Designed for systems-based traders and trading technologists, Allied Testingâs Exchange Simulator addresses the growing disparity between complex trading algorithms and the methods used to test them. It brings the added dimension of realism to the testing of automated trading strategies in the European equity markets.
The Simulator can be used in a number of different modes. In emulation mode, it is designed both for testing connectivity and as a configurable, always-on UAT exchange for generating fills. The suite of tools also provides the unique capability to trade against historical liquidity and receive realistic fills, based on statistical models of stock-specific behaviour. Historical flow scenarios can be generated, either from historical L3 exchange feeds or - using proprietary technology - the orderbook can be reconstructed using historical L2 depth and trade data. It does this by inferring order flows from changes in price and volume, using models of how an individual stock typically trades.
When user-orders start trading against historical liquidity, the Simulator takes over the task of keeping the orderbook in shape by adjusting the prices of incoming historical orders and generating completely new simulated orders to replicate typical reactions to a userâs own trading activity on the book.
âThe simulated behaviours depend on the typical dynamics and microstructure properties of each instrument; so you will see typical queuing behaviour for liquid stocks with large spreads and, where there are spaces in the orderbook, simulated orders will compete for the spread by jumping ahead of large orders in the queue. Apart from being based on an empirically justified, continuous, event-driven approach, this Simulator provides much greater realism in fill generation than can be provided by estimating fills based on simple âif-thenâ rules,â explains Jon Davidson, VP of Product Development.
The Simulator provides a market gateway including a .NET API, as well as industry standard protocols including FIX. Designed specifically to integrate seamlessly into an algo-trading environment, the Simulator can replace outgoing connections to the market and provide price feed reflecting the current state of the orderbook, using RMDS or custom price feed. It is also highly configurable, and can easily be customised for new markets and even alternate asset classes, using the semantically rich config library.
The flexibility to run the Simulator as a replacement for UAT markets saves valuable testing time, often wasted while UAT markets are unavailable or closed. Furthermore, the ability to test strategies against historical flows in a variety of market conditions, either step-by-step, in debugging mode or at high speed, in strategy-stressing mode, gives quant-traders, strategist and technologists the ability to gain a deeper level of understanding of and confidence in their strategies before releasing them to market.