Contents of this article
With advances in science and technology, many areas of our lives are now carried out or supported by machines and modern technology. Finance, and investment finance in particular, is no exception. Artificial intelligence (AI), machine learning and other technologies are being applied more and more, most notably in stock trading robots, or simply trading bots, which help investors improve how they trade.
What is a stock trading robot?
Contrary to what many people imagine when they hear the word robot, stock trading robots do not have a physical form like the ones we usually picture, such as Asimo, the cute robot that is a source of pride for the Japanese, or more recently Sophia, the world’s first robot citizen, with a woman’s face.
A stock trading robot is a tool that can execute trades (place orders) automatically, based on a set of rules defined by its developer. These rules are built on input data such as time, price and volume, or on any mathematical or quantitative model.
What is a stock trading robot?
According to Investopedia, an algorithmic trading robot is a computer program that can generate and execute buy and sell signals in the market. A trading robot usually includes entry rules that signal when to buy or sell, exit rules that indicate when to close the current position, and rules that determine trade size (position sizing rules).
Key features of stock trading robots
Building an algorithmic trading strategy
One of the first steps in building and developing a trading algorithm is to capture the core features that every algorithmic trading strategy needs. The strategy must also be grounded in observation of market conditions and the economic context, and the mathematical model used to develop it must rest on sound statistical methods.
Next, we need to define the inputs for the robot’s analysis. For a trading strategy to run automatically, the robot needs to identify persistent market inefficiencies. In economic theory, an inefficient market is one in which asset prices do not accurately reflect their true value for some reason, leading to a loss of social welfare. In practice, most markets show some degree of inefficiency and, in extreme cases, an inefficient market can be an example of market failure.
Algorithmic trading strategies that follow a strict set of rules exploit the “behaviours” of the market. For a behaviour to form, it has to repeat many times. The algorithm tracks such behaviours closely so that the programmer can build their own strategy. Moreover, without identifying the cause of a market inefficiency, there is no way to tell whether a strategy’s success or failure is down to chance.
Today, trading strategies for robots are usually built by exploiting an edge in areas such as macro news (changes in interest rates, taxes, etc.), fundamental analysis (revenue, profit, etc.), statistical analysis (correlation, variance, etc.) and technical analysis (moving averages, the RSI, etc.)…
Backtesting and optimization
Backtesting focuses on verifying whether the trading robot is doing what you want, and on understanding how the strategy performs across different timeframes, asset classes and market conditions, especially during “black swan” events such as the 2007–2008 financial crisis or the recent Covid-19 pandemic.
In addition, once you have a trading robot, you will want to maximize its performance while reducing bias. To maximize performance, you first need to choose a good performance metric that captures both risk and reward (for example, the Sharpe ratio).
To maximize performance, you first need to choose a good performance metric that captures both risk and reward (for example, the Sharpe ratio).
Overfitting, meanwhile, occurs when your robot relies too heavily on past data, creating the illusion of strong performance; because the future never unfolds exactly like the past, the plan can then fail. Training the robot on more datasets, removing irrelevant input features and simplifying the model can address this problem.
Pros and cons of stock trading robots
As in any other field that applies information technology and machines, using robots for stock trading has the following pros and cons:
Pros
A stock trading robot executes trades at the best price, at the right moment and instantly, avoiding the risk that the price moves in the meantime.
Unlike humans, who are prone to errors of their own making (human errors) in any manual task, a stock trading robot places orders accurately, with fewer input mistakes and none of the delays that come from slow human execution. For example, when you want to buy a stock at price X, you could easily type price Y by mistake without noticing, and when the order fills at Y, it hurts your investment. With a stock trading robot, trades happen automatically, with a far lower chance of error than with a human.
Trading robots also remove emotion and psychology from the trading process. Psychology has a huge influence on every investment decision, especially for new “F0” investors. They are easily swayed by other people’s opinions and by the flood of information in the market. As newcomers who lack experience, F0 investors’ psychology is easily shaken, their emotions take over, and they cannot stick to their investment discipline.
Trading robots, on the other hand, have no emotions; they follow the trading rules set by the developer from the outset, so their trades are more objective and effective.
Cons
Alongside the pros, trading with robot systems also has some drawbacks:
The input data for stock trading robots’ analysis is usually price and liquidity, and rarely includes a company’s financials, the factors that create its intrinsic value. As a result, if the market is being manipulated, a robot may well make inaccurate trading decisions or issue buy signals on poor-quality stocks.
In addition, trading robots are only truly effective when a stock has a clear price trend. When a stock is moving sideways, these robots may issue a buy signal and then immediately a sell signal, flip-flopping constantly and forcing investors to switch positions quickly, which takes a heavy toll on their psychology.
Finally, because of the features described above, these robots run on the original algorithm built by their programmer, so investors cannot intervene in the system to find stocks of their own choosing.
In summary, a stock trading robot can be a useful tool for investors if the investment strategy has been built carefully and effectively. Whether to use a robot for stock trading depends entirely on your own capabilities and needs, so every investor should research thoroughly before deciding whether to use this tool.
Happy Trading!
The “Stock Trading For A Living” course
Green Chart designed this course to give you a solid grounding in fundamental analysis & stock valuation, and to show you how to integrate it effectively into your existing trading system, especially if you already trade with technical analysis. The programme includes 5 theory sessions and 6 months of hands-on trading guidance in the live market with the lecturer.
Find out more about the course here.
