Reducing Latency in Financial Networks
The importance of minimizing latency in data center networks has been discussed intensely in the last year, mostly in respect to trading in financial markets. This became especially apparent during Facebook’s IPO in May when NASDAQ experienced latency – the time between the initiation of a transaction and its execution. This left several investors wondering if their transactions had been executed. Now, NASDAQ OMX Group is looking into setting up a $40 million fund to compensate those who lost money.
Beginning with NASDAQ in 1971, electronic trading has grown rapidly in the last 41 years. In the U.S. market alone there are more than 12 regulated security exchanges, approximately 40 alternate trading systems (ATSes), 9 equity options markets, and 13 futures exchanges. The increasing number and types of devices involved in a transaction makes latency a fact of life. Data center, metro, and wide area networks all generate delay due to the time required to send data over a distance, and through networking devices such as routers and switches. Additional latency is caused by the processing of requests by computers and their associated storage media.
While key to profitability for financial networks, service and network latency is becoming a critical factor in the design and management in data centers of all types. The meaning of latency differs depending on the area of application. In electronic networks, it refers to the time between two events. A related term, “jitter,” is the amount of variation in latency. An exchange may have an average latency of 100 milliseconds, but if its jitter is 500 milliseconds or more, then it will be viewed as less dependable than systems with less jitter. Sensitivity to latency and jitter varies by the type of data being transmitted.
Efforts in Reducing Latency
Today’s market exchanges are striving to reduce latency, not only for key transactions, but for all exchange processing. The goal is to reduce the exchange order response times. If there are significant deviations in latency – the time that it takes for a transaction to receive, process, and acknowledge an order as measured from the customer side of the exchange firewall/ gateway within the transaction’s data center – exchange users must base their trading calculations on larger than the average value.
Processing exchange trades are subject to computational delays. This can be anything from market opening and closing, breaking news, and algorithmic and high-frequency trading, and causes the amount of market data and trading to increase dramatically. Not only do these situations increase computational delay, but they also increase network traffic that can result in additional delays. Such events results in jitter, which wreaks havoc on mathematical trading algorithms dependent on specific latencies for specific markets. Improving these delays are the province the trading application developers, along with their software suppliers.
One way to avoid low latency is using algorithmic trading. It works by breaking larger trades into smaller ones and distributed into multiple exchanges, so it does not affect the price at any one exchange. This means less price slippage from when the buy/sell decision is made until the trade actually happens. This type of trading is being adopted in a growing percentage of trades in equity markets.