Market data volumes are on the rise, and exchanges are challenged to minimize latency, reduce the cost of consumption and generate revenue.

High frequency trading and the proliferation of trading venues in the U.S. and Europe are driving market data volumes ever higher. As volumes increase, firms upgrade their technology so they can handle the flow, which enables them to produce even more data. This scenario has had an impact on market participants’ ability to  onsume data quickly and on exchanges’ ability to sell market data products. To this end, exchanges are allenged to put in place an infrastructure that can keep up with rising data volumes and a market data strategy that ensures a revenue stream.

Many new venues use market data as a loss leader while they build market share. So, if exchanges are going to charge for market data, they need to maintain a high quality offering. For starters, exchanges need to recognize that their customers must cope with data from many venues. An exchange’s technology investment will be multiplied several times over for their downstream users. Exchanges need to understand how their customers access data and need to work closely with the customers to ensure they can meet their needs.
 

Randall Hopkins, Senior Vice President of Global Data Products at NASDAQ OMX, explains that customers pay a direct cost for market data. Moreover, they need to procure bandwidth to bring the data in house as well as acquire computer systems, feed handlers and databases to process the data.
 

“The all-in cost varies depending on the product set the customer chooses and the technology configuration,” he says. “But the market data itself typically represents 25-33% of the total cost.”
 

It is important for exchanges to develop strategies and  implement technologies to minimize latency for data consumers and to decrease the total cost of data consumption. Not everyone requires the lowest latency data. Some market participants do not mind data being a millisecond or two slower, but they want to pay less for it. “Everybody talks about latency, but volume is the elephant in the room,” says Stewart Orrell, Senior Manager in Financial Services at Equinix, a provider of global data center services.
 

Obtaining data feeds over a cross connect in a data center is typically more cost effective than running a line to another location. Further, not all data sources are in the same place, and connectivity requirements vary between customers. Ultimately, the breadth and depth of connectivity choices will determine the way firms procure bandwidth.
 

Some vendors want to distribute market data over the Internet or into clouds. They will want to be close to the access points for the different distribution mechanisms, such as an Internet-peering platform or a cloud distribution node.
 

“In each of the regions several market data hubs will develop similar to the Internet model where customers have access to diverse networks and various sources of data,” says Orrell. “They also will have space to grow their processing plants for the market data in a cost-effective way.”
 

Meanwhile, the industry will continue to develop ways to compress and conflate data so it can be distributed through different mediums. FAST and ZLIB are two compression technologies already in use. But exchanges still have to offer uncompressed feeds because some business models are predicated on receiving information in uncompressed forms.

Each region has its own unique set of challenges. Europe is similar to the U.S. in that there is a proliferation of venues, and data volumes are increasing. In the U.S., market centers are located mainly in New York and Chicago, and customers try to locate their servers close to the data sources. But in Europe, market centers and customers are more geographicallydispersed in London, Frankfurt, Paris, Amsterdam and Zurich. Network connectivity between the different cities is important. 
 

In Asia, tr ading volume is increasing, but it is not as high as in the U.S. and Europe. Execution venues have not proliferated; markets tend to be closed, and they do not interact together. That dynamic will likely change in the near future as the entry of alternative venues creates more competition. At that point,  exchanges will have to determine where to centralize market data processing in a geographic area that is widely dispersed. “This situation is going to impact clients more than exchanges, but exchanges will still have to figure out how to get data to their customers in a cost-efficient way,” says Orrell. For example, what is the best way to get data from Tokyo for customers with data centers in Singapore?”
 

Ultimately, technology is the key enabler which will have to be continually improved to keep pace with increasing market data volumes. Customers will look for innovations that will enable them to manage data effectively and derive maximum value. The burden is on exchanges to meet those needs.