Algorithmic Trading: Meaning, Benefits and Concerns – Explained, pointwise

Introduction

The Securities and Exchange Board of India (SEBI) recently came out with guidelines for stock brokers who provide services relating to algorithmic trading to investors. The aim of the guidelines is to prevent instances of mis-selling by the brokers. SEBI had noticed that certain stock brokers provide algorithmic trading facilities to investors through unregulated platforms. In the last few years, algorithmic trading and co-location have been in the news for both good and bad reasons. Algorithmic Trading, also known as Algo Trading, became legal after SEBI approved Direct Market Access (DMA) in 2008. Major concerns were raised about algo trading when it was revealed that the National Stock Exchange (NSE) gave preferential access to a few algo traders in 2015. In December 2021, SEBI issued a discussion paper proposing that all orders emanating from stockbrokers’ Application Programming Interface (API) be treated as algorithmic trading, raising concerns that such restrictions will stifle the growth of algo trading in India.

What is Algorithmic Trading?

Algorithmic trading is computer assisted buying and selling of stocks. It is also known as automated or programmed trading. In algorithmic trading pre-programmed computer strategies execute buy and sell trades depending on set parameters, instructions or market pattern and conditions. The instruction can be to buy or sell a particular number of share(s) of a specific company as soon as the price reaches a certain pre-defined level.

The trades happen at a super-fast speed by the use of advanced mathematical models that involve automated execution of trade. The execution speed is so fact that even a split-second faster access is considered capable of bringing huge gains to a trader. The algorithms are run on computer systems owned by traders (and not investor). The algorithm automatically executes the order with no human involvement from either the broker or the investor. The algorithmic trading system automatically monitors the live stock prices and initiates an order when the given criteria are met. This frees the trader from having to monitor live stock prices and initiate manual order placement.

Assume that an investor/trader wants to buy shares of XYZ Corporation if it rises by 2% over its previous closing price and if trading volumes on that day are higher than the average volume of the last 10 days.

The investor can put this order through a broker’s website or app, and wait for the order to be executed. However, the investor will only be able to specify the price, not the volume. Another way is to monitor live stock prices and manually place the order when the order is bet.

A third way is to have a software with an algorithm that will place the order only when both conditions are met. This will be done automatically without human intervention.

Source: Turing Finance. The system picks information from database (e.g., market data related to price) and other latest information. The algorithm compares the data with instructions (model) and executes the order. With advancements in Machine Learning, the models can improvise/adapt themselves for even better/faster executions.

What is an Application Programming Interface (API)?

An API or application programming interface is a set of programming codes that analyses data and sends instructions between one software platform and another.

Investors/traders typically look for opportunities by screening stocks based on certain parameters (like trend of price). Earlier, they had to identify such stocks on a separate application and then place trades with their broker. In other words, the trader’s screening software and the broker’s software could not directly communicate with each other. But with brokers providing APIs, traders can connect their screening software with the broker’s API offering access to real-time prices. The trader’s screening software will identify opportunities based on the live prices, and automatically place orders.

Every broker has an open Application Programme Interface, or API. API allows a trader’s software to connect to the broker terminal. Brokers, at their end, have an order management system which is used to place an order electronically.

What is the current status of Algorithmic Trading in India?

Algorithmic Trading was allowed in India in 2008.

According to a report by the National Institute of Financial Management (under the Union Ministry of Finance), ~50%+ of total orders at both NSE and BSE are algo trades.

Many brokers in India have started providing Application Programming Interface (API) access to their clients. This has allowed small investors (retail investors) to invest trough algo trading. Companies like Zerodha, Upstox, 5 paisa are amongst the few brokers who provide algo trading platforms to retail clients for algo trading.

What are the benefits of Algorithmic Trading?

First, algorithmic trading enables faster execution of trades and thus increases the efficiency.

Second, It reduces chances of human error. It also eliminates human emotions which might impact actual execution of order (like last minute dilemma whether to buy or sell).

Third, It makes the market broader and improves the market quality. From a broker’s or trader’s perspective, the algo can be set in such a way as to get the best possible price.

Fourth, algos are executed automatically and with great speed, and substantially reduces the risk for traders.

Fifth, it improves the liquidity in the market. An algo keeps throwing orders in the market and then withdrawing it. This is something that SEBI plans to regulate more closely now. The bottom-line is that this aggressive interplay of orders helps improve the liquidity in the market and facilitates the execution of transactions seamlessly.

What are the concerns related to Algorithmic Trading?

First, it increases market volatility (i.e., large variation in prices, market instability). Even a small fall in the market can trigger a mass sell order, leading to a crash. Algorithms lack human intuition and hence may not be able identify events of panic selling during market crash further exacerbating such crashes.

