own USD drift.
Why FDUSD and USDe specifically?
Because they have no other honest live signal. FDUSD's depth lives almost entirely on Binance FDUSDUSDT ($41M daily, no comparable USD-quoted venue). USDe's main spot book is Binance USDEUSDT ($1.8M daily, same situation). The USD-anchored bench `stablecoin-peg` (№ 014) covers USDC / USDT / DAI which have real USD-quoted venues (Kraken, Bitstamp, Curve). This companion bench is the only place where FDUSD and USDe can be measured on a live order book at all.
Did USDe really flash to $0.65 on Binance?
Yes, on Binance USDEUSDT specifically, the night of October 10 2025. A sharp liquidation cascade on perpetual futures cleared the orderbook well below the 1.0000 reference for several minutes. Ethena's USDe is not reserve-backed. it's anchored by a delta-neutral basis trade on perps, so when perp markets dislocate, the spot can follow without any redemption mechanism stepping in fast enough. The depeg event flag on this bench fires within 5 minutes of any pair breaking the [0.97, 1.03] band, so a recurrence would be visible live.
How is this different from stablecoin-peg (bench № 014)?
Bench № 014 measures USDC / USDT / DAI against $1.00 USD using Kraken USD pairs, Bitstamp USD pairs and Curve 3pool on-chain. Honest USD anchor, real peg measurement. but it can't cover FDUSD or USDe because no deep USD-quoted venue exists for them. This bench (№ 015) measures USDC / FDUSD / USDe against 1.0000 USDT on Binance. Same venue, same cadence, same currency. one is a peg-robustness measurement, the other is a swap-cost measurement on the deepest CEX. Read them together for a complete picture.
Which stable should I hold on Binance for the lowest rotation cost?
USDe currently has the tightest pair at 0.025% (p99 spread vs USDT, 24 h). The leaderboard re-sorts every minute. For inventory sizing, use p99 rather than p50, because the spread during a forced rebalance under stress is what determines actual cost, not the calm-minute median. Combine with the success-rate column and the fundamental reserve-risk story (which this bench does NOT capture) before committing meaningful inventory.
Source code github.com/OpenChainBench/OpenChainBench/tree/main/harnesses/stablecoin-peg
