Capital Market

1000

Algorithm Trading - Catalyst to Revolutionise

Indian Financial Markets

As in all other fields of life, technology seems to be revolutionising every activity in a fundamental way,
forcing us to rethink the way we do things. The recent advancement in digital technology has been a
game changer for the financial sector. It has touched upon almost every aspect of the market, whether
it is in terms of the huge increase in accessibility of the market, the speed of making transactions or the
ease with which resources can be mobilised and distributed across different segments. In this article,
the author has deliberated upon one such technological advancement, i.e., Algorithm Trading. Read
on to know more….
Almost three decades back, India started its
rendezvous with technology in securities markets.
When screen based trading replaced the open outcry
system in 1994, far ahead of time as compared to
several developed markets, we embarked on the
CA. Praveen Garg first generation reforms in securities market. It
(The author is member of the Institute was followed by electronic holding of shares and
and Joint Secretary (Financial Markets), setting up of world class depositories. Around 2000,
Department of Economic Affairs, Ministry
of Finance, Government of India. He can be SEBI ushered in the second generation reforms
reached at [email protected].) with Internet Trading by which a client sitting in

96 THE CHARTERED ACCOUNTANT JANUARY 2018 www.icai.org

 Capital Market
1001

any part of the country could trade using Internet from 17 to 13 milliseconds. HFT activity, benefited
as a medium through brokers' Internet trading from technology that reduced delay, or latency, to
and order routing systems. In 2008, we started the markets. In addition, by co-locating their servers
the third generation reforms with the launch of with market servers at an exchange, the algorithmic
Direct Market Access (DMA). DMA is a Computer traders, including HFT firms, increased the speed
to Computer Link (CTCL) facility which allows at which they could access the markets. Speed may
brokers to offer institutional clients’ direct access to not necessarily result in greater translation of those
the exchange’s trading system through the broker’s orders into trade, which is conducted on a time-price
infrastructure without manual intervention by priority basis by the exchanges. But speed matters.
the broker. DMA offered direct control of clients Asian countries were one of the first movers to
over orders, faster execution of client orders, and embrace HFT in order to bring in market liquidity.
reduced risk of errors associated with manual order In developed markets, Algorithmic Trading stands at
entry, greater transparency, increased liquidity, about 80% of the total turnover. In the NIFM study, it
lower impact costs for large orders and better audit has been found that around 50% plus of total orders
trails. DMA also helped in better use of hedging at both NSE and BSE are algo trades–client side.
and arbitrage opportunities through the use of Proprietary side algo trades are 40% plus of total
decision support tools like algorithms for trading. orders placed at both the exchanges. More than 80%
The use of algorithms has matured enough and has of the algo trades are generated from the co-location
now manifested as High Frequency Trading (HFT) at both the exchanges.
and colocation which has increased the speed and One of the biggest advantages of Algo Trading
efficiency of trading manifold. The fourth generation is the ability to remove human emotions from the
reforms have already started with block chain markets, as trades are constrained within a set of
and distributed ledger techniques which have the predefined criteria. As NASDAQ puts it, the two
potential for redefining the trading platforms. emotions that lead to poor decisions that algo traders
Algorithmic Trading or simply Algo Trading are not susceptible to, are fear and greed. Further,
involves the use of a basic algorithm i.e., a set of rules it provided the traders, the ability to backtest. They
or instructions, to feed orders into the market at pre- can run the algorithms based on past data to see if
set intervals to minimise market impact cost. At its it would have worked in the past and help them in
complex form, it may entail many algorithms that are removing any flaws before it is actually run. Thus,
able to assimilate information from multiple markets Algos crystallise past trading wisdom of years to
in different assets and to use this to implement a high- corner every profitable opportunities and arbitrage.
speed, multi-asset trading strategy that transacts As machines take over, transaction cost reduces since
numerous inter-related trades in fractions of a traders need not be glued to their systems, monitoring
second1. These mathematical models analyse every the news and price movements. Particularly for big
quote and trade in the stock market, identify liquidity institutional players, Algo Trading can be used to
opportunities, and turn information into intelligent break up a big purchase into small parcels so that the
trading decisions. Fuelled by economics, it is maths price of an asset is less affected than if it were to be
and science combined; and rules and logic married; bought in one single act of purchase. It helps short
high-speed trading and “High-Frequency Trading”
(HFT) refers to automated trading conducted at
millisecond or microsecond or nanosecond speeds Algorithmic Trading or simply Algo Trading involves
throughout the trading day. It is a type of algorithmic the use of a basic algorithm i.e. a set of rules or
trading only. The rise of HFT has led to the practice instructions, to feed orders into the market at pre-
of co-location–putting the servers of trading firms at set intervals to minimise market impact cost. At its
exchange itself or closer to the exchange servers so complex form, it may entail many algorithms that are
that order sending or receiving time can be cut short. able to assimilate information from multiple markets
Michael Lewis in his book, Flash Boys mentions in different assets and to use this to implement
a high-speed, multi-asset trading strategy that
about a $300 million project for constructing an 827
transacts numerous inter-related trades in fractions
mile long cable that cuts straight through mountains of a second.
and rivers from Chicago to New Jersey-with the
sole goal of reducing the transmission time for data

