There’s low latency, as any business model would appreciate, and then there’s ultra-low latency as in the financial markets environment. Microseconds in latency can quickly make the difference in millions of dollars lost or won, especially where high-frequency trading techniques are engaged. In other words, we’re talking 100 times faster than it takes for a human to blink..

Exchanges continue to upgrade their systems and slash the time it takes to accept and handle orders. Traders, fund managers and other liquidity houses have taken aggressive steps by colocating computer systems in close proximity to the exchanges to further ring-out latency issues.

Joel Hasbrouck and Gideon Saar examined the issue of microsecond latency in a paper entitled, “Low-Latency Trading,” (October 2010), in which they completed a comprehensive study on the influence of low-latency traders in the market environment. They established latency as the sum of three components: The time it takes for information to reach the trader, the time it takes for the trader’s algorithms to analyze the information, and the time it takes for the generated action to reach the exchange and get implemented. However, this metric is typically measured as the processing delay from the entry of the order (vendor’s computer) to the transmission of an acknowledgment (from the vendor’s computer).

A wealth of statistical information was presented, illustrating the interaction of traders and the microsecond environment. And while their activity aggregates to affect characteristics of the financial markets like the short term volatility of stocks, the total price impact of trades, and the depth of the market – details on the methodologies used to support the ultra low-latency strategies were also discussed.

The researchers utilized NASDAQ order-level data in the study, where they established two core types of activities in the microsecond environment: one by traders who respond to market events, and the other by traders that operate according to a schedule (ie., accessing the market every second) in a time series of market activity – based on wall-clock time.

Drivers of Fundamental Change in the Way the Equities Market Operates Today

Automated trading systems utilized by various organizations are designed to access the market on a per second basis of clock-time periods. These periodicities, as they are referred to in the study, demonstrate the impact of different algorithms – usually used by buy-side managers to minimize trading costs. The interaction drives intense episodes of order submissions and cancellations of limit orders, though they start and stop abruptly..

Much low-latency activity is also driven by traders that respond to market events. Electronic market making firms and statistical arbitrage operations in hedge funds for example, impact the market with high frequency episodes.

While these types of activity can be generally perceived as “scary” in some instances, the bigger picture conclusions of Hasbrouck and Saar showed evidence that increased low-latency activity improves market quality in the form of liquidity, smoothing short-term volatility. After all,the new market reality is based in algorithms, interacting with other algorithms. Further, the study also did not support thoughts that conventional measures of liquidity, typically familiar to long-term investors, have worsened in light of the millisecond pace.

The Old-School Analogy

Authors Hasbrouck and Saar referenced the days of lighter technology, and the old floor-based exchanges in an interesting analogy to the speed advantage presented through today’s low-latency methodologies.. The NYSE for example, was once a physically finite space, where would-be traders could buy a membership, and gain direct access to the market. This presence “on the floor” gave traders a timing advantage, as off-floor traders encountered delays in executing orders.

Fast forward to today’s environment – the exchange is simply a computer server. The new floor – is the physically compact rack on which the server sits. In other words, by paying a fee traders rent a slot for their computers next to the exchange‘s server, or colocating as the practice is termed. This design gives the traders a timing advantage as colocation minimizes the time it takes to receive and analyze information from the exchange (via direct data feeds) and issue orders back to the exchange. Meanwhile, other traders off the co-located environment face more traditional network delays.

The Real Economic Cost Of A Delay

Consider the frequent potential of a single, randomly timed market event – such as an announcement by a key market player… Opportunities can be captured by a first-mover that takes a long or short position against other traders lagging in awareness…Microsecond response can represent the difference in millions of dollars over the course of numerous opportunities presented. Of course, ongoing daily imbalances of supply and demand created within the market itself can also present profitable opportunities to traders that are fast enough to respond.

FiberMedia Group’s Financial Services Practice has excelled in providing virtual and physical infrastructure solutions to capital markets customers. Our Secaucus data center is strategically located and provides micro-second delivery to the U.S. capital markets. This combined with our VMWare based private cloud platform for the financial community provide numerous options for the buy and sell side community.

Visit www.fibermedia.net for more information on our financial industry solutions.