Posted by Dan Hubscher

CFN offers the lowest latency from Chicago, New York, to  BM&F Bovespa

The trading business feels like a fight, now as ever, with the threat of sweeping regulations as the most pressing concern of the moment.  This business even sounds like a war, too – with algorithms names like “Raider” and “Sniper;” and with terms like “dark pools,” and “low latency arms race” drawing focus from regulators and media alike.  But the war-like aspect remains because trading is a highly competitive business.

The imperative to increase market share will remain a top priority this year along with risk management and regulatory compliance, and the technology required to compete is available to everyone.  As Apama has expressed before here, the markets are still driven by those with the flexibility to quickly adapt to new regulations, the insight to understand new market behaviors, and the imagination to conceive a trading strategy that can capitalize on the opportunity.  It’s no wonder that the relatively small number of firms using high frequency trading strategies are responsible for over 70% of US equity trading volume.  These pressures push the rest of the capital markets in the same direction and the trend is unlikely to reverse.

On February 8th 2010, Apama announced that Banco Fator Corretora, a Brazilian bank and brokerage firm, has deployed the Progress® Apama® Algorithmic Trading Accelerator. Apama plays a critical role in Banco Fator’s new electronic trading strategy, enabling it to more effectively develop high frequency, proprietary trading tactics, achieve rapid customization, and perform low latency execution of trades on behalf of its buy-side clients.  Banco Fator is also working with its clients to design customized algorithmic trading strategies that provide them significant competitive advantage, and the bank explicitly emphasized the importance of providing its clients with a fast method to enter the high frequency trading business.

Why so much emphasis on high frequency trading (HFT)?  Reasons will differ among traders and regions, but a short primer on HFT and some ideas are here.  The Brazilian market has expressed a strong opinion on the matter, with over 15 customers deploying Apama internally to automate execution and/or alpha-seeking strategies in the past year-and-a-half, with many further rolling out the platform to downstream clients.

So, do you have an opinion on HFT as well?  What characteristics should a platform for HFT have to enable you to be more competitive?  Let us know – leave a comment, or take our poll.

These days trading floors may have disappeared, but the prize still goes to the swift and nimble in the market. And as electronic trading has taken hold, firms with the fastest network and performance across all of the links in the trading chain will be at the head of the queue, able to react speedily to changing market conditions. Once measured in seconds, latency is now calculated in microseconds, and as markets continue to be volatile, speed of thought and execution remain paramount for many firms., Vijay Kedia Flextrade Systems – Read More

CFN Services has announced reduced latency between Toronto Financial District and  NY/NJ Financial District. CFN Services, continues their commitment to continually evaluate and improve network latency to all major Global Exchanges. The ability that CFN has to quickly react to the market and the needs of their users provides them a differentiation in the ultra low latency space. CFN  offers turnkey Global Exchange Ultra Low Latency Connectivity, and also provides custom network design and optimization for their users. CFN brings a long history of experience and expertise in the fiber network space and that value is passed along to their customers to help provide them the edge they need to increase their execution success in high frequency trading.

Highlights of the Tabb Group – 2010 Market Structure Panel. Interest in high-frequency trading is at an all-time high, but profit-taking from high-frequency trading strategies focused on low latency is getting tougher.
“The window of opportunity to get into high-frequency trading is almost closed,” said Mark Casey, president of CFN Services, a network provider. He defined high-frequency trading as strategies whose underpinning is low-latency order placement and execution.
“If you’re competing primarily on latency, it’s very, very, very, very difficult,” added Nigel Faulkner, chief technology officer for the equities technology group at Goldman Sachs. More

Zero latency is the optimal trading speed that all firms hope to be able to achieve at some point in their life. But until that can become a reality instead of a dream, there is a race to keep reducing latency from all the aspects of the trading process. This is from the server to the messaging to the sending and receiving data, and each piece must be optimized to reach its zero latency goals. For example there is a data center in Weehawken, NJ houses five major exchanges including: the New York Stock Exchange, Philadelphia Stock Exchange, and BATS Trading. This idea is a way to ensure there is almost zero latency built into the trade to those exchanges but not really the complete puzzle. Even though the data centers that host the exchange as a simple cross connect, is rarely the only exchange a firm is trading on at one time, so how do we close the cap? The solution is simple. All one needs is a carrier who can transport the data when latency and performance are important but not that critical at the moment. When the latency however reaches a level that is critical to your company’s success and for your strategic advantage, you made need to relook at your approach and consider a more agnostic approach. This allows you to work with each carrier and see which one is able to provide you their best solutions between locations. This however may not be the best solution if you have multiple desires and needs for this problem. In the race to Alpha, the combination of Proximity Hosting paired with Low Latency Transport is the Key to your companies success.

