Algorithmic trading is a step-by-step instruction for trading actions taken by automated systems. It refers to any order that is generated using automated execution logic. As the word ‘algorithmic’ suggests, this trading involves complex formulas, mathematical models and delegation of decision making to the algorithms, in turn, enabling traders to generate a large number of orders in a small interval of time, and at the same time, react to opportunities that may exist for fractions of a second. Algorithmic trading is the use of computer and defined set of instructions about placing trades at a speed and efficiency unavailable with a human trader. This type of trading can be traced all the way back to the ‘70s’ New York Stock Exchange and the introduction of Designated Order Turnaround (“DOT”) which facilitated routing of orders from traders to specialists on the exchange floor and an eventual acceptance of electronic trading. This trend came to India in 2008 (amidst the global recession) with the introduction of Direct Market Access Facility (“DMA Facility”) allowing institutional trades without manual intervention. The main aim of Algo-trading was faster communications with exchanges to gain advantage of speed and quantity of trades over competitors. Foreign Institutional Investors were allowed to use DMA facility through nominated managers from February, 2009. Now, a huge part of trading in India is done through Algorithmic Trading.
June 2010 saw the introduction of Co-location services at National Stock Exchange (“NSE”). SEBI, vide circular CIR/MRD/DP/07/2015 dated May 13, 2015 and circular SEBI/HO/MRD/DP/CIR/P/2016/129 dated December 01, 2016 laid down the ‘Guidelines to ensure fair and equitable access to the Co-location/proximity hosting facility offered by stock exchanges’. This has provided the vehicle to high-frequency traders to capture such trading opportunities. This saw major national and international broking firms signing up for racks. In 2010 NSE enabled the Financial Information Exchange (“FIX”). At present, NSE provides Co-location facility for NNF trading on NSE.
High-frequency Trading (“HFT”) involves algorithmic strategies which get executed in an automated way in quick time, usually on a sub-second time scale. It is a subset of algorithmic trading comprising of latency-sensitive trading strategies and deploys technology including high-speed networks, colocation, etc. to connect and trade on the trading platform. Its success is largely attributed to their ability to react to trading opportunities that may last only for a very small fraction of a second.
According to the International Organization of Securities Commissions (“IOSCO”) following are some of the characteristics to identify HFT:
- Sophisticated technological tools for pursuing a number of different strategies;
- Algorithm through the investment chain: analysis of market data, deployment of strategies, minimization of costs and execution of trades;
- High daily portfolio turnover and order to trade ratio
- Flat or near flat position at the end of the trading day (little or no risk is carried overnight saving the cost of capital associated with margined positions)
As with the introduction of any new technology, it takes time for the regulators to come up with an efficient law. The same has been the case with Algo-trading in India. Inevitably, there are the risks of market manipulation and abusive practices lest regulated cleverly. It cannot be ignored, however, that Algo-trading adds liquidity to the market and narrows the bid-ask spread, increases the competency of the market and brings to investors more advantageous prices. While the basic and fundamental notion of ‘reasonable investor’ gets substantially moved if one’s traders are non-reasoning algorithms; these algorithms if put in place after due consideration may be able to tackle fraudulent trends better than human capabilities. Critics have raised that Algo-traders with their highly efficient software and complicated algorithms anticipate large moves in certain stocks like huge purchases by Mutual Funds (“MF”) and then buy stocks at lower price selling it back in less than a second to MFs at a higher price. Front-running becomes another worry.
The other issues are recognized by SEBI are Price volatility, market noise (excessive order entry and cancellation), Cost that high-frequency trading imposes on other market users, a Technological arms race, Limited opportunities for regulators to intervene during high volatility, Strengthening of surveillance mechanism, etc. There is also the added concern of unfair access and inequity to non-colo and non-HFT participants vis-à-vis the participants that use trading algorithms and co-location to trade.