Regulatory Issues Raised by the Impact of Technological Changes on Market Integrity and Efficiency, IOSCO Consultation Report, July 2011
i

www.icai.org THE CHARTERED ACCOUNTANT JANUARY 2018 97

 Capital Market
1002

the fairness and integrity of the markets. At the same

time, it cannot be denied that even at present, access
The main objective of Algo Trading is not necessarily is unequal as trading member’s servers are located
to maximise profits, rather to control execution costs
at varying distance from those at exchanges. Some
and market risks but the lack of control has led to
systemic risks, such as flash crashes. feel that competition is not materially distorted, as
long as any market participant wishing to obtain
co-location space can do so on fair terms and if it
term traders and market makers to take advantage is the case then the rules apply to all. But it needs
of short-term market inefficiencies like arbitrage. to be seen whether HFT coupled with co-location
As the cliché goes, every coin has two sides. The is accentuating the present minor disparity and
proponents of Algorithm Trading argue that it has whether the increased costs of operation for smaller
brought depth and much desired liquidity into the players are acting as entry barriers and thus, anti-
market while the counter view is that it brings price competitive. A study needs to be undertaken specific
volatility and noise into the market. Algo Trading to India, to look into the statistically significant gains
requires a lot of technical programming skills, which an HFT player is getting as compared to a normal
can take quite a while to learn. Further, automation broking firm.
can lead to lack of control. The main objective of SEBI came out with its first comprehensive
Algo Trading is not necessarily to maximise profits, guideline on algos in March 2012 which were further
rather to control execution costs and market risks revised in May 2013. In its Consultation Paper of
but the lack of control has led to systemic risks, such August, 2016, SEBI has proposed certain options like
as flash crashes. As pointed out by International Minimum Resting Time for Orders, Time Stamping,
Organisation of Securities Commissions (IOSCO)2, Frequent Batch Auctions, Random Speed Bumps or
if trading venues experience outages, and particularly Delays in order processing/matching etc. Separate
if this is the result of technology not being managed queues for co-location and non-co-location orders
effectively, investor confidence and market integrity have also been proposed. In fact, certain measures
as a whole can be negatively impacted. have already been put in place to improve Order-
Meanwhile, it has also been proved in the past that to-Trade Ratios. SEBI has received comments from
Algo Trading and HFT can be used to manipulate various stakeholders of the securities market on this
market using techniques like quote stuffing , layering and is examining the same.
(spoofing) and momentum ignition. Evidence IOSCO Principles on Market Integrity
suggests that market manipulating algorithms lead and Efficiency released in October 2011 has
to decreased liquidity, higher trading costs, increased recommended that ‘Regulators should require, that
short term volatility, impact performance and fill trading venue operators provide Fair, Transparent
rates, and result in massive price moves, backed and Non-discriminatory access to their market
by false volume. Increased speed provides limited and to associated products and services.’ IOSCO
opportunities for regulators to intervene during high Recommendation 3 says, “Regulators should seek
volatility/uncertainty. Further, cascading effects and to ensure that trading venues have in place suitable
negative network externalities of technical errors trading control mechanisms (such as trading halts,
cannot be overlooked. volatility interruptions, limit-up-limit-down
While algos are broadly accepted as a matter of controls, etc.) to deal with volatile market conditions.
principle, questions have come up on the necessity Trading systems and algorithms should be robust and
of adopting HFTs and co-location. As rightly pointed flexible such that they are capable of dealing with,
out by International Organisation of Securities and adjusting to, evolving market conditions. In the
Commissions (IOSCO), Co-location raises issues case of trading systems, this should include the ability
related to potential distortion of competition to adjust to changes (including sudden increases)
between market members, equal access of the in message traffic”; we may take steps for policy
market and the cost of such services. The fact that formulation keeping these principles in mind.
some participants may receive information on order With the changing market landscape, we may
book, trading interest and executions sooner than conclude to say that Indian market has embraced
others, and have their orders entered in the trading Algo Trading wholeheartedly, the need of the market
system more rapidly, may raise questions regarding is to enhance market infrastructure. 
2
Mechanisms for Trading Venues to Effectively Manage Electronic Trading Risks and Plans for Business Continuity; Consultation Report (April 2015), IOSCO

98 THE CHARTERED ACCOUNTANT JANUARY 2018 www.icai.org

Success is how high you bounce when you hit bottom. - George S. Patton