Arbitrage Trader’s Livelihood at Risk
Thousands of day-traders earning a living from arbitrage opportunities in Indian stock markets may become extinct in the next few years,
thanks to the accelerating interest in algorithmic trading.
The rising number of brokerages with algorithm, or computer programme-driven trading, may turn out to be a second turning point in Indian markets after the introduction of electronic trading a decade-and-a-half ago, which shuttered about 20 stock exchanges and many local brokerages, but expanded the trading community and volumes by leaps and bounds.
“Manual arbitrage as a business is finished, the future belongs to traders with superior technology as far as this line of business (arbitrage) is concerned,” says BR Bagri, chairman of the New Delhi-based BLB Securities. The National Stock Exchange, which controls more than three-fourths of the trading volumes, has approved applications of 200 of its members, roughly a fourth, to trade using algorithms, said a senior exchange official who did not want to be identified. But he said all of them may not do so right away.
Algorithmic trading refers to automated trades executed through software programs which do not require humans to place orders. There could be thousands of programs written to buy or sell a security, currency or commodity at a particular level when one or more factors emerge. Those programs are so fast that people who look at various developments and decide trade would be left way behind because a machine has done it in milliseconds.
For example, a program could be to sell the stock futures of a particular company and buy the stock if the futures price is x% higher than the stock price. Also, it could be to compare a set of variables — if rupee is more than 45 to the dollar, and crude oil is less than $60 per barrel — then the software would sell Infosys futures and buy HPCL shares.
Other than investors who buy for a long term and traders who buy and sell on a daily basis to profit from minor movements, there’s a section, called arbitrageurs, which looks to benefit from distortions in prices despite public information.
Profits from arbitrage have slumped in the past decade as investors across the country have access to same stock prices unlike in the past where various cities had different prices at a given time for the same share. Also, the common trading cycles between the National Stock Exchange and the Bombay Stock Exchange shrank arbitrage.
Now, with software programs taking over trading, it may well just disappear.
“The annual returns from jobbing (arbitrage trading) were about 15% till 2007, but the advent of algorithm trading has shrunk the returns considerably… to about 6% last year, which is what you can earn from a liquid mutual fund,” says Suresh Mehta, chairman and managing director of Dhyan Stock Broking, which employs 50 jobbers.
While trading volumes could surge as fat-cat brokerages hire maths wizards from the best of institutions, it could lead to severe disturbances in the market as was the case during the credit crisis. Some experts have said the credit crisis was accentuated by algorithmic trading which triggered millions of trades due to fast-evolving developments. However, some like Goldman Sachs and JPMorgan, with superior programs, have reportedly benefited through trading even during the crisis.
Algorithmic trading can create a class conflict too between haves and have-nots of technology. There could be charges of discrimination if two members of the same exchange are not on an equal footing. Even some US legislators are planning to restrict high-speed trading. Senators Charles Schumer and Ted Kaufman have sought a review of whether high-speed trading gives unfair advantages to firms with the fastest technology.
In India it may not immediately lead to a surge in volumes since Indian markets still don’t have the depth of the western markets and related markets such as commodities, currency are controlled, say brokers.
NSE has also signed up with 60 members for co-location server facility, priced at Rs 22.5 lakh which allows brokers to place their servers in the same place as NSE’s trading engine which will give them quicker access to price feed. The broker with the server next to the trading engine will get a price feed updated every 3-4 milliseconds, while a broker with a server in Kohima will get a price feed updated every 35-40 milliseconds.
“But it is not as easy as most people would like to believe. You just can’t pick one of the software off the shelf and start trading. If there is the slightest of problem, the results can be disastrous,” Mehta says.