Overview: What we’re analyzing and why it matters

Perpetual futures continue to be the go-to instrument for crypto derivatives traders in June 2026. Since our March update, several microstructure and tooling changes have shifted the true “all-in” cost of trading perps: fee schedules, spread and price impact, funding dynamics, liquidation design, and transaction-ordering (MEV) behavior. If your strategy targets a 10–30 basis-point (bps) per-trade edge, venue frictions of 15–60 bps still decide whether you keep that edge. The numbers tell a different story now than they did three months ago—this update gives you the concrete benchmarks and tactical rules to act on today.

Background: What’s changed since March 2026

Two developments have been decisive since March:

  • Wider adoption of intent/private-relay markets and solver auctions. More L2s and several relayer networks pushed intent-based matching and solver markets into production. That reduced median submit-to-inclusion drift on majors, but the benefit is uneven across relayers and pairs.
  • Execution specialization and venue features on both sides. CEXs lowered public taker fees to compete and rolled out faster delta-hedge APIs and improved index-smoothing for marks. DEX perps matured concentrated-liquidity primitives and fee-subsidy programs (maker-side incentives) targeted at mid-sized tickets.

These changes don’t eliminate venue-specific frictions; they re-distribute them. You must re-measure for your ticket size, frequency, and risk tolerance.

Data & evidence: Breaking down all-in cost (June 2026)

1) Explicit fees: narrower but not irrelevant

Market-observable fee ranges in June 2026:

  • CEXs: retail taker fees have compressed to roughly 3–5 bps per side (0.03%–0.05%) on major venues; maker fees often rebate or sit near 0–2 bps negative for high-volume clients.
  • DEX perps (L2): protocol fees still list low (0–3 bps per side). But add-ons matter: relayer/solver fees (1–4 bps), gas/settlement costs (typically 0.5–1 bps on modern L2s for small tickets), and any premium for private-bundles.

Put simply: the headline 1–2 bps gap is smaller than before. For high-turnover strategies (50–200 round trips/month) even a 1–2 bps difference compounds meaningfully, but for discretionary traders fill quality and adverse selection still dominate.

2) Spread and price impact: the dominant execution tax

Updated practical observations:

  • Top-of-book spreads on BTC/ETH on major CEXs range 3–6 bps, but visible depth at those prices often supports only small tickets (sub-$50k). Impact scales nonlinearly: a $100k sweep can produce 15–30 bps impact on some pairs.
  • L2 DEXs with concentrated liquidity primitives now show comparable effective spreads for mid-sized tickets ($5k–$50k). For very large tickets (> $100k on BTC, > $50k on smaller alts), marketable impact often remains worse on DEXs unless you access dedicated AMM tiers or aggregated maker inventory.

Actionable benchmark: measure a 30-day realized implementation shortfall (IS) for your typical order size. If your average adverse move between signal and fill exceeds ~15 bps, prioritize execution tactics (slicing, posting, venue rotation) before worrying about fee negotiation.

3) Funding and basis: persistence is the exploitable ingredient

Key June 2026 findings:

  • Funding rates across venues are lower on average and less persistent than in 2024–25 because automated funding hedging bots shortened mean reversion. Large, persistent funding edges (>20 bps/week) are now rare on majors.
  • Delta-neutral funding strategies remain possible but fragile: once you include round-trip slippage, gas, and delta-hedge hedging tail risk, realistic net funding capture for disciplined strategies is typically 0–10 bps/week, not the 20–40 bps sometimes quoted in early 2025.

If you see advertised weekly funding >20 bps, ask whether persistence and entry/exit costs make it real.

4) Liquidation mechanics: improved transparency but complex effects

June updates:

  • Partial-liquidation and pre-liquidation buffer features are now common on both CEXs and DEX protocols. These reduce single-shot wipeouts but increase fee churn and can amplify realized costs during volatile moves.
  • Some large CEXs standardized mark smoothing and publicized their mark-index methodology; on-chain DEXs continue to let protocols configure oracle cadence, which can sometimes lag aggressive off-chain moves and produce costly divergence during fast markets.

Trader rule of thumb: with >10x leverage or around scheduled news, choose venues where you can reproduce the liquidation trigger from public data. If you can’t, treat that opaqueness as a risk premium—add 5–15 bps to your required edge, depending on leverage and event risk.

5) MEV and ordering risk: lower median, higher variance

What changed in June:

  • Private relays and solver markets reduced median sandwich and frontrunning impact on majors; typical submit-to-inclusion drift for small-to-mid tickets on the best relayers is now ~1–3 bps.
  • However, during congestion or cross-rollup settlement events, MEV spikes remain — transient events where drift jumps to 10–20 bps. Alts and low-liquidity pairs still see material MEV and extractive ordering behavior.

Measurement is essential: log reference price at submission and inclusion, and track rolling 30–90 day median and 95th percentile drift. If your 95th-percentile adverse drift exceeds your tail tolerance, change execution mode or use a different relay.

