What is Algorithmic Trading?
Algorithmic Trading, Quantitative Trading
What is it?
The most basic question here is what is Algorithmic Trading.
Basically, it stands for conducting trading in secondary financial markets with the help of computers based on a set of algorithms. It can be anything such as:
If tomorrow rains, buy the stocks of Coke-Cola.
If a stock has dropped its price in 10 consecutive days then buy.
Sell my portfolio once I made a 10% profit
What constitutes a set of algorithms?
Since an all-round trading algorithm is what everyone wants, some simple simple questions need to be defined:
When to buy?
How much?
When to sell?
Everyone has their own answers to the previous questions, and the answer is often called a "strategy". I guess no one aims to lose his/her wealth as quickly as he/she can (Will there be philanthropists in the financial markets 🧐?)
Do you have some examples of strategies?
The strategies can be of multi-purpose. For example, sell as quickly as they can, buy but try causing minimum price volatility, or the most common one: buy and sell to earn some money.
While nothing is easy when it comes to money talk, there are still some patterns or methods you can follow.
Market Directional Trading
The most common one is to bet on the market trend of a stock. If you foresee a stock will go up -- buy! if you predict a stock is going to drop -- short!
Arbitrage
Even if it is said that the financial market should be efficient but it's not always the case especially when it lacks liquidity. If you spot a price error based on your observation then do it quickly before anyone else. Then you will be rewarded for the rectification.
Any real examples?
Yes of course. I can give you some ideas about profit-making strategies.
If you see the forex market, where 1 USD = 140 JPY, 1 NZD = 90 JPY and 1 USD = 1.6 NZD. You should:
Borrow 1000 NZD
Sell NZD to 90000 JPY
Sell JPY to 642.86 USD
Finally convert USD back to 1028.57 NZD
Yes, that's how you capture a market error and make a $28.57 profit. However capturing market error often requires to be quick, quick and quick, which often be called High-Frequency Trading.
The speedrun game of High-Frequency Trading may not suit everyone, another option is to find your mathematical interpretation of the market. Imagine that you believe the valuations of Coke and Pepsi should be the same given that what they sell are almost identical (Yes, they taste the same 🤥!)
You notice that Coke's stock price is normally 5.5 times Pepsi's stock price
One day, there is no special news about Coke or Pepsi, but the price difference turns to 6 times.
Capture the chance, short the Coke and long the Pepsi.
If you bet correctly, which means Coke and Pepsi return to equilibrium of 5.5 times, congrats! Buy back Coke, Sell your Pepsi and keep the 0.5 times as your profit.
How to make an informed guess?
Congrats! Now you are touching the essence of Algorithmic Trading. While the methods are diverse, some useful domain knowledges are:
Pricing and Valuation
It is always important to have a rough idea about how your trading instrument is priced/valued. Such as the Black-Schole model.
Data Science
Some statistics tools to analyse the financial time-series data, such as Econometrics, Stochastic Process, or recent AI models.
Optimisation
To have an idea about managing your portfolio to achieve a maximized return.
Control Theory
How you should take your wealth movements and market impact into account for your investment decisions.
Last updated