Jogi Narain said that trading desks could monitor and control latency effectively within their own network, but also needed to look to infrastructure outside of their control to gain a full picture of where fluctuations could be occurring.

"Latency invariably equates to the network. Processing done on your own servers, your own network, routers, switches, these are all things you can control," Narain told the TradeTech Europe conference this week.

"But as soon as you step outside the firm, whether it is in a co-location site, a proximity site, or in your own offices, you have various companies that provide you with physical infrastructure - this would be the telcos," he added.

"One thing you have to continuously review every three to six months is the performance on their network. What speeds are they publishing and does that performance match what you are seeing on a daily basis?"

Narain went on to explain that because telco companies are constantly making changes to their infrastructure, whether that be alterations or upgrades, speeds can change month by month. He described it as a "never ending battle" to not only measure performance on an internal trader's network but also that of the telcos.