What is algo trading?
Algo trading-short for algorithmic trading-is a computer programme that automatically places buy or sell orders based on a set of pre-defined rules.
Typically, retail investors must place trades manually through the brokers’ trading app or platform. In algo, software does the trades on its own when certain predetermined conditions are met. You could set an algorithm to buy a stock if it falls below ₹100 or sell if it crosses ₹150.
How different will the new formalised algo regime for retail investors be?
Until now, rules for a lot of retail algo trading were grey, with open APIs or application programming interfaces (digital tools from brokers that allowed traders to use their own algos) and little regulatory oversight. From August 1, all that changes as retail algo trading comes under formal regulatory supervision and only approved software will be allowed to be used for trading.
Can all retail investors access the formal algo trading in the new regime?
All retail investors can access algo trading only if registered with stockbrokers and as per the conditions laid out by Sebi. If you’re using a third-party algo-from a fintech or algo vendor-it must be empanelled with the stock exchange, and your broker must do the necessary checks before allowing it. The algo itself must be registered with the exchange.
Can I use my own algo?
Yes. You need to register it with the exchange through your broker if the algo fires off orders fast enough to cross a limit, measured as the number of orders per second. If the number of orders is below that threshold, the algo need not be registered.
Who can offer algo trading software to retail investors?
You can’t just download an algo software from the internet and start firing trades. Only algo providers empanelled with the exchange are allowed to offer this. These firms can provide it only through a registered broker. Every algo strategy must be registered with the exchange and will be assigned a unique ID.
Will new rules give retail investors a fair shot against institutions and proprietary desks?
Automated trading has mostly been the preserve of institutional investors, trading desks, and FPIs. Big trading firms like Jane Street use super sophisticated algorithms that execute orders in microseconds. By formalising it, Sebi has tried to democratise it.
But that doesn’t mean that the playing field is level-the big players will continue to have an advantage in terms of speed, flexibility and firepower.
How different will the algo software for retail investors be from what the big players use?
The algo software used by retail investors is not the same as what big institutions or proprietary desks use. The retail version will have to go through strict checks and approvals, while institutional and prop trading desks use in-house algos, which are customised and faster.
Several large institutions connect directly to the exchange through direct market access (DMA)-a system that allows them to place orders directly, bypassing the broker. This route is not available to retail investors. DMA gives algo trades the edge in terms of faster execution. While retail investors will also be allowed to access algo software, the level of regulatory controls is tighter, while speed and flexibility are modest compared to what institutions use.
Is it safe for retail investors to use algo software? What if the algos act up?
Sebi’s safety nets are aimed at ensuring that the algo trades do not snowball into a bigger issue. While regulations ensure oversight by brokers, exchanges and Sebi, one key safeguard is the ‘kill switch’, which allows exchanges to halt trades from a particular algo. This is aimed at averting a systemic disruption on account of one algo.
Will algo trading make a meaningful difference to the way retail investors trade?
Algo software for retail traders may not be as efficient as what institutions use, but it does open possibilities. Automation brings discipline to trading, while retail investors can do rule-based trading strategies consistently without much effort. So, retail algos may not outrun big machines, but it’s a good place to start.
