Machines vs humans; algo trade impacts stock market traders
Concerns are rising among traders like Kumar about Algorithmic trading (or simply Algo trading), computers monitoring and trading in stock markets. The computers can do millions of transactions in matter of seconds with better connectivity and processing speed, compared to an ordinary human trader.business Updated: Aug 17, 2015 19:46 IST
When Bhaskar Kumar (name changed), a stock market trader wanted to buy shares of Bajaj Auto in mid April, he found that no matter how many times he tried, they went to a strange trader who could always place orders for a slightly higher price – often as low as 50-60 paise – than his order.
Finally he gave up the plan. Kumar did not know then that the other trader he lost out was not a human being, but a computer programmed with a trading algorithm.
Concerns are rising among traders like Kumar about Algorithmic trading (or simply Algo trading), computers monitoring and trading in stock markets. The computers can do millions of transactions in matter of seconds with better connectivity and processing speed, compared to an ordinary human trader.
Echoing the concern, the Association of National Exchanges Members of India (ANMI) on last Wednesday wrote a letter to Securities and Exchange Board of India asking it to protect interest of small investors by regulating algo trading. “There is no equity in equity markets. The market puts ‘moneyed’ people with technology like algo trading in an advantages position, compared to ordinary retail investors,” says V Nagappan, who runs Oriental Stocks, a leading stock broking firms in Chennai.
The regulators are concerned too. In June, RBI in its Financial Stability Report (FSR) warned Indian market of High Frequency Trading (HFT) – the high speed form of Algorithmic trading – posing “risks in the form of increased possibilities of error trades and market manipulation”.
RBI attributed certain abnormal activities in the market to Algo trading. In August, the market regulator SEBI was reportedly considering regulations to restrict it.
In Indian stock markets the Algo trade volume in the cash segment was around 40% of total trades in NSE and BSE in March 2015, a substantial increase since 2008 when it was allowed first time here.
In comparison, it accounts for around 70% of equity trading in the US markets.
The mysterious black-box
“High Frequency Algo Trading requires the ability to react to events in the market very fast. It is the biggest contributors of liquidity to the markets,” says Rajib Ranjan Borah, Co-founder and Director of Quant-Insti, one of India’s major Algo trading training institutes.
Algo traders claim it has immense advantages over the conventional trading. It brings a lot of money to the market. It reduces bid-ask spreads—with faster connectivity you are able to find out where the best price is offered, and trade at that price.
They also claim it improves trade volumes and price efficiency, and reduced market volatility. But many still perceive it as mysterious black-box owing to its complexities.
American financial journalist Michael Lewis in his celebrated book “Flash Boys” accused it of rigging markets. “People no longer are responsible for what happens in the market, because computers make all the decisions,” he says in the book.
DMA and Co-location
Algo trading has been there in the US since the 1980s. In India it was allowed only in April 2008 with the introduction of direct market access (DMA) -- allowing of electronic ways of interaction with the stock markets.
It got a boost when ‘co-location’ was allowed in India. Co-location refers to an exchange allowing some traders, usually large traders, to place their computer servers next to that of the servers of the exchange for a price.
This allows faster communication between the two. It means those traders can trade faster than others. This has been criticized by many for putting smaller traders in a disadvantage position. Business Line on August 9 reported that SEBI is considering installing a two-queue system for traders, one for brokers with a co-location advantage and another for those without.
It is also considering a lock-in period for algorithmic trades. “Co-location means those who can pay get the advantage in trading. That is unfair,” says Nagappan.
In India, the regulator SEBI allows algorithmic trading in all segments -- equities, derivative and commodities. When asked about the regulatory preparedness, SEBI replied to HT that it has issued broad guidelines on Algo/HFT trading in 2012 and 2013 and has specified regulatory requirements for exchanges offering Co-location facility, in May, 2015.
The regulations include minimum order level checks, a consolidated audit trail, and framework for penalizing cases of high order-to-trade ratios.
New algorithms are also tested before they go live. SEBI also said it has constituted a Technical Advisory Committee (TAC) to advise it on addressing technological challenges, including co-location, algorithmic trading, and smart order routing.
Algo trade – India Vs west
The major malpractices of algo traders globally have been spoofing and quote stuffing. They place orders with no intention of executing them, creating an illusion of demand to get favorable prices.
“India has the Security Transaction Tax which is on the value of the trade rather than the number of transactions. In the US, for instance, you have to pay the same amount of tax for a $100 transaction and a $10 million transaction. But it is not so in India,” says Manish Jalan, CEO Samssara Capital Technologies LLP, an algorithmic trading company based in Mumbai.
This prevents too much High Frequency Trades happening in a short time that could potentially lead to flash crashes, he says. “We don’t have too many exchanges.
This also ensures there are no unfair advantages to Algo traders through inter-exchange arbitrage,” he adds. Usually the algo-trading companies in the US make profit by finding out those exchanges where the value of a particular share would be lesser by slightest margin from other exchanges.
With the multi- exchange connectivity and speed, they then buy large number of those shares and sell those in slightly higher prices in other exchanges.
The profit is made through increasing the trade volume. “Algo trading is affecting genuine investors,” says Nagappan. “Unlike the computers an ordinary trader would not be able to keep track of the market each minute. And he loses out,” he adds.
“Since conventional traders are finding it difficult to compete with algorithms for the same task there is a lot of hue and cry. Like any other industry, things evolve and the old ways need to catch up by learning new skills. There is always a lot of resistance to change, but eventually the fittest survives,” says Borah.
“Any trading strategy that could manipulate the markets can be executed manually or algorithmically. It is the trading strategy which aims to manipulate, and not the platform,” he adds.