The Evolution of Algorithmic Trading in India

Author: Raghu Kumar, Co-founder, RAIN Technologies.

I have been trading exclusively with proprietary, automated trading strategies since 2006. My introduction to algorithmic trading was in 2005 during my sophomore year at the University of Illinois, Urbana-Champaign (UIUC), whereas an uninspired 19-year-old, I was seriously starting to question the merits of my decision to pursue a career in Actuarial Science. I had no interest in becoming an actuary but didn’t know what else I should be doing. My 2.5 GPA at the time did not help matters, either.

But the introduction to algorithmic trading was by happenstance since I never had an interest in trading or investing. In fact, I thought the stock markets were just a sophisticated version of casinos, except that the traders that I did know never had the guts to admit that they were gamblers. I respected traditional gamblers, in that sense. At least a gambler knows that he’s a gambler.

But all of that changed when I stumbled upon an article on the legendary Turtle Traders. The story of Richard Dennis was something straight out of a Hollywood movie, and it convinced me that anybody could be taught rules-based trading principles (for a more in-depth analysis of the Turtle Trading System, visit here). Dennis was able to teach his Turtle Trading System to 15 complete novices (his “Turtles”), and the Turtles amassed over $150 million over 4 years in their own right using the rules-based system taught by their respected mentor.

I graduated quickly and pursued a full-time career in algorithmic trading. From 2006–2008, I solely traded the “spot” (also known as Over-The-Counter, or “OTC”) Forex markets. Using algorithms built-in house, I deployed arbitrage based trading models which scaled and generated millions of dollars. In 2008, the housing crisis and external competition began impacting our trading profits, so we shut down the Forex book and started contemplating our next move. Little did I know India would be my calling card.

2008: A Momentous Year for Algorithmic Trading in India

In mid-2008, I stumbled upon a circular issued by SEBI. In it, I found out that India was opening up its exchanges to algorithmic trading with the introduction of Direct Market Access, or DMA for short.

What is DMA? DMA allowed brokers, at the time, to open up their infrastructure to their institutional (non-retail) clients and provided these same clients the ability to place trades powered by algorithms. In other words, for the first time, clients could trade with no human intervention.

This was a big deal! Sure, it wasn’t open to the retail public (yet), but the capacity was there to build rules-based trading systems and deploy them. I began building arbitrage-based models for the Equities, Futures, and Options segments on both the Bombay Stock Stock Exchange (BSE) and National Stock Exchange (NSE).

As can be evidenced from the above graph, the percentage of turnover powered by algorithms has steadily gone up since 2011 on BSE’s Cash (Equities) segment. After making the move to India, we were able to trade our models by essentially working out an arrangement with one of the leading institutional desks in India. In due time, we applied for our own brokerage licenses and begin trading as our own proprietary trading firm.

2010: Smart-Order-Routing is Launched

It is a well-known fact that the Bombay Stock Exchange (BSE) and National Stock Exchange (NSE) compete against each other and vie for the same market participants. This is especially true since the same securities are cross-listed on both exchanges!

In 2010, SEBI approved the launch of Smart Order Routing (SOR), which essentially, overnight, made it possible for investors to place their trades without having to worry about which exchange was providing the better price. This provided a welcome positive boost in the stock markets. Stock market participants were able to place trades confidently, knowing that the best price would be executed upon. From a liquidity standpoint, transaction volumes increased across both exchanges.

Enter: TBT and Co-location Services

In January 2010, the NSE started offering Tick-By-Tick (TBT) data to its members (for a hefty fee). Around the same time, it also started offering co-location services. Co-location provides members the ability to place their servers in the exchange’s premises in return for a fee. Interestingly, the NSE is now engulfed in a whistleblower scam that centers around the allegation that certain members were given preferential treatment over others, which enabled these same members to amass massive profits. The odd aspect to the story is that, as a regulator, SEBI should have released a discussion paper/circular to approve such services from being offered; however, SEBI stayed mum on the topic for two years.

In 2012, after hearing complaints from the BSE about the NSE’s offering of TBT/Co-location services without a green light from SEBI, SEBI issued a circular laying out the framework for fair practices to be followed.

In 2013, the BSE also started offering co-location services to its members at just one-fifth of the price that the NSE charged its members. It also started offering Tick-By-Tick data.

Opening it Up to Retail

Both the NSE and BSE opened up algorithmic trading to the retail population, but the guidelines have never been crystal clear. Technically, both exchanges require that each algorithm must be approved through a rigorous process. However, with the advent of API’s being offered by brokers such as Zerodha’s Kite and Upstox’s Developer API, the retail population can technically circumvent the approval process. In other words, algorithmic trading is truly open to the retail population in India.

This is a welcome change. With the steady rise of algorithmic trading in India, we can expect the sophisticated retail trader to make full use of these new-age API’s and build out custom, powerful, proprietary rules-based trading strategies. In due time, we can expect the percentage of turnover powered by algorithms, in India, to rival the numbers of the United States, where over 90% of all trades are powered by algos.

Richard Dennis once said “I always say you could publish rules in a newspaper and no one would follow them. The key is consistency and discipline.” There is no better way to stay consistent and disciplined with your trading than with a rules-based approach powered by algorithms. At RAIN, since we solely build algorithmic trading strategies for India’s capital markets, we are excited to see what the future holds for the algorithmic trading community in the country.