I analyze different approaches to creating bots and look at their effectiveness. To answer the question of whether bots are effective, I’ve spent a lot of time researching in the field of neural networks, algorithmic trading, and crypto trading, and can now offer conclusions for further study and practical application. In this blog, you’ll learn about the most important ridges and burnt spots of trading bots.
Let’s start at the beginning. Futures trading bot is not new, must confess, there have been numerous books written about them, and they are utilized by financial institutions and banks. But do they truly assist? The notion of algorithmic trading has generated a lot of debate, with several studies and reflections on the subject, but no clear conclusion.
The majority of these bots were written by either trader with some ideas and thoughts or programmers without trading expertise. I’ll do my best to explain this topic. As a developer, as well as a trader. As someone who has lost money on trading but gained a lot of money.
You may construct your own bot in a few hours if you are a software developer or understand the fundamentals of writing scripts. However, will it be profitable? Will there be enough money coming in each month? Will he ever earn $100,000 per year? I’ll answer these issues and provide you with some pointers on how to continue in this post.
What is a trading bot?
A trading bot is an algorithm that transforms market conditions into decisions on transactions (usually buy, sell or hold).
Trading bots type:
All types that any trader could take since a bot is an algorithm written by programmers, they can put any logic into it.
According to the type of traders and types of trading, the following options can be distinguished:
- Long-term traders (long-time traders) — investors;
- Swing traders – those who place trades for a week, month or year;
- Day traders — those who place a small number of trades during the day without moving them overnight;
- Scalpers are those who put up a lot of trades per day, an hour, or even a minute.
All these types of traders can be implemented in the form of a bot.
What are the types of bot implementations?
Simply stated, this is the meaning of algorithmic trading. Because the bot is a full software that automates the trader’s trading strategy in automated mode.
- Neural network or artificial intelligence — there can be simple bots with one perceptron based on several neurons, complex LSTM networks, or even news analysis based on keyword heuristics, and semantic search with NLP.
- Quantitative trading — based on a strategy that combines any criteria for making a decision, it can be a comparison of indicators, price behavior, pattern search, etc.
- Semi-automated bot, alerts (semi-automated, alert bot) — will use some algorithms to tell traders what to do. This type includes indicators based on a strategy.
- Genetic algorithms — it could be attributed to a part of machine learning/neural networks, but in fact, it is not sufficiently studied to define this type of algorithm as an approach to machine learning. Implementations of genetic algorithms vary, are being researched by universities around the world, and are part of a broader topic beyond the scope of this publication.
So, having answered four fundamental questions that will already give you a basic idea of how to write your bot, I’m ready to start. But what’s next? We can consider calculating a bot’s KPIs now that we know what it is.
Measuring the bot’s quality
To measure the quality of the bot, we can use the technique of backtesting.
Reverse testing is the process of generating trading signals based on previous data that has now been removed from the market. It includes both simulated markets (so-called side cases, such as a crisis, steep declines, or market growth) and actual markets.
Let’s take some heuristics to work with, such as the amount of cash available, the time duration it operates, stop-losses, and profit targets. We may employ additional heuristics based on any trading strategy we choose (for example, neural network bots can utilize market conditions to determine their parameters).
The best way to determine the quality of a strategy is to write its prototype using TradingView. Using their documentation, you can easily write a strategy in Pine Script (which has a very simple syntax).
As you can see in the picture above, I implemented a simple strategy based on a long order after each green candle.
However, despite the exponential growth of the account in TradingView, this strategy is not applicable on any of the exchanges, since it uses a stop loss of 0.01% and a profit of 0.5%. At this percentage, you can automate trading and even place a stop-limit order in its range, but you will not be able to survive with the commissions offered by the broker.
We may develop our bots’ strategies for TradingView faster before using the actual market or a demo trading environment by utilizing sufficiently extensive analytics from TradingView. Particularly if we intend to build a server bot with a complex structure and a simple user interface.
After you see something like this:
You may consider using a real trading bot if you have a reasonable net profit combined with a percent profit over 60 (at least). Is this to say that bots are helpful? Can we answer this question right now? Spoiler alert: Yes. They are effective. But let’s look at it closer.
Do trading bots help you make money?
Yes, of course. That’s why many hedge funds, banking structures, and large financial companies hire specialists in machine learning and algorithms. These people are responsible for the introduction of automated trading bots for trading in large markets with big money.
Trading bots can help you trade or invest more efficiently, but they should be carefully planned and thoroughly tested before being used on real money. I’m currently finalizing my own day trading bot, and I also maintain a bot that notifies me about potential purchases/sales of investment assets over long distances (1-2 transactions per month). This
Make sure you’re familiar with the technologies you employ to build algorithmic trading. I’d also like to caution against purchasing pre-made bots that are available at a premium (many of them operate on models for only 1-2 weeks, after which they are discarded).
Later, I’ll write several articles on how to utilize algorithmic bots effectively, how to create effective strategies in Pine Script, and so forth. Continue reading to see what I’ve been up to lately!