Execution of trades on stock exchanges based on predefined criteria and without any human intervention using computer programs and software is called algorithmic trading or algo trading. While being a subset of algorithmic trading, high-frequency trading involves buying and selling thousands of shares in fractions of seconds.
In the US and other developed markets, High-Frequency Trading and Algorithmic trading accounts estimated 70% of equities market share. This form of high-speed trading rose 12 per cent on the Bombay Stock Exchange, to account for almost 30 per cent of total trades. Its share is higher on the National Stock Exchange, with nearly 46 per cent of trades happening on the platform, according to latest reports.
The Year 2008: Beginning Of Algorithmic Trading In India
On April 3rd 2008, Securities & Exchange Board of India (SEBI), started allowing Direct Market Access facility which allows buying or selling of orders by institutional clients without manual intervention by brokers. Direct Market Access (DMA) enables clients to access the exchange trading system through brokers’ infrastructure but without manual intervention.
Amongst the global recession, this decision by the Indian regulators was a welcomed change by the entire banking and securities market change.
It was expected that this change would result in greater transparency, increased liquidity, lower impact costs for large orders, better audit trails and better use of hedging and arbitrage opportunities.
Investors in Indian markets, on April 4th 2008, received direct access to the exchange’s trading system, albeit through the broker’s infrastructure, a practice very popular in developed markets. At this stage, only institutional investors were allowed to access DMA. Nevertheless, the facility brought down costs for the institutional investor as well as help in better execution by cutting on time spent in routing the order to the broker and issuing the necessary instructions.
April 29th 2008, this facility had already become popular enough to have a sustained flow with more and more players signing up for the DMA facility. The list of pending applications was dominated by foreign entities. FI’s & FII’s like UBS, Morgan Stanley, JP Morgan and DSP Merrill Lynch were the entities awaiting approval. Edelweiss Capital, India Infoline and Motilal Oswal Securities were among others who had submitted their request to the stock exchanges.
By July 31st 2008, leading brokerages along with stock exchanges were preparing the ground for operationalising Direct Market Access (DMA), Brokerages such as Citi, Merrill Lynch, Morgan Stanley, JP Morgan, Goldman Sachs, CLSA and Deutsche Equities had started holding test runs of their DMA software, in an attempt to synchronise it with the systems at the stock exchange.
Fast Spread Of Algo Trading In The Early Years
In India, Foreign Institutional Investors (FIIs) were allowed to use DMA facility through investment managers nominated by them, from February 24th 2009.
On June 22nd 2009, Credit Suisse’s Advanced Execution Services (AES) unit launched algorithmic trading in Indian equities. The AES suite of algorithms included traditional algorithmic strategies that seek to divide trading volumes up over time and strategies that seek to trade at the Volume Weighted Average Price of a stock.
NSE’s Contribution To The Industry
The National Stock Exchange (NSE) started offering additional 54 colocation server ‘racks’ on lease to broking firms in June 2010 in an effort to improve the speed in trading.
Deutsche Bank, Citi, Morgan Stanley, Goldman Sachs, and MF Global were among the foreign broking firms which availed of the facility. Motilal Oswal Securities, JM Financial and Edelweiss Capital figured among the prominent domestic firms who signed up for the racks. Local brokerages like Globe Capital, SMC, Global Vision, East India and iRageCapital had also opted for the facility. Not surprisingly, with a few weeks of offering this facility, there was a long period of waiting up to 6 months to get a space on the server racks!
It was clear to the Indian exchanges and regulatory bodies that Algorithmic Trading is well-received by the institutional clients and banks in the country and its demand would continue to rise. This was the time when exchanges started improving their offerings in the automated trading domain, financial technology companies started offering automated trading platforms and SEBI continued to regulate the markets.
NSE Adapts FIX Protocol
May 12th 2010, NSE moved to enable the Financial Information Exchange (FIX) protocol on its trading platform boosting transaction speed for overseas investors using direct market access. A fund manager sitting outside India can buy or sell shares hereby routing their orders through the broker’s system to the exchange system but without any manual intervention from the broker.
In layman’s terms, the orders coming from the system of Foreign Institutional Investors (FII) outside India is in a language different from the one understood by the NSE’s system. What the FIX protocol does is to quickly convert into a language understood by the stock exchange. This reduces the time taken for the transaction to be executed.
NSE and BSE had provided a suitable environment for algorithmic trading to grow over the past few years. To read about how the trading ecosystem had evolved and which facilities are available in the Indian markets, read our other post here.
Brokerage Industry Started To Change
August 17th 2010, broker commissions had started shrinking as a result of an increasing number of institutional clients warming up to the Direct Market Access (DMA) concept. The brokerage industry warmed up to this change by offering automated software to the changing market demands. Listed below are few vendors in India who provide API for HFT:
Market data is provided by Globaldatafeeds for backtesting and paper-trading. Another stack is called Presto provided by Symphony Fintech. This has a few disadvantages in terms of how they setup costs. Usually, a broker had to enter into a specific agreement with clients for whom they permit the DMA facility, which clearly stated that the client will use the DMA facility only to execute his own trades and would not use it for transactions on behalf of any other person/entity.
Regulations In Indian Stocks Markets
Every year SEBI comes up with regulations required to be followed by traders and brokers to keep the trading industry safe and risk controlled. Over the years, SEBI streamlined DMA facility for trading by issuing a list of dos-don’ts for traders and brokers. To read about SEBI’s recent announcement regarding the algorithmic trading industry in India, go to the post here.
Risk management is critical with algorithmic trading. That is why, for any HFT algorithm to be approved by the markets, exchanges require a firm to undergo a series of stringent tests if it intends to trade through HFT. These tests include the number of orders that would be placed per second, the maximum order value of any order placed, and the maximum traded quantity during a particular trading day.
