Cookie Notice

Products

Exchange Simulator (SIM)

Exchange Simulator (SIM) is a product aimed at back-testing of intraday trading strategies. SIM can act as both an exchange emulator and a historical data re-player thus providing an environment where stat-arb and other systems-based traders can run their intraday trading algorithms and strategies. A set of statistical and expert models incorporated into SIM allows it to reproduce a realistic behavior and reaction of the market. They include:

  • Temporary market impact via a closed-loop control model which dissolves user intrusions in form of order submissions so that any deviation from the historical orderbook is compensated via a sequence of automatically generated artificial order submissions and cancellations. Thus the model can replicate a temporary market impact with no permanent impact caused by user intrusions.
  • Preservation of reasonable spread value, realized by introducing a possibility to change submission price of historical orders. The limit price for artificial order submissions is changed accordingly to elasticity model which takes into account a distance between historical mid-price and historical submission price.
  • The real-life effect called diming is accounted for in the submission price probability distribution generation. It occurs when a large volume appears at a small distance from the mid price. The large volume causes traders who wish that their orders execution is not delayed significantly, to put the orders at the lower distances.

 

SIM can also be used to cover functional, performance and capacity testing for complex trading applications by:

  • emulating various trading application inputs including exchange connectivity, market data, and order-flow
  • acting as a data-driven re-player of detailed, historical trading activity on major stock exchanges
  • simulating market reactions, generating new market activity and counter-flow in response to non-historic, user-generated orders.

 

Competing Products

  • Leading stock exchanges and few commercial vendors offer facilities or tools that allow users to test trading algorithms, strategies and systems but under restricted conditions that offer neither a realistic simulation of a liquid market nor the ability to replay and interact with historical trading activity.
  • Competing products “emulate” the rules of an exchange: they are, in essence, an emulator box containing the mechanical rules of the exchange + an empty order book. Users can test whether their orders will be executed by that exchange, in accordance with its rules.
  • SIM goes further and simulates a stock’s trading behavior on an exchange: in addition to an emulator box, it comprises historical data flow, user-generated flows and model generated flows. Users can test both whether their orders will be executed and how they will be executed.

 

Unique Features

  • In “playback mode” SIM replicates precisely the historical record of activity on an exchange over a period of time and this can be repeated any number of times, at normal and accelerated speed
  • In addition, SIM in “simulation mode” produces realistic, simulated counter-flows, in response to user-generated submissions, cancellations and trades
  • Strategies can be run on SIM either “deterministically” (the same strategy always produces the same results, which is useful for replaying and identifying problems) or stochastically (the same strategy, run multiple times produces a distribution of results)
  • Models are tuned at the level of individual stock behavior
  • Users can replay and interact with historical order books, not just market data snapshots
  • SIM can use historical exchange transaction files
  • It can use generic market data (trades and snapshots) to recreate flows and historical order books using a proprietary orderbook reconstruction technique
  • SIM offers a variety of interfaces for convenient plug-in to algo-trading environments, including FIX for integration into existing algo-trading frameworks or COM for linking to Rapid application development environments such as Excel and VB. It also provides a number of ways of publishing market data
  • Users can trade multiple stocks or baskets of stocks, in real time or ASAP mode

 

Exchanges Supported

  • SIM supports LSE (SETS), EuroNext, Xetra, OMX, Milan, and Madrid exchanges, and allows end users to easily configure it to support arbitrary exchanges. User can also specify trading phases including accumulation, opening auction, intraday auctions, closing auction, continuous trading, trading at last, etc.; supported order types and validity conditions; and any differences in matching rules in auctions.

 

Exchange Simulator Model Methodology is available for download below:

Download PDF file

back