From the April 01, 2012 issue of Futures Magazine • Subscribe!

Software Review: OptionsCity Freeway

Addressing Time-to-Market for the Professional Algorithmic Trader

Freeway adds the notion strategy jobs to modularize development. Jobs are then uploaded to the second major component of the Freeway product — the Freeway server — for test and execution. By supporting multiple asynchronous jobs that the Freeway server executes concurrently, the developer is able to assemble large algorithms a component at a time. The Freeway product comes with a number of pre-written jobs that the trader-developer can use to boot-strap development (plenty of help getting the initial “Hello, World” program up and running). A click on the front-end upload key and a strategy is sent to the Freeway server. The Freeway front-end is used to manage the server and it provides status information and graphical displays of algorithm execution.

The Freeway server, a Linux-based machine co-located within the exchange, handles all the details needed to interface to the exchange and clearinghouse. All the major U.S. exchanges are supported, with plans for a number of foreign exchanges in 2012. Here, the detailed plumbing needed to execute against the exchange is implemented, eliminating time-consuming development, while providing a high-performance run-time environment. The server manages the execution and data flow between strategy jobs, including proprietary queuing mechanisms to handle data bursts during high volume and volatile market activity. The current Freeway server supports round-trip transaction times with the exchange of less than several hundred microseconds.

The OptionsCity team spent more than five years tuning and providing the necessary support for seamless integration with the exchange. Debugging is accomplished with an OptionsCity co-resident exchange emulator, or via a separate OptionsCity server that simulates the exchange. Freeway clients run their code on a dedicated server, and multiple servers can be configured, based on the number of exchanges traded and the scope of the trading. The user either can subscribe to a third-party data service for historical data or capture real-time data for testing. While Freeway server configuration will have a number of variables affecting pricing, a $10,000-per-month figure was quoted by OptionsCity as a representative price point.

This reviewer spent an afternoon testing a CME Market Depth bid-ask queue breakout strategy with the help of an OptionsCity developer. It was impressive to see how readily the pre-canned Freeway APIs and jobs could be used as a starting point and the strategy brought up quickly against the exchange emulator.

It’s all about time-to-market. Using industry-standard Java programming, the Freeway strategy APIs and front-end, and the Freeway server direct exchange interface, the OptionsCity product has solved many of the programming and infrastructure development costs of professional algorithmic trading. Because algorithms run on a co-located Freeway server, development can occur anywhere on the Internet. The professional programming enterprise can look at the Freeway product as a means of cutting development and time-to-market costs. The OptionsCity Freeway product has moved the company from an options focus to cover the breadth of stock, futures and forex markets, as well. If faced with the challenge of entering the business of high-frequency algorithmic trading, this reviewer definitely would examine this product to reduce time-to-market costs. 

For 20 years Michael Gutmann was a software engineer and manager at Intel Corporation. He recently published the second edition of “The Very Latest E-Mini Trading: Using Market Anticipation to Trade Electronic Futures, 2nd Edition.” Mike can be reached via www.anticipationtrading.com.

Comments
comments powered by Disqus