Q&A: Giles Nelson, Strategy Director, Apama

At the time when regulators around the world are blaming algorithmic trading for all the evil in stock markets, Giles Nelson, co-founder of Apama, the most used algo, says it is becoming a technology. In an interview with Palak Shah & Ashish Rukhaiyar, he talks about the growing influence of algorithmic trading in India. Edited excerpts:

How is algorithmic trading evolving? What is its future in India?
It is becoming a future technology. Nearly 70 per cent of equity trades in the US originate from algorithmic trading in one way or other. It is catching up in Europe, Brazil, Australia and Asia. In India, spending on softwares for algorithmic trading is going to be easily worth $100 million in two years. The market share of algo trade will rise from 15 per cent at present to 50 per cent in the next three years.

The number of trades on the NSE (National Stock Exchange) are 10 times that of the London Stock Exchange. Both NSE and BSE (Bombay Stock Exchange) are offering co-location facility; smart order routing and mobile trading has now been allowed, too. Also, commodity exchanges are catching up on algo trading. It is being put to use in foreign exchange derivatives too.

That’s an enormous reason to do algorithmic trading. Smart order routing will force inefficiency out of the market, reduce price discrepancies between the two main equity exchanges and increase competition. NSE’s process of validating every algo was putting a significant brake on its growth. The process is unsustainable and will be short-lived.

There are concerns that algo trading leads to distortion of price and its discovery, that it encourages excessive speculation.
Criticism has been going on for over a year after the ‘flash crash’ in the US stock market. Nobody has yet understood what caused the flash crash. There were other reasons, including the structure of the market. Algo technology itself is difficult and neutral; it is not right to only blame it for anything that goes wrong in the market. Sometimes, massive control and malicious entries behind it is where difficulties lie.

About 70 per cent of equity trades in the US coming from smaller high-frequency trading firms, using different algo trading models, is a potential issue. There may be circumstances of price distortion. But there is more liquidity in markets which have high algo trading. It reduces spreads between prices, makes it cheaper not only for the algo trader but also the investor.

Algo trading is banned for things that it is not responsible for. There has been controversy about dark pools. First of all, naked access should be banned. All orders that are through electronic means or manually should go through a pre-trading process. It will prevent instances like last year in the UK, where a global oil trader bought seven billion barrels of oil when he was drunk. This impacted the global crude price.

Is there an adequate regulatory framework? Are you offering complex algo patterns in India?
Mandatory pre-trade risk and market surveillance should be there within the regulators and the exchanges. Regulators have the rear-view mirror approach when it comes to understanding market softwares. They do not have the capability to know what is happening on a realtime basis. The technology that SEC, the US stock market regulator, was using was two decades old. Regulators have to catch up here. All parties in the trading cycle should take more responsibility to ensure appropriate risk control and surveillance.

We are offering vanilla algorithms and smart order routing in India through Apama. It is an algorithmic trading software used for multi-exchange markets such as US and Europe. The software is a transparent box and can be looked into. It offers not only a platform but also tools to change algos with a changing market. We have Reliance, HDFC as our clients, but we have started marketing it only this year.

How big is the third-party algo vendor business? What are the profit margins?
It’s difficult to answer. Global business is probably a couple of billion dollars a year, when you look at all the platforms, connectivity, the hardware, and so on. We are a $500-million organisation, with operations in 180 countries. Apama, which is used in risk management and surveillance, apart from trading, earns us a number of million dollars per year. Quite a significant number.