Risks in Algorithmic Trading and Measures Taken By SEBI
Many individuals confuse algorithmic trading with quantitative trading. All quantitative trading setup is also an algorithmic setup; however, this is not always the case.
An algorithm is the automation of any given setup so that trades are created automatically. Algorithm trading (also known as Black Box Trading) is a very sophisticated trading technique that use complicated mathematical formulae and models to make rapid judgments and transactions in financial markets.
Aside from earnings, algorithm trading provides market knowledge that may be used to help make long-term trading decisions.
There are no hard and fast rules for constructing algorithms; they evolve in the same way that nature does.
Risks in Algorithmic Trading:
While speed is one of the benefits of algorithmic trading, it comes with the danger of losing a lot of money.
High volatility: Algorithms may increase the bid-ask spread or suspend trading in response to changes in market circumstances, resulting in excessive volatility and in some cases reduction in liquidity.
Machine Dependence: When you switch to an algorithm, you are fully reliant on the successful and efficient operation of your servers. If your system hangs or is unable to connect to the exchange, you may be unable to trade until the problem is resolved.
Chain reaction: Because global markets are highly integrated, a slowdown in one market spreads to other markets and asset classes, causing a chain reaction. (As was the case throughout the subprime mortgage crisis.)
Incorrect Algorithms: A flawed algorithm can lead to errors in transactions and market manipulation, resulting in millions of dollars in losses in a relatively short amount of time.
Lack of Knowledge: Because brokers provide a wide range of algorithms, the buy side lacks the tools to assess which algorithms are appropriate for their portfolio. This reduces the accuracy of algorithm evaluation. The ready to use algorithms provided by the 3rd party brokers
Many algorithms employ identical fundamental functions to operate on current market conditions, which might result in undesirable results.
Imbalances in the market: While only a few traders can afford sophisticated equipment that executes orders, the majority still trade manually. This generates market fragmentation, which leads to liquidity in the short run.
Measures Taken By SEBI:
To make algorithmic trading more accessible to investors, the Securities and Exchange Board of India (Sebi) established recommendations. It requested that stock exchanges implement shared co-location services in order to save expenses.
Orders created utilizing automated execution logic are referred to as algorithmic trading. Members can co-locate their servers on the exchange’s premises for speedier trading access. Sebi presented a system for penalizing excessive order-to-rate ratios.
A high order to trade ratio indicates that many orders are not turned into transactions. The regulator also directed bourses to give free tick-by-tick data stream (TBT Feed) to all trading members.
This will allow every trading member to know the best pricing of any particular stock in real-time, allowing everyone to benefit equally from price changes in stocks.
Separately, Sebi established the so-called product appropriateness framework to provide ordinary investors with a safety net when investing in equity derivatives. There is a concept known as product appropriateness.
Investors in stock futures will be required to report income in their tax returns, and if the investment is not consistent with income, brokers will be required to do due diligence, according to a news release from Sebi.
Sebi tightened the eligibility conditions for scrips trading in the futures, options, or derivatives market. It stated that scrips that do not match the higher compliance standards would be forced to settle physically.
SEBI has released paper on December 9, 2021, expressing concern over algorithmic trading by retail investors and seeks views from the industry and the public on the potential threats it has. Resulting in Sebi’s circular putting immense pressure on the brokerage houses to “A stockbroker distributing the API is also supposed to ensure that any trading happening through its API has to be tagged with the unique algo ID provided by the stock exchange after the approval is granted for the algo.”
This requirement as posed by SEBI looks fairly simple on papaer but in fact it is not. The trades follow a pre-fixed architecture that recognizes the order parameters and has to be similar for all the orders across all the brokers for the order matching engine to match the orders. This system has been in place since the advent of electronic trading. And changing the age old infrastructure in place will come at a great cost, which is ultimately going to be trickled down to retail traders.
Sebi has also strongly asked the brokers to get the algo approved in order to minimize systemic risks. The problem arises that the algorithms are proprietary to the market participants and getting the algorithms audited is like sharing the Coca-cola recipe, which will curb the interest of the public slowly adopting algo trading.
Nitin Kamath, founder of Zerodha, has stated in his reply that there are many ways in which third party software can generate orders through api’s and mask it as a manual order while pushing the trade into the system, so the regulation will deem useless in such cases. Just like making a robot mannequin and programming it to punch orders on a keyboard, just like a human trader.
However, it is equally critical to construct such algorithms with the highest precision and to provide high-level security to protect them from cyber-attacks.
