Kim Man Li has been developing algorithmic trading products, or ‘algos’, for nearly a decade. As head of electronic trading sales in Asia for Goldman Sachs, he has seen the debut of algos in India and expects more Indian investors to adopt it as an efficient and anonymous trading technique. He tells Ashish Rukhaiyar that India will soon see more complex high-frequency trading strategies. Edited excerpts:
It has been some time since algos made their debut on Indian bourses. How are market participants taking to it?
The popularity of algorithms has grown over the past few years, the main driver being increased education by brokers, with the buy side realising the advantages of electronic trading. The number of orders traded through algorithms has grown exponentially. As a result, buy-side traders are looking towards more advanced strategies. Basic volume-weighted average price (VWAP) and time-weighted average price (TWAP) algorithms have their place, but more advanced users are looking at variations that can provide them an added advantage.
With growing awareness, buy-side traders are also looking to dissect the actual logic of algorithms to understand the subtle differences between various broker offerings. Overall, India continues to be a high-growth and a high-focus market for Goldman Sachs electronic trading.
Are algos used in India the plain vanilla ones, or have complex strategies also made their debut?
The majority of such orders are still the more traditional, or plain vanilla, algorithms. That being said, clients have started exploring use of ‘advanced’ features and complex strategies within these algorithms. Onshore broker proprietary trading desks are utilising more of the advanced “statistical arbitrage” strategies. And, with the advent of co-location in India, we will see more complex strategies, with a bias towards high-frequency trading.
Algos, at times, have been blamed for accelerating a fall in the broader markets, as they are able to capture the minutes of inefficiencies in pricing and valuations. Do you agree?
Algorithms can be tuned to operate within specified price and volume limits. They should be programmed to work within the risk management framework set down by exchanges and regulators.
Given the additional care taken by exchanges and regulators in approving algorithms, we feel an accentuated move in prices caused by algorithms should not be a common phenomenon. Additionally, Goldman Sachs sets high standards for its algorithms, which are tested vigorously before being released to clients for use.
A host of brokerages and fund houses do not have resources to develop their own proprietary algo. Does this provide global entities, including Goldman Sachs, an opportunity to market their software?
Goldman Sachs has been a market leader in this space and assigns a large amount of resources to make sure it is always on the forefront of technology and service. Our experience and incumbent technology situates us in good stead whenever we enter any new market. However, a great deal of work is still required to make technology and logic work effectively in every kind of market and to make sure that they strictly adhere and confirm to local regulatory requirements.
Fund houses in India usually do not develop own algorithmic trading engines, as it is not part of their core business. They will always look to leverage broker-developed systems. Domestic brokerage houses are now seeing the advantage of electronic trading and are moving towards developing their own algorithmic offerings.
Also Read
Have you seen any increase (or decrease) in the quantum of trades routed through algos in the recent past? Some market players say this number is abysmally low at four-five per cent of the total daily trading turnover.
The number, as a percentage of daily trading turnover, might be low. But, one has to look at these numbers over time. In our experiences with other markets, we have seen algorithmic trading grow rapidly. We expect a similar trend in India. According to our estimates, DMA/algorithmic trading will make up 15-20 per cent institutional market flows.
What is the growth potential of programme trading in India? Will more brokerages and mutual fund houses start using algos?
Absolutely, the growth in use of algorithms will increase over time. Programme trading will slowly move towards newer complex formats, with clients looking to trade waves of their baskets through different algorithms. Overall, the buy side will start viewing algorithmic trading as an efficient and anonymous way to trade baskets into the market, and volumes will only grow from here.