MENTAT
← funding arb · catalog

day 2 of 5

Picking the venues

Why Hyperliquid + Nado, and how to evaluate a perp DEX before you put size through it.

<!-- SKYTO: review/edit before publishing -->

The strategy doesn't care which DEX you pick — it cares about the properties of those DEXes. Course of action: pick venues with characteristics that match the trade, not venues with the loudest brand.

The properties that matter

1. Funding payment cycle. Hourly is best for arb. 8-hour cycles work but compound slower and react slower to flow shifts. Anything longer is a research project, not a trade.

2. Open interest depth. Your fill price scales with size relative to OI. A $10K position in a $500K OI market is going to move price; a $10K in $80M OI is invisible. Rule of thumb: stay under 1% of OI on entry.

3. Funding rate cap. Some venues cap funding (Binance: ±0.5%/8h, ~547% APR). Caps protect you from the venue's manipulation but also cap your upside in extreme regimes. HL has higher caps than most CEXes; that's a feature for the strategy.

4. Oracle robustness. The funding calculation depends on the index price. If the oracle is fragile or thin, the venue itself is fragile. Look for: multiple feed sources, on-chain attestation, no single-keyholder admin override.

5. Withdrawal posture. "Custody risk" is a polite phrase for "what's the exit if the team goes dark". Decentralised execution on HL is good. Permissioned backstops are bad.

Why HL + Nado specifically

Hyperliquid: runs on its own L1, hourly funding, deep OI on majors (BTC/ETH/SOL/HYPE), low latency, no KYC. The on-chain orderbook makes it the closest thing crypto has to a CEX-quality matching experience while staying non-custodial.

Nado: newer, but the funding cycle and orderbook design are HL-influenced. Two reasons it earns a spot:

  1. Splits venue risk — if HL halts, Nado lets you continue.
  2. Funding rates between the two often diverge — you can long the cheaper side and short the expensive side, capturing the spread instead of either absolute rate.

If you only run on one, run on HL. Two-venue structure is what unlocks the "advanced" version of this strategy in day 4.

What to check before you size up

□ Funding rate (8h annualised) — is it positive?
□ Funding rate cap — what's the worst case I'm signing up for?
□ OI on the target market — am I under 1%?
□ Last-1h volume — is anyone actually trading this?
□ Spot venue for the hedge — same exchange? cross-margin?
□ Withdrawal queue depth (manual check) — how long if I want out today?

The agent we'll wire up on day 4 runs this checklist for you continuously. For today, run it manually on three markets you're considering. Note which ones you'd say no to and why.

quiz · 3 questions

Q1. When evaluating a perp DEX for funding-arb, the most important variable is:

  1. The aesthetic of the front-end
  2. Open interest depth and how often funding pays
  3. Whether the team is anonymous
  4. The size of the airdrop teaser

explain → Depth bounds your size; payment cycle bounds how fast your edge can compound and how predictable your cash flow is.

Q2. Hyperliquid pays funding every:

  1. 1 hour
  2. 4 hours
  3. 8 hours
  4. Every block

explain → HL's hourly funding cycle is a feature for arbitrage — frequent rebalancing opportunities and faster information signal.

Q3. A 'venue concentration risk' issue means:

  1. Too many venues to integrate
  2. Too much capital on a single exchange that could halt withdrawals
  3. Funding rates being correlated
  4. Latency between venues

explain → If FTX-2025 hits, single-venue concentration kills you. Splitting across HL + Nado is the cheapest insurance you can buy.

homework

List 3 perp DEXes you've heard of (besides HL + Nado). For each, write one line about what worries you (regulatory, liquidity, oracle design, custody risk, etc).

e.g. dYdX v4: thin spot pair coverage. GMX: slippage on size. Drift: oracle reliability...

Homework is acknowledged, never graded — do it for yourself.