×
Home Companies Industry Politics Money Opinion LoungeMultimedia Science Education Sports TechnologyConsumerSpecialsMint on Sunday
×

Indian exchanges spend more to speed up trading

Indian exchanges spend more to speed up trading
Comment E-mail Print Share
First Published: Mon, Dec 14 2009. 12 57 AM IST

Updated: Mon, Dec 14 2009. 12 57 AM IST
India may still be in the early adoption phase as far as algorithmic trading goes, but its stock exchanges are making large investments to facilitate its progress. Some of the investments are directed at reducing latency, which is the time delay in data being transmitted from the exchange server to that of the member broker as well as the time taken for a member’s order to reach the exchange’s system. Low latency is critical for algorithmic trading, and exchanges, brokers and traders worldwide are investing millions of dollars to enhance trading efficiency.
The National Stock Exchange (NSE) has offered co-location facilities to its members, who can set up automated trading systems in the same building as the exchange. The idea is that the time taken for market data going out from the exchange and for order messages to come in from members will be reduced. By how much? With most of its members connected on leased lines, data is already moving at a fraction of a second. Using co-location services can help traders reduce latency by a few milliseconds (one millisecond is one thousandth of a second).
Graphics: Sandeep Bhatnagar / Mint
In advanced markets, every millisecond counts for algorithmic traders— especially in a market such as the US equities, where two-thirds of all volumes come from high-frequency trading firms. In the Indian context, high-frequency trading is at a nascent stage. Many algorithms used in the Indian markets are to facilitate better order execution. Here, shaving a few milliseconds off the time taken to send an order may not be a high priority. Still, there are categories of traders who are involved in statistical arbitrage and the like, who would want to reduce latency.
Besides, algorithmic trading is slated to grow manifold in the Indian markets, and it’s good that Indian exchanges are making such investments. NSE sent out a circular on 25 November saying members could avail of a faster broadcast of market data, up from 100 messages per second to 200 messages per second. A day later, the Bombay Stock Exchange sent out a circular saying that it has implemented FCAST, which enables a faster broadcast of market data compared with its predecessor DCAST. Besides, it has recommended that all its members should upgrade to 2 Mbps lines and this transition should be completed in the first half of 2010.
NSE has since gone a step ahead and has said in a circular sent earlier this month that it would make tick-by-tick data available to its members. In essence, data on every order entry, modification, cancellation and trades will be made available to members on a real-time basis. Currently, only the bids and offers at the top five prices are disseminated to members. Those who avail of this facility will be able to see the entire order book for a security, except of course, full details of orders where only a specified “disclosed quantity” is revealed.
Again, only a small sub-set of the Indian market participants may benefit from such large volumes of data and decide to invest in such infrastructure, which require dedicated leased lines for each market segment. But the good thing is that the service is available for whoever is ready. Of course, much of this is also happening because of the increase in competition among exchanges.
Write to us at marktomarket@livemint.com
Comment E-mail Print Share
First Published: Mon, Dec 14 2009. 12 57 AM IST