Changing Trading Ecosystem
Update: TBT (Tick by tick market data) is now available in multicast and not just in TCP/IP format. Compliance requirements have changed. Empanelment process has also been changed significantly.
In this video, we learn about how the changing trading environment towards Algorithmic Trading and how it has become more conducive to Algorithmic Trading. Please note this video is created from one of our old webinar held on 23, Sep 2013.
The following topics are introduced in this short video:
- Colocation facilities: Reduce the time taken for orders to reach the exchange. A reduction in a half millisecond is a big improvement in HFT trading strategies.
- Tick by tick data: Create order book of any depth using tick by tick data
- Normal vs. Bucket feed: Subscribe to only a few instruments for which you require the data
- High Capacity Interactive Lines: Send up to 400 messages every second to exchange
- Smart Order Routing: Implement systems which can pick any of the multiple exchanges to send orders at a more favourable price
- Speedy empanelment process: Streamlined processed to get your strategy empanelled at the exchange
- New membership categories: Alpha category memberships available at NSE
- Compliance: More stringent and vigilant
These are some of the new initiatives which have taken place in NSE, the leading stock exchange in India, located in Mumbai.
Future Of Algorithmic Trading In India
With several amendments over the years, India provides a good opportunity for HFT traders due to a number of factors such as colocation facilities and sophisticated technology at both the major exchanges; a smart order routing system; and stock exchanges that are well-established and liquid.
Given the rapidly growing trend and demand of HFT and Algorithmic Trading in developing economies & emerging markets, there have been efforts by various exchanges to educate their members and develop the skill sets required for this technology-driven field. To empower the trader to face the challenges in trading QuantInsti has joined hands with NSE to provide short Management Development Programs in Algorithmic Trading. These are two-day programs to initiate the brokers and traders into this complex and challenging multi-disciplinary field, comprising of sessions on Statistics, strategy writing and using financial computing tools. Read about upcoming NSE’s Management Development Program on Algorithmic Trading here.
Frequently Asked Questions about the Future of Algorithmic Trading
Here are some of the most commonly asked questions which we came across during our Ask Me Anything session on Algorithmic Trading.
Question: Is Algorithmic Trading legal in India?
Reply: Definitely Yes! April 03, 2008 is when SEBI allowed algorithmic trading in India, so since then it has been legal.
Question: How tedious is it to get legal approval for any automation? How confidential and secured it will be if it goes to automation after approval, is approval process and infrastructure cost affordable for retail traders?
Reply: The approval process is not that costly, but yes the infrastructure, if you are going for HFT can be a big burden if you are a retail trader or individual trader but you can do automation and that would not be a huge cost as such.
Assuming this is from an Indian market perspective, India has a peculiar regulation which says that you have to approve each and every strategy before you take it live. This is different from most of the developed market regulations in which you have to get the platform approved and then you can code any strategy you want to on that platform. Same goes for other developing markets like Thailand where you have to get every algorithmic trading strategy approved before you can automate. The regulation demands that the broker should take the approval on your behalf, you as a retail trader cannot go to the exchange and ask for approval. The cost depends on the broker but technically it’s not that costly.
Question: What are the approvals you need before going algo?
Reply: We touched upon this in brief in one of our previous questions but it depends on which geography you are trading into. In case you are trading in the CME, SGX or Eurex then the approval required is more of a conformance test which means that you will be taking approval for your trading platform. Once it is approved you can code whatever strategy on it and send out orders.
In case you are in geographies like India or Thailand then you will need to get your strategies approved and for that what you will be doing is creating a document for each strategy and sending it out to the exchange for approval. If you are a member of the exchange yourself you can send it directly and if you are not a member with the exchange then you send it through a broker. The process in India involves (can vary for different exchanges) to get the strategy signed from the auditor, participate in a mock trading session, then you demo it with the exchange, post that you get an approval from the exchange and then you start trading. That’s the rule you have to follow for each strategy.
Question: How is a strategy confidential if it is going through the approval process?
Reply: The exchanges generally do not focus much on the strategy but more on the risk management. The focus is that your strategy should not create havoc for the market or for them, which is the key concern for the exchange and not what your strategy does. They would ask you about the strategy at a broad level but I don’t think it goes to a level where your IP is threatened.
Question: How risky is algorithmic trading towards manipulation such as colocation?
Reply: Colocation is not manipulation. It’s just a facility provided to you. It’s like saying how risky it is if you are travelling by air by spending more as compared to someone who is travelling by train to a destination, you are reaching faster but you are paying for it and you are getting it so it’s a fair market, you pay for what you get.
For those who colocation matters and for most of the exchanges across the globe it is not that expensive hence the exchanges also have been pretty responsible. Even in India you can get half racks (which is 21 units) you can place a good number of servers in half rack and that comes to around 50,000 rupees a month. I am not saying it’s very cheap but it is not that stringent if you are trading into strategies which are depended upon colocation for which every millisecond matters.
You can review the European regulation applied in the 27 EU member countries, specifically for algorithmic trading in this blog.
If you’re a retail trader or a tech professional looking to start your own automated trading desk, start high-frequency trading training today! Begin with basic concepts like automated trading architecture, market microstructure, strategy backtesting system and order management system. QuantInsti® aids people in acquiring skill sets which can be applied across various trading instruments and platforms.
One can learn about Algo Trading through this starter pack course on Algo Trading through Trading strategies, ML, Stat Arb and much more.
Disclaimer: All data and information provided in this article are for informational purposes only. QuantInsti® makes no representations as to accuracy, completeness, currentness, suitability, or validity of any information in this article and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. All information is provided on an as-is basis.