The types of Automated Trading Algorithms
In recent times, the proliferation of flexible programming languages such as Python along with the advancement of data engineering and science methods have dramatically increased the popularity and accessibility of algorithmic trading methods, both in professional financial institutions and by retail enthusiasts.
In this article, I will try to outline and explain the different types of algorithms and their uses. I will delve into the different types in more detail in future articles, as well as publishing guides on how to get started in Python, so stay tuned!
What is algorithmic trading?
Before we go too deep, let’s settle on a definition of what is algorithmic trading, as the term can have multiple meanings depending on who you ask.
Going by Investopedia’s definition: Algorithmic trading is the use of process- and rules-based algorithms to employ strategies for executing trades. This can either be to optimally execute trading conditions (by reacting at speeds not possible for humans for example) or to initiate trades based on various quantitative strategies (e.g. pairs trading).
An algorithm is a set of instructions for solving a problem or accomplishing a task. Every computerized device uses algorithms to perform its functions.
So the term then is relatively open, encompassing anything that uses a set of defined rules to trade instruments, which starts to explain some of the confusion or interchangeability of the term.
Now we’ve got that out of the way, let’s focus on the different types:
Trade execution algorithms
When you hold a large block of security (say $1bn of AMZN stock) you’d be hard pushed to find a buyer on the other side of the transaction willing to purchase the whole block at market value or even close to it; the act of you selling so much of the stock changes the buy/sell dynamic of the stock and causes a significant drop in the market. To mitigate this, amongst other similar effects known as “slippage”, traders & institutions employ the use of trade execution algorithms.
Want to read this story later? Save it in Journal.
The basic idea is to systematically split a larger order into many smaller orders based on the available liquidity.
Three of the most commonly used trade execution algorithms are Time Weighted Average Price (TWAP), Volume Weighted Average Price (VWAP), and Percent of Value (PoV).
I will cover these types of algorithms in future articles.
Strategy implementation algorithms
These are what the majority of people envision and refer to when thinking of algorithmic trading.
Designed to process real-time (or batched) market data and create trading signals to be executed by trade execution algorithms. This may involve automatically rebalancing portfolios when certain pre-specified tolerance levels are exceeded, searching for arbitrage opportunities, creating trade signals based on news sentiment analysis, and producing trading signals from technical analysis.
These algorithms are mostly used by retail investors, hedge funds, asset managers, and prop traders.
I will cover the main types of strategies and the implementation of one using Backtrader and Python in the next article.
Stealth/gaming algorithms
If you can’t beat them, join them! These algorithms are designed to take advantage of the price movement caused when large positions are filled (usually through trade execution algorithms) and also to detect and outperform other algorithmic strategies.
Due to their nature of feeding off signals of larger and more prominent strategies, these algorithms are most effective at small capital sizes.
Another method which is now outdated would involve traders using liquidity detection algorithms to decipher whether there are large orders existing in a matching engine by sending out small orders to look for where large orders might be resting. When a small order is filled quickly, there is likely to be a large order behind it.
These algorithms are largely out of date now, given the advancement in trade execution times and the advancement of trade execution algorithms and dark pools.
Electronic market making
Traditionally, a market maker would provide liquidity in the market by both buying and selling assets, taking a slice off the top through the bid-ask spread.
Trading algorithms exist that provide the role market makers once played. This has evolved into what is known as Passive Rebate Arbitrage.
In large use by most investment banks and retail investment platforms (such as plus500 or Trading212, this is how they come up with their pricing) these algorithms allow for increased efficiency, reducing latency and spread margins.
Conclusion
The growth of algorithmic trading means it is now responsible for the majority of transactions on exchanges globally and is in use by some of the largest and most successful hedge-funds, including Renaissance Technologies and Two Sigma.
While written material and video content on this topic is continually growing, it still a topic that can be difficult to access and fully understand for beginners and even finance professionals, with a lot of people viewing it as a “black-box”.
Hopefully, this series can help you start (or brush up) on what is a fascinating and growing topic.
📝 Save this story in Journal.
👩💻 Wake up every Sunday morning to the week’s most noteworthy stories in Tech waiting in your inbox. Read the Noteworthy in Tech newsletter.
