Mark Prices

  • Updated

What is the Mark Price and why it matters

The mark price is a reference price used for key risk processes such as unrealized P&L, margin calculations, and liquidations. Its purpose is to reflect a fair and robust estimate of market value, even when order books are thin or temporarily distorted.

Rather than relying on a single last trade or the top of the order book, Deribit’s mark price combines multiple market signals and applies several stabilizing mechanisms. This helps ensure that positions are not liquidated due to short-lived price spikes, manipulation attempts, or illiquidity.

Warning

Due to the factors mentioned, particularly the lag in genuine market moves and flash crash protection, the mark price should not be relied upon for making trading decisions. For example, in futures trading, sudden price movements often result in the mark price falling outside the bid-ask spread.

Key building blocks of the mark price

1. Implied forward prices from options (dated futures only)

For dated futures, Deribit enhances the futures order book with implied forward prices derived from options markets.

  • Call and put options at the same strike are combined using call-put parity to infer a synthetic forward price.

  • Multiple strikes are evaluated, and the strike with the tightest implied bid-ask spread is selected.

  • The resulting implied bid and ask are added to the futures order book as virtual liquidity.

Rationale: 

Options markets are often more liquid than long-dated futures. Using option-derived prices improves price discovery when futures trading is sparse, while remaining grounded in tradable market data.

2. Volume-weighted average price (VWAP)

Instead of using the best bid and ask alone, Deribit computes prices using a VWAP over a configurable depth of the order book.

  • A VWAP is calculated separately for bids and asks using a defined notional depth.

  • The mid price is then locally bounded to remain within a narrow range defined by the best executable bid and ask prices, ensuring consistency with observable market quotes.

Additional safeguards ensure the resulting price stays close to executable market quotes.

Rationale: 

VWAP reduces sensitivity to small orders or spoofing at the top of the book and better represents realistic execution prices.

3. Term structure consistency (dated futures only)

For dated futures, prices across expiries are expected to follow a smooth term structure.

  • Prices from nearby expiries on the same index are used to interpolate or extrapolate a fair premium for a given expiry.

  • The interpolated price is bounded by current market quotes.

  • The mark price is then constrained to stay within a configurable percentage range around this term-structure estimate.

Rationale: 

This prevents illiquid contracts from drifting to unrealistic prices relative to neighboring expiries, while still respecting observable market prices.

4. EMA smoothing of the futures premium

Rather than smoothing the raw futures price, Deribit applies an exponential moving average (EMA) to the futures premium, defined as:

Futures price − Index price

  • The EMA is calculated over a defined time window.

  • The smoothed premium is then added back to the index price to form the mark price.

Rationale: 

Smoothing the premium preserves responsiveness to genuine market trends while filtering out short-term microstructure noise. It also ensures the mark price remains anchored to the index.

5. Dynamic final bounds around the index

As a final safeguard, the mark price is constrained within bandwidths around the index price.

Rationale: 

These bounds reduce the risk of unnecessary liquidations caused by extreme or transient price deviations, especially as expiry approaches.

End-to-end mark price workflow (simplified)

  1. Derive implied bid and ask prices from the most liquid option strike (dated futures).

  2. Compute VWAP-based bid and ask prices from the futures order book.

  3. Calculate and locally bound the mid price.

  4. Apply term structure bounds (dated futures).

  5. Smooth the futures premium using an EMA.

  6. Add the smoothed premium to the index price.

  7. Apply final dynamic bandwidth limits.