Skip to main content
This page covers best practices for makers on TrueCurrent. Following these guidelines will improve your fill rate, protect you from adverse selection, and keep your maker standing in good shape.

​
Pricing fundamentals

worst_price scale: The v2 indexer endpoints return worst_price in human-readable scale (e.g. "14.85"), not 1e18-scaled integers. Reading it as a raw integer from a v1 endpoint (which would return 14850000000000000000) will cause worst price exceeds limit rejections. Always use the v2 endpoint values directly. Scale spread to volatility: Your spread represents the compensation for the risk you take by being on the other side of a trade. In low-volatility periods, tight spreads win more flow. In high-volatility periods, widen accordingly, as the risk of being picked off is higher. A simple starting model: 
spread = base_spread + volatility_factor × realized_vol_30s
Scale spread to order size. Larger orders expose you to more market impact when you hedge. A basic model: 
spread = base_spread × (1 + size_factor × quantity / reference_quantity)
Account for your inventory. If you’re net long INJ from previous fills, you’re already exposed to downside. When the next request is for a taker long (you’d go shorter), you may want to tighten your spread to encourage the fill. For a taker short (you’d go longer), widen slightly.

​
Numerical spread construction example

Below is a worked example of building a competitive two-sided quote for an INJ/USDC RFQ. Inputs:
  • Mid-market: $25.00
  • 30-second realised volatility: 0.08% (annualised ~40%; intraday spike)
  • Request size: 500 INJ (~$12,500 notional)
  • Reference size: 100 INJ
  • Current inventory: +300 INJ net long (skew toward selling)
Step 1 — Base spread: 2 bps = $0.005 per INJ (0.02% of mid) Step 2 — Volatility adjustment: 0.08% × 0.5 = +0.04% → add 2 bps each side → spread = 4 bps each side Step 3 — Size adjustment: factor = 0.5, size ratio = 500/100 = 5 → 1 + 0.5 × (5-1) = 3× → spread = 4 bps × 3 = 12 bps each side Step 4 — Inventory skew: +300 INJ net long → tighten offer (selling), widen bid (buying). Apply −2 bps to offer, +2 bps to bid. Final quote:
  • Bid: $25.00 × (1 − 0.0014) = $24.965
  • Offer: $25.00 × (1 + 0.0010) = $25.025
  • Half-spread: 10 bps bid / 10 bps offer (asymmetric due to inventory skew)
4 bps maker fee impact: Filling the $12,500 request costs $5.00 in maker fees. The minimum spread to break even on fees alone is $5.00 / $12,500 = 4 bps each side. The 10 bps half-spread above earns $12.50 gross spread minus $5.00 fee = $7.50 net spread income, before hedging costs.

​
Response rate

Always try to quote inside the collection window. TrueCurrent currently uses 500 ms. A consistent response rate is important both for your standing and for the quality of the RFQ system. Traders who request quotes and receive none get their order cancelled - a worse experience for them and lost flow for the maker pool. Fail gracefully. If you genuinely can’t price a request (e.g. your pricing feed is down, your margin is depleted), it’s better to not submit a quote than to submit one. The system handles no-quote correctly: the request expires and the taker is told to retry. Monitor latency. The collection window includes network round-trip time. If your system is co-located far from the indexer, account for that latency. A quote that arrives after the active window is ignored.

​
Tracking the mark price

To minimize adverse selection and avoid rejected quotes, makers should price around the same reference the contract validates against: the current onchain mark price. TrueCurrent’s RFQ validation uses mark_price directly. The market’s mark_price is the reference used for quote validation, worst_price validation, trigger evaluation, and liquidation checks. Position P&L display uses the streamed indexPrice, so do not use UI P&L or index price as the source for contract-facing quote bands. To track it:
  1. Query the relevant derivative market from the Injective exchange module.
  2. Read the current mark_price.
  3. Normalize price decimals and tick size before using it in quote logic.
  4. Refresh continuously and treat stale or unavailable mark price data as a no-quote condition.
You can still use external venues, private feeds, and inventory models to calculate your own fair value. Just keep contract-facing validation centered on mark price: quotes too far from the current mark will be rejected even if your offchain model says they are fair.
Last modified on May 27, 2026