If you would like to read more on this topic please visit our CFN Services website.

Performance targets in financial trading networks are more stringent than in any other type of network deployed today. Message latency and data loss result in missed trading opportunities and have a direct impact on revenue. The standard of performance needed for success is determined by competition from other trading organizations, and this has led to an ‘arms race’ in which latency targets have steadily shrunk from tens of milliseconds to milliseconds, and even to microseconds. CorvilNet is a network monitoring and analysis system by Corvil, designed to provide visibility into network performance at the timescales which matter for financial trading. Traditional network monitoring tools which make measurements over periods of seconds and longer, fail to identify conditions which can add hundreds of milliseconds of latency. CorvilNet addresses this deficiency with a set of new

The latency that messages experience in practice can be strongly affected by system interactions and usage patterns

Learn WP – Measurement & Characterization of Latency in Trading Networks

Every Microsecond Counts Today brings modern technology and with it, an increasing number of Internet-based applications that require an end-to-end latencies, requiring the need latencies to be transmitted at a rate of milliseconds or even microseconds. Many applications though still demand that the latency remain stable, otherwise having little to none ‘jitter’.  These new applications range from the popular multimedia services to things like voice-over-IP, multi-player gaming and even video conferencing. Since these applications are growing at such a successful rate, customers are placing just as increasing demands on operators to help provide and manage the networks needed to meet these new stringent demands. Unfortunately many of the tools that are currently available for use are not able to create accurate measures of the latencies in the magnitudes that are being demanded, nor can they even detect or localize any loss of spikes. This leads to a loss at extremely small time scales like tens of a microsecond.

Here is an example, a trading network that connects a stock exchange to the number of data centers where automatic trading applications run. In order to stop any unfair opportunities, network operations personnel must be able to ensure that the latencies between the exchange and each of the data centers are within 100 microseconds of each other.

This problem can be solved, just like many others by using current routers that typically support two distinct accounting mechanisms: SNMP and NetFlow. Each router cannot do the work all on its own, it needs the other to help pick up what it cannot. SMNP can only provide cumulative counters, which can be very useful when estimating load, however it cannot provide the latency estimates.  On the other hand NetFlow can give samples and timestamps to help with all the receiving packets and also allows the company to get a calculating of the latency requirements from multiple routers.

It used to be that latency was defined as the time it takes for a signal to travel from one point to another in telecom networks and was of much concern to only people using specific applications such as international voice calls transmitted via satellite that relied on latency. These calls are sensitive to  any high levels of latency, in which  they had negative impacts on the voice quality of the call. Recently latency has become a big interest to a broader audience then before. Businesses have increased their reliance on telecom networks and have been working to improve processing speeds for computer networks and the Local Area Networks (LANs)/Wide Area Networks (WANs) that interconnect them. This has lead to a combination that allows one to make the speed at which a signal travels in a telecom network more noticeable and critical.

Telecom networks and the LANs/WANs connected to them are extremely complex. This means that one should expect performance variances between the different networks and other implementations that may come along with them.    The difference is largely based upon the specifics of the network one is working with. The needs to be an understanding however of some of the drivers of the latency, that can be used to paint a very useful picture to help evaluate the network’s performance.

Once you start processing the information, the processing itself takes place in every part of the network, from simple conversions of characters on the screen to the data bits being transmitted over the vast telecom network. If there are any processing delays, they are just added by the transport network and can vary by the type of connectivity service that was purchased.

If you want to read more on the topic visit CFNServices case study.

Fiber‐optic communication is defined as the method of transmitting information from one place to another by sending pulses of light through an optical fiber. The definition of zero network latency is a network transporting their information at the speed of light. This being said, the further the light has to travel, the higher the latency. Fiber networks normally do not follow a direct line between any who two locations. Instead they follow geographic contours that relate to roads or railroad tracks to name a few.  It is important to understand the exact route that your circuit will follow so that you can eliminate any wasted distance.

In the environment, there are two main options for how you can extend the transmission of light through the fiber; optical amplifiers and regenerators.   Amplifiers and Regenerators are placed along the fiber optic route to help ensure that the signal is able to travel the full distance, while keeping its’ strength and information. There is only one main difference between the two and that is that an amplifier adds light to the existing wavelength to help increase the strength of the signal without having to add more latency, whereas a regenerator will add a substantial amount of latency to add to the strength of the wavelength. Both of these require both a lit and dark fiber solution. A dark fiber is a term that is used to describe fiber optic strands that are not being used or has no equipment being used on the ends of the fiber. Its contrast is called a ‘lit’ fiber, meaning that it is an active fiber optic cable. Unfortunately, due to so few telecom companies still offering dark fiber, the available dark fiber route may be much longer then an alternative lit circuit oath from another provider.

If you want to read more on this topic, take a look at our case study.