ATTEMPTS BY SEBI
SEBI, vide its Circular CIR/MRD/DP/ 09 /2012 laying down ‘Broad Guidelines on Algorithmic Trading’ dated as early as March 30, 2012 provided that, in order to ensure maintenance of orderly trading, exchanges shall put in place effective economic disincentives w.r.t. a high daily order-to-trade ratio of algo orders of the stockbroker and also take measures to impede any possible instances of order flooding by algos, to identify dysfunctional algos (i.e. algos leading to loop or runaway situation) and giving the exchange power to block the brokers terminal in the last instance. It laid down minimum levels of risk controls by vesting Stock Exchanges with powers to lay down price bands, quantity limits, and value per order to keep algo-trading in check.
Vide Another Circular No. CIR/MRD/DP/16 /2013 dated May 21, 2013 SEBI, taking into consideration the need of the hour and security concerns, tightened the responsibility on the brokers/ traders to conduct system audit every six months and the duty to report deficiency and serious deficiency at priority to the exchanges. Further, the stock exchanges had implemented a framework of economic disincentives for high daily order-to-trade ratio of orders placed from trading algorithms by prescribing penalties in form of ‘charges to be levied per Algo orders ‘at various levels of daily order-to-trade ratio. The present circular directed doubling of the charges. An additional penalty was prescribed to discourage repetitive instances of high daily order-to-trade ratio.
SEBI vide Circular dated May 13, 2015 laid down guidelines for fair and equitable access to co-location facility and to ensure that the facility of co-location/ proximity hosting does not compromise integrity and security of the data and trading systems.
SEBI vide Circular SEBI/HO/MRD/DP/CIR/P/2018/62 dated April 09, 2018 laid down measures to strengthen Algorithmic Trading and Co-location/ Proximity Hosting Framework by introducing Managed Co-location services under the supervision of Stock Exchanges. It also provided for the provision of Tick-By-Tick Data Feed (“TBT”) free of cost to all participant. TBT data feed offered by stock exchanges provides a detailed view of the entire order-book including details relating to addition, modification and cancellation of orders and trades on a real-time basis. It further made provision for compulsory Unique Identification of Algorithms to ensure enhanced surveillance and prescribed testing requirement for testing of Software and algorithm.
SEBI caused anonymous entity to undertake a simulation exercise, to understand the efficacy and outcome of various mechanisms mentioned in the discussion paper
On August 05, 2016, SEBI came out with a Discussion paper on ‘Strengthening of the Regulatory framework for Algorithmic Trading & Co-location’ wherein it made several valuable suggestions such as the introduction of a Speed Bump Mechanism similar to that implemented by the Securities Exchange Commission (USA) and TSX Alpha Exchange (TSXA). This mechanism involves the introduction of randomized order processing delay of few milliseconds to orders with a view to discourage latency sensitive strategies which would affect much of HFT but not so much deter non-algo order flow where delay in milliseconds in insignificant. Further, it was suggested to introduce a Minimum Resting Time of 500 milliseconds with a view to dampen ‘fleeting’ and ‘vanishing’ of liquidity caused by modifying or canceling the order or placing a new order as a reaction to new developments. Another suggestion was to exclude at least one trade for a set number of order messages sent to the trading venue by bringing in a Maximum Order Message-to-Trade Ratio. A Time Ratio to set a time interval for matching of orders short enough to allow for opportunities for intraday price discovery, but long enough to minimize the latency advantage of HFT to a large extent with a view to nullifying the latency advantage of the co-located players to a large extent. And Finally that Separate Queues for colo orders and non-colo orders which includes implementing an order handling architecture comprising of two separate queues for co-located and non-colocated orders such that orders are picked up from each queue alternatively to give a fair chance of execution to non-colo participants and to address concerns related to being crowded-out by orders placed from colocation.
The only suggestion acted upon by SEBI was to make available TBT for free of cost. These suggestions are fundamental in ensuring not only fairness to non-algo/ non-colo participants but also to improve the present trends of persistent order cancellations as also chances of manipulation.
– Manal Shah