Ithaca Matching Engine

The Ithaca Matching Engine (IME) is a multi-product cross-orderbook matching engine.

IME has four key elements:

1. Replication – IME decomposes financial contracts into packages of ‘atomic’ instruments using put-call parity relationships. Therefore, orders are matched at the atomic instrument level.

2. Conditional orders – ‘synthetic’ orders are formed by replication principles while the matching engine algorithm ensures simultaneous consistent execution

3. Mixed Integer linear Programming (MIP) optimization – MIP allows searching for clearing prices and associated sets of consistent orders satisfying them that maximize executed volume and satisfy best execution requirements.

4. Portfolio Dominance

Buy and Sell orders, including orders for pre-packaged and structured products and conditional orders are matched based on the principles of maximum matched volume and minimum surplus across all orderbooks covered.

If several possible sets of matching prices exist that would result in the same amount of total matched volume and the same volume of surplus across all orderbooks, the set of match prices within the group of possible sets that results in the smallest total price deviation to the previous set of reference prices across all orderbooks will be the determined set of match prices.

Once the set of match prices for all contracts is being determined, all orders that are being allocated a fill at the determined prices are filled accordingly and are either removed from the orderbook or their live outstanding size updated.

If there are surpluses of match-able quantity on one side of an orderbook, all orders on that side are partially filled according to their original order size relative to the overall executable quantity.

Fully executed orders are removed from the Orderbook at the end of the Matching Phase. Partially filled orders are automatically updated in the Orderbook to appropriately reflect the new and remaining size and state of the order.

Across all products at ‘ an auction within an auction ‘ the concept of portfolio dominance allows for the protocol to accommodate orders that would not have found direct matches within the orderbook, by acting as the implied counterparty. As long as at expiry the sum of all potential liability of the protocol is less than the sum of its asset side in all scenarios, the protocol creates more fills for users.