Bringing more control to the management of trading traffic.

Exchanges today operate in a digital world that links together two basic communication systems: the Local Area Networks that operate inside the firewalls of a trading organization and the Wide Area Network that lies outside their control. The Internet is, of course, the world’s largest Wide Area Network.

In many topological configurations, network engineers use an Internet Gateway Device—which is a hardware and software system—to add functionality to the connection between a Local Area Network and a Wide Area Network.

This technology has great potential in terms of the network architecture of an exchange, because of its ability to provide more control over incoming trading traffic. So for the sake of simplicity, let us call this particular market-focused application an “Exchange Gateway Device” or EGD.
In terms of implementation, the EGD will be integrated into the Local Area Network of an exchange inside the exchange’s security firewall. By design, it will be able to use the exchange’s encryption keys to decrypt incoming traffic from the Wide Area Network. As a result, it will be able to break down the contents of incoming messages and identify the trading symbol, price, type of trade request, and other pertinent information.

Once this step is completed, the EGD will store each trading message in an electronic message queue. After a defined Queue Interval timeframe, these stored trade requests will be transferred to a Release Batch where individual trades will be sorted in an order determined by the Fairness Algorithm. Trades will then be released to the exchange’s trading computers for final fulfillment.

In the traditional approach, trading messages received by an exchange are forwarded to trading computers for fulfillment in the order in which they were received. But this process ensures that traders who successfully exploit latency will gain an unfair advantage.

The EGD, on the other hand, provides the capability to negate the effect of the latency of trades and establish an equitable queue order determined by the controlling Fairness Algorithm—a flexible algorithm that can be configured in a number of ways.

Here’s one example of a practical configuration.

To ensure trading fairness and efficiency, an exchange could configure the Fairness Algorithm to set the Queue Interval at approximately one second. The decision could also be made to establish a sort sequence in which “buys” in the Release Batch would be processed first followed by the “sells.”

In each buy or sell sub-group, trades would be arranged in the appropriate “limit order” price sequence and processed in descending order according to price.

This approach would ensure that buys with the highest price get matched with the highest priced sells. Market orders would be ranked at the top of each group in the sort sequence.

Practical experience and an analysis of trading activity will provide the insights needed to fine-tune both the Fairness Algorithm and the Queue Interval to achieve the optimal level of performance and fairness for all of the stakeholders involved: exchanges, investors and regulatory organizations.
Naturally, these controlling algorithms can be formulated in a variety of ways. But the goal would be to establish a “Fairness Algorithm” that would take latency out of the factors in trading activity and empower trading from all locations however distant from the exchange while maintaining the optimal level of trading efficiency we would in fact expand the market opportunities in this shrinking worldwide marketplace. In addition, the exact nature of the algorithm and any changes to the algorithm should be made public at all times to ensure transparency.

Because of its ability to negate latency as a significant factor, the EGD’s Fairness Algorithm will also serve as an effective countermeasure against snooping, the intentional corruption of packets, and “front-running” Since market participants wouldn’t be able to take advantage of reduced latency, they will have no incentive to engage in these unethical practices.

Of course, the development of the Fairness Algorithm would initially be a challenging task requiring the active involvement of all market stakeholders. It would also undoubtedly involve some trial and error.

But in a market environment that is spinning out of control, it is important to remember this point: It is technically possible to achieve the goal of global latency equalization with an EGD and an effective Fairness Algorithm.

It’s really not that difficult from a technical standpoint.

An effective tool for regulatory oversight.

Another key advantage of the EGD is that it can be designed to automatically compile a comprehensive audit trail of all trading activity. This audit trail can then be analyzed by regulators to ensure equitable trading, maintain transparency, and identify any anomalous activities.

In fact, regulators could develop automated systems to scrutinize trading on an continual basis and make sure that the published Fairness Algorithm is, in fact, the one in use and is achieving its stated goals.

Thanks to these capabilities, the EGD will increase the administrative control of self-regulating organizations, improve the ability of regulators to do their job, and make trading more transparent—all of which will help restore confidence in our financial markets and the exchanges that support them.