This article was first published by the Global Association for Risk Professionals on June 19, 2016
Six years since allowing algo trading, India wrestles with questions familiar across financial markets
Since first being allowed in 2008, algorithmic trading has grown to account for 40% to 50% of the turnover on the National Stock Exchange (NSE) of India.
Widely regarded as a disruptive technology, and creator of advantages for some that harness it, algo trading was defined by the Securities Exchange Board of India (SEBI) as “any order that is generated using automated execution logic.”
The late Gangadhar Darbha, who was executive director and head of algorithmic trading strategies at Nomura Securities, said, “Algorithmic trade per se is nothing but a reflection of what happens in our brains. Algo trading is the use of electronic platforms for entering trading orders with an algorithm which executes pre-programmed trading instructions whose variables may include timing, price, or quantity of the order, or in many cases initiating the order without human intervention.”
Institutional investors, insurance companies and mutual funds use algorithmic techniques for portfolio rebalancing and risk control amid large order flows on either side. Anonymity and mathematical logic to break large orders into small pieces can counter adverse price moves and manipulation.
Financial firms apply the cutting-edge technology to product development and innovation in such areas as liquidity-seeking, cross-asset and multiple-exchange trading.
While it is difficult to know the strategy or logic being applied by looking at the trade data, it seems unlikely that the algorithms act on fundamental information about companies or the economy. Algo trading takes into account quantitative information regarding trends, reversion to mean, arbitrage, etc.
It is argued that computer-assisted trading improves liquidity in the markets by breaking down large trades into smaller trades, reduces bid-ask spread and lowers risks of adverse selection and trade related price discovery. Terrence Hendershott, Charles M. Jones and Albert J. Menkveld, in “Does Algorithmic Trading Improve Liquidity?” (Journal of Finance, Vol. LXVI, No. 1, February 2011), argue that algo traders are more likely to be providers of liquidity. They also increase the speed and efficiency of trades and reduce costs of trading.
There are also purported disadvantages.
There have been many instances of flash crashes on exchanges around the world, in which automated trading has been implicated, whether due to erroneous coding, a systems failure or cascading effects of certain trades. Algo is also often blamed for causing large fluctuations or volatility in the markets.
On October 5, 2012, the CNX Nifty Index of top 50 companies traded on India’s NSE fell nearly 16% within seconds (before rebounding), causing panic amid traders and institutional players, as stop losses were triggered. Similarly, the Bombay Stock Exchange (BSE) had to annul all trades on “muhurat day” in 2011 due to extraordinary volumes. (Muhurat day is a special trading session on Indian bourses to mark to beginning of the new financial year on the Hindu calendar. It coincides with the popular festival of Diwali.)
Algo trading’s effect of improving liquidity has been scrutinized. It is alleged that algo trades focus on a few large stocks, resulting in short-term liquidity improvement only for those stocks. If this also means that trading gets concentrated in fewer stocks, then this is not good for an exchange. And if algo trading magnifies panic, it can have an avalanche effect on prices.
Technology is reducing the cost of trading and attracting large volumes. But that requires exchanges to improve the efficiency and capacity of their data servers, matching engines and bandwidth, which in turn increases infrastructure cost and has to be added back to the brokers transaction cost.
Haves and Have Nots?
Does algorithmic trading "discriminate between rich and influential brokers and common investors/retail investors and create inequality” among constituencies on the BSE and NSE, as has been alleged by the Intermediaries and Investor Welfare Association? Algo traders have a technological advantage over the regular market participants, who are not able to afford such high fixed costs.
The Technical Advisory Committee of the SEBI, it was reported in April, noted that a certain broker benefited from loopholes in the systems architecture of NSE, which were not prevented by the exchange. Larger questions regarding the role of the regulator in coming out with policies on co-location and algorithmic trading are being raised as well. The surveillance systems of the exchanges and regulatory actions against manipulative activities have not kept pace with the improvements in technology and the complexity of algorithms.
Interestingly, rampant accusations that flash orders favor insiders led some lawmakers in the U.S. to urge in 2009 that the practice be banned by the Securities and Exchange Commission. It remains legal. The Indian counterpart, the SEBI, has set an agenda for itself to come out with a discussion paper on high-frequency trading, or algo trading, in the coming three months, which could lead to regulations of those activities.
Perhaps the right way to look at it is as Darbha once said in an interview: “Algorithmic trading is not just a facility, but an aid. While algorithmic trading gives you freedom to trade, it does not replace fundamental research. It only enhances trading efficiency.”