Second, algorithmic trading is susceptible to system failure risks and network connectivity errors etc.

Third, the accuracy depends upon the robustness of algorithms. A faulty algorithm can result in large losses to the traders/investors. In many cases, software engineers writing the algo strategies may lack understanding of working of markets/trading of shares resulting in poor algorithms.

Fourth, although it reduces chances of human errors but can’t eliminate them. Algo trades are susceptible to fat finger trades and algos misfiring. Fat finger trade is a human error while punching an order. This can include entering a wrong value in terms of price or quantity or selection of the wrong execution action such as buy or sell.

Fifth, The trading strategies for algorithms are often written and analysed by the concept of ‘backtesting’. Backtesting is analysing past historical financial data to generate a set of trading signals. There are three problems with the approach; (a) Bulk of the back testing is generally based on data analysed over the last two-three years. However it is a narrow horizon and may fail to identify a bubble (when share prices are exceptionally high) or bust (when share prices are extremely low say during a recession); (b) Past performance cannot guarantee future returns; (c) The strategy writers and the marketplaces remain unregulated as of now.

Sixth, algorithmic trading is susceptible to scams like co-location scam as happened in the National Stock Exchange.

Source: The Times of India

Seventh, For the algos deployed by retail (small) investors using APIs, exchanges are not able to identify if the particular trade emanating from API link is an algo or a non-algo trade. This kind of unregulated and unapproved algos pose a risk to the market and can be misused for systematic market manipulation as well as to lure the retail investors by guaranteeing them higher returns.

What are the SEBI Regulations/Guidelines regarding Algorithmic Trading?

Brokers need to take the approval of all algos from the stock exchanges. Each algo strategy has to be certified by Certified Information Systems Auditor (CISA)/ Diploma in Information System Audit (DISA) auditors. They need to inform the exchanges (BSE, NSE etc.) of any changes to the algos.

All orders emanating from an API should be treated as an algo order. The APIs to carry out algo trading should be tagged with the unique algo ID provided by the stock exchanges.

All algo orders have to be routed through broker servers located in India. Also, all algo orders have to be tagged with a unique identifier provided by the stock exchange in order to establish an audit trail. This allows the exchange to know if an order is an algorithmic one or non-algorithmic.

Brokers shall ensure that appropriate checks are in place so as to allow only exchange approved algos.

The stock broker will be responsible for all algos emanating from its APIs and redressal of any investor disputes.

What are the shortcomings/concerns related to the SEBI regulations?

First, the third party algo providers are not regulated. There is also no investor grievance redressal mechanism in place.

Second, getting the requisite permission from the stock exchanges is a tedious process, brokers may have to stop using the API system. This will impact development of the market.

Third, submitting algo programmes to exchanges for their approval would mean that vendors may have to reveal their approach. There is a possibility of the intellectual property (IP) being compromised—some brokerages can easily replicate the successful algo strategies.

What are the global rules on Algorithmic Trading?

In April 2016, the U.S. Securities and Exchange Commission (SEC) approved a rule proposed by the Financial Industry Regulatory Authority (FINRA) that would require algorithmic trading developers to register as securities traders. The move was primarily aimed at reducing market manipulation.

In the UK, any market participant must notify the regulator Financial Conduct Authority if it is engaging in algorithmic trading. A firm must provide information like nature of its algorithmic trading strategies and details of the trading parameters to the FCA.

What should be done going ahead?

First, there is a need to establish an investor grievance redressal for trades undertaken by third party vendors.

Second, the third-party algo providers must also be tightly regulated.

Third, SEBI must tighten the regulatory norms so that the Co-location scams that happened in NSE can be prevented.

Fourth, there is a need to raise awareness among the retail (small) investors in the stock market about the risks associated with algorithmic trading, so that they are able to use such investing strategies more prudently.

Conclusion

India’s dematerialised (demat) accounts have crossed the 100 million mark, up from about 40.9 million in March 2020. This number is further going to increase in the coming year due to digitalisation and more awareness about the economy. As the participants in the stock markets increase, so do the risks in the market. In this context, the role of the market regulator becomes critical in order to protect new or small investors while also ensuring accountability on the part of the various market players. SEBI’s recent guidelines against algorithmic trading facilities for investors provided by unregulated platforms are commendable. SEBI must expand these regulations to cover the unaddressed aspects of algo trading and protect investors’ interests.

Syllabus: GS III, Indian Economy and issues related to growth.

Source: Mint, Indian Express, The Hindu BusinessLine, The Hindu BusinessLine, CNBC