Multiple perspectives: voices you should weigh

CEX operators and market makers

CEX firms point to lower public taker fees, faster hedging rails, and improved transparency in mark methodology as evidence that they provide superior execution on majors. Their internal metrics show sub-10 bps all-in for many retail-sized round trips on BTC/ETH, but they caution that this depends on actual fill size and not just top-of-book spreads.

DEX protocol teams and relayers

DEX teams highlight composability, reduced custody risk, and intent-based execution as advantages. They note that maker-incentive programs and concentrated liquidity on L2s deliver attractive effective spreads for mid-sized flows. Their counterpoint: smart-contract and oracle risks remain non-trivial and require monitoring.

Active traders and prop shops

Execution desks increasingly adopt a venue-agnostic approach: instrument, measure, and rotate. For latency-sensitive and large-ticket flows they prefer venues where hedges can be sourced deterministically; for mid-sized or alt pairs they route to liquidity pools with favorable depth after accounting for relayer fees.

Implications: practical guidance for traders in June 2026

Stop judging venues by headline fee tables alone. Build and update a simple all-in cost card for each market and order size you trade, with these inputs and pragmatic thresholds:

  1. Explicit fees: CEX taker ~3–5 bps per side; DEX protocol 0–3 bps plus relayer/solver 1–4 bps.
  2. Realized spread crossed: compute over 30 days; if >10–15 bps for majors at your size, optimize execution.
  3. Impact at size: if sweeping levels pushes you >20–30 bps, slice or change venues.
  4. MEV/ordering drift (DEX): if median submit-to-inclusion drift >3–5 bps or 95th percentile >10 bps, use private-bundles, solver markets, or avoid marketable entry.
  5. Liquidation cost: quantify via historical mark vs index divergence; add 5–15 bps to required edge if triggers are opaque.

Two updated examples (June 2026):

  • $10k BTC round trip, marketable: CEX: taker fees 4 bps/side = 8 bps; spread+impact ~10–15 bps = ~18–23 bps total. DEX (L2) with top relayer: protocol+relayer ~5 bps round trip; spread+impact ~12–18 bps; MEV ~1–3 bps = ~18–26 bps total.
  • $100k BTC sweep: CEX: taker fees + spread/impact often 30–50 bps total unless you have OTC or internalization; DEX: unless you access dedicated concentrated-liquidity tiers or maker aggregation, expect 40–70 bps impact and variable MEV—test on a small pilot before routing large tickets.

The winner varies by order type, ticket size, and your access to relayers or private liquidity. The numbers tell a different story for different traders—measure and adapt.

Outlook: what to watch over the next 6–12 months

  • Broader relay standardization: Expect more standardized APIs for intent and bundling across L2s; that will lower median MEV but also commoditize relayer fees.
  • Liquidity concentration and cross-venue hedge costs: Market makers will continue to favor venues where hedging is cheapest. Watch cross-venue settlement costs (wallet-to-exchange and rollup bridges) as a driver of pool liquidity.
  • Regulatory signals: Jurisdictional licensing updates and clearer custody rules will redistribute flow. Maintain multi-venue access and re-test your all-in card after any major regulatory development.

FAQ

Is a DEX perp always safer because it’s self-custody?

No. Self-custody reduces counterparty insolvency risk but adds smart-contract, oracle, MEV, and bridge risks. “Safer” depends on which risk you prioritize. Quantify both sides for your size and time horizon before choosing a venue.

What’s a realistic target for all-in round-trip cost on BTC/ETH perps?

For active traders on majors, a realistic all-in round-trip target is 15–30 bps for small-to-mid tickets (sub-$50k). For larger tickets expect 30–60+ bps unless you have special liquidity access or OTC hedges. If your measured all-in exceeds these bands, either widen your edge or improve execution.

How should I measure MEV impact practically?

Log the reference price at submission (mid or official index) and the price at inclusion or fill. Track rolling 30–90 day median and 95th-percentile drift. If median drift >3–5 bps or 95th >10 bps for your tickets, ordering costs are material—use private relays, bundling, or post liquidity instead.

Limit orders or marketable orders on DEX perps—what’s best?

Prefer posted (limit) liquidity when latency and urgency allow; it reduces MEV and adverse selection. Use intent/private-relay marketable execution when you’ve validated that the relay materially lowers submit-to-inclusion drift for your ticket size.

Single best metric to track per venue?

A rolling 30-day realized implementation shortfall (bps) for your typical order size. It bundles fees, spread, impact, and ordering costs into one number and enables apples-to-apples comparisons across venues.

Bottom line: the venue choice in June 2026 remains non-binary. Tooling improvements narrowed some gaps, but hidden frictions—fill quality, MEV tail events, and liquidation design—still move P&L more than headline fees. Measure everything you can. Fees are visible; fill quality and hidden costs are where profits disappear.