Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Low latency drives HK data center services for financials

Sheila Lam, Computerworld HK | March 26, 2013
The ability to trade securities at nanosecond speeds has become a reality.

The ability to trade securities at nanosecond speeds has become a reality.

Cloud computing, hosting services, network connectivity and data analytics technologies have matured, and low latency trading is picking up in Asia's financial markets.

The Bank of England estimates that low latency trading accounted for 5-10% of equity volume in 2010 with rapid growth potential. That potential is driving more hosting services players to the market and transforming IT spending patterns among financial institutions in Hong Kong.

How fast is "fast"?

Low latency trading refers to the network connections and computing technologies used by financial institutions to connect to stock exchanges to execute financial transactions, according to Wikipedia.

Latency in trading could be found in the time it takes for: 1) securities trading information to reach the trader, 2) the trader's algorithms to analyze the information, and 3) the generated action to reach the exchange and be implemented. Delay in any of these processes can affect the financial institution's investment strategy and directly affect its revenue.

Low latency trading has moved from executing a transaction within several seconds to milliseconds, to microseconds, and now nanoseconds. Nowadays, a millisecond improvement in network speeds offers competitive advantage for financial institutions.

"The game has become near real-time," said Cyrus Daruwala, managing director at IDC Financial Insights Asia Pacific.

Deadly spikes

The demand for a low latency trading environment has changed the way trading firms and stock exchanges interact.

The major challenge for most stock exchanges, including the Hong Kong Stock Exchange (HKEx), is to provide an environment capable of processing billions of bidding requests without faltering when trading volumes spike.

"When trading volume goes up, real-time analytics and back office applications to process bidding requests create a demand-spike for compute power," said the IDC analyst. "The application can process billions of transactions, but can the infrastructure and network handle that?"

If a stock exchange relies solely on its own IT infrastructure--a rigid datacenter facility and infrastructure designed for a finite number of transactions--the entire system can stall when trading requests hit its limit.

"The system will hang, because it can't handle that volume," said Daruwala. "Trades drop or become corrupted, meaning that when the traders click 'enter', the trade does not register."

To provide a high availability trading environment, stock exchanges are turning towards cloud computing and hosted services. By providing high availability SLAs, cloud providers like Verizon and BT provide supporting infrastructure for stock exchanges to handle spikes.

When trade volume reaches the limit of the exchange's infrastructure, "the cloud providers provide additional compute power to handle the additional workload," said Daruwala. "This can only be provisioned through cloud."

Demand for analytics


1  2  3  Next Page 

Sign up for CIO Asia eNewsletters.