How It Works

The Trading Loop

Strike runs continuous short-duration prediction markets (default: 5 minutes). Each market asks a simple question:

Will BTC/USD be above $X at time T?

Where $X is the strike price (captured from Pyth at market creation) and T is the expiry timestamp.

Traders express their view by buying outcome tokens on the orderbook:

  • Buy YES if you think the price will be above the strike

  • Buy NO if you think the price will be below the strike

Step by Step

  1. Market opens — A new market is created with a strike price and expiry. The orderbook begins accepting orders.

  2. Place orders — Traders submit limit orders at their desired price (0.01–0.99). A YES token priced at 0.70 means "70% chance price is above strike." Deposit collateral or existing outcome tokens to back your orders.

  3. Batches clear — Every ~3 seconds, all pending orders are matched at a single uniform clearing price. If bids and asks cross, a clearing price is found that maximizes matched volume. The oversubscribed side gets pro-rata partial fills.

  4. Claim fills — After a batch clears, traders claim their filled amounts. You receive outcome tokens (if buying) or collateral (if selling).

  5. Trading halts — When less than one batch interval remains before expiry, the book stops accepting new orders. The final batch clears normally.

  6. Market resolves — After expiry, anyone can submit a signed Pyth price update to resolve the market. The contract verifies the update cryptographically and determines the outcome.

  7. Redeem winnings — Winning outcome tokens redeem 1:1 for collateral. Losing tokens are worthless.

What Makes FBA Different?

In a continuous orderbook, the first order to arrive gets priority — this creates speed races and MEV extraction. In a Frequency Batch Auction:

  • All orders within a batch window are treated equally (no time priority within a batch)

  • Everyone gets the same clearing price (uniform price)

  • Oversubscribed sides are filled pro-rata (fair partial fills)

  • Makers have time to cancel stale quotes before the next batch

This is the same mechanism used by traditional stock exchanges for opening/closing auctions, adapted for on-chain prediction markets.

Last updated