Algorithmic Trading Future

What is the future of Algorithmic Trading?

As I write this answer in NSE about 55% of the Orders are fired by Algo trading. In developed markets this percentage is even higher probably around 75%.

Let’s look at the evolution of Trading in exchanges.

1900 - 70s:

Open outcry Sytem:

Open outcry is a method of communication between professionals on a stock exchange or futures exchange typically on a trading floor. It involves shouting and the use of hand signals to transfer information primarily about buy and sell orders. The part of the trading floor where this takes place is called a pit.

In an open outcry auction, bids and offers must be made out in the open market, giving all participants a chance to compete for the order with the best price. New bids or offers would be made if better than previous pricing for efficient price discovery. Exchanges also value positions marked to these public market prices on a daily basis.

This is how it used to be:

From 1970s to 2000s:

Conversion to Electronic Trading:

A computer software program that can be used to place orders(Bid/Ask) for financial products over a network with a financial intermediary. Various financial products can be traded by the trading platform, over a communication network with a financial intermediary(Brokers, Exchanges et all) or directly between the participants or members of the trading platform.

This is how it looks like:

From 2000:

Algo Trading:

Algorithmic trading is a method of executing orders using automated pre-programmed trading instructions. Most of the Investment banks, Hedge funds, Institutional investors adapted to this Algo trading method.

This is how it looks like :)

Below few examples that gives a glimpse of how much markets are depending on the Algo trading these days.

Flash crash of Dow Jones in 2010:

There was a famous flash crash happened in 2010 where Dow Jones almost plummeted around 9% for about few minutes. There are many documentaries, books that explained this flash crash. But to keep it simple this flash crash was created as a result of automated bots fighting over each other. There was a really big Short order(probably by a mistake) that entered the market(according to references the order was about 4 billion USD). This big order was immediately detected by the high frequency algorithms across the exchange and immediately withdrew all the Bids in the markets. As a result momentarily there were no Buyers in the exchanges and Stocks started to fall.

Here is the wiki link if you want to know more about the 2010 Flash Crash.

2010 flash crash - Wikipedia

Elon Musk Tweet:

1st May 2020 - Elon Musk tweeted that “Tesla stock price is too high imo”. With in few minutes Tesla stock fell about 10%. His own tweet costed Elon Musk 20 Million USD.

These days live news feed from many sources (ET, WSJ, Bloomberg, Reuters & Twitter et al)are being read by Algo softwares. These softwares perform Sentiment Analysis on all the news feeds and assign a positive and negative ranking. Based on these rankings Algo Softwares fire up Buy or Sell orders automatically. Here is my answer on this.

Now there is a whole new avenue ruling Stock markets these days i.e; Machine Learning. Using Machine Learning we can train the systems to learn new patterns themselves rather than Programmers feeding the patterns like momentum, Eliot Waves et al. Humans can identify probably 100 patterns by themselves but with Machine Learning algorithms Programs can learn infinite themselves by themselves and make them better each day.

Now to your question on future of Algorithmic trading: Until about last 5 years only big institutional investors and Investment banks can afford the infrastructure needed for Algo trading. But now with many brokers offering their own APIs to automate Orders, providing historical and tick data, even Retail investors can adapt to Algo Trading with minimal infrastructure.