The future of AI trading

The finance industry has one of the highest adoption rates of AI. With machine learning developing faster every day, speculation on what the future could bring is exciting those in the sector. Dive into what we think the future of trading could look like.

• What does AI trading look like today?
• What will the future look like?
• Trading robots and virtual assistants
• The use of data
• Strategic technical analysis
• Automated risk management
• Sentient trading
• The future of regulation and accountability
• Alternatives to AI with City Index

Before we look at the future of AI, it’s important to understand the current landscape and machine-learning capabilities.

What does AI trading look like today?

Today AI trading technology uses algorithms and machine-learning techniques to learn from vast amounts of historical data and market patterns, to execute trades in different markets.

These systems collect large amounts of financial data from various resources, transform this into a consistent format and extract features or indicators that are relevant for technical analysis.

Currently, the performance of these models is monitored by programmers and developers to improve accuracy and mitigate possible errors, as AI-driven engines are only powered by historical data. They are continuously adapting the systems as they can be inaccurate in situations of high market volatility.

It is important to consider that there are a few benefits and risks of AI trading. Examples of the benefits include faster analysis, the automation of certain trade actions or the possible mitigation of risks and human errors. While the risks include data bias, over-optimisation and ethical concerns such as accountability and lack of transparency. 

AI trading strategies

In the current trading climate, different types of trading strategies are carried out with the help of AI-driven systems:

Algorithmic trading is a type of strategy that uses computer programs to automatically execute trades based on precise instructions.

High-frequency trading (HFT) is a method of trading that uses powerful computer programs to conduct a large number of trades in milliseconds. It a type of algorithmic trading strategy that uses high speeds, turnover rates, and order-to-trade ratios to take advantage of very small, short-lived opportunities in the markets.

Machine-learning-based prediction is a strategy that uses tools that can learn from themselves, to make more accurate predictions about future market movements. It focuses on creating computer algorithms that can automatically improve their performance through experience.

Natural Language Processing-based sentiment is a type of strategy that uses natural language processing (NLP) techniques to analyse news articles, social media posts, and other forms of text data to measure the overall sentiment or attitude towards a particular instrument or market, identifying patterns within subjective material.

What will the future look like?

We know that AI is growing rapidly - at least half the firms surveyed by the consulting company McKinsey in 2022¹ were employing AI in some manner, compared to only 20% five years earlier - but how will it change the way we trade? Below we discuss some hypothetical future scenarios inspired from recent developments.

Trading robots and virtual assistants

There are already existing trading robots which allow you to program specific rules to buy or sell markets, automating orders and saving traders huge amounts of time.

A trading bot is a colloquial term for a software program that helps traders buy or sell instruments at a given point in time. A popular example is TrendSpider, which provides advanced technical analysis and enables you to automatically trigger an event when certain conditions from your strategy are met. These algorithms have already changed the way many traders plan their strategies.

In the future, trading bots might turn into independent AI virtual assistants that can help with real-time trading. Traders might even be able to talk to their virtual assistant just like they would to a colleague or friend, as it talks them through stock reports or forex trading examples in a personalised manner, according to the trader’s strategy or level of understanding.

Artificial intelligence assistants might even show users how to trade forex pairs based on the current levels of volatility, or suggest an index for a trader to short, for example.

The use of data

All traders might have access to large data sets if regulators ask big tech firms to start sharing market data to improve transparency, accountability and market integrity. This means these data sets may influence traders’ decision- making.

Technical indicators could also become advanced enough to factor fundamental analysis into their predictions. Therefore, virtual assistants might combine both technical and fundamental analysis in their suggestions, saving time for both traders and investors

The widespread use of big data might also lead to a concept such as ‘smart reports’ reports – for example, based on consumer trends and company spending. These self-updating reports might enable traders to see any information that could affect their portfolios instantly. It could change automatically without traders having to look at the data themselves.

Strategic analysis

As technology advances and market structures evolve, we can expect approaches such as HFT to develop and play a significant role in AI-driven trading strategies. This could mean that due to the speed of movement, most execution and analysis could be left to AI systems, and human traders could focus on strategic analysis – looking for ways to improve their strategy rather than using raw data or technical analysis tools.

In addition, strategies given by AI systems and virtual assistants could become completely personalised - new strategic indicators could suggest a style based on a trader’s preferred risk-reward ratio and previous trading trends. For example, If a trader has a history of day trading and going short, AI could suggest the best next step

Automated risk management

When it comes to risk management, striking the right balance between innovation and risk mitigation could be crucial in shaping the future of AI trading.

The key difference with the present might be that an AI-powered virtual assistant will take the profit and loss for a trader’s entire portfolio into consideration, looking at previous strategies and actions and adjusting individual trades depending on the desired risk profile.

For example, a trailing stop could not only move based on the market price, but it could also consider the current balance of other positions. Virtual assistants might know exactly what you are looking for and the risks associated with it, so they will plan your trades accordingly to mitigate the risks as much as possible.

Sentient trading

The ability to process and understand human language is becoming increasingly important in AI trading. Natural language processing (NLP) techniques, enable algorithms to extract valuable insights from news articles, social media, and other textual data sources.

By incorporating NLP capabilities into trading strategies, AI systems might be able to measure market sentiment and react accordingly, improving decision-making processes. AI algorithms might no longer base their strategies on objective market data, they might also analyse human emotions and reactions to predict price movements according to the types of emotionally based decisions companies or individual traders might make.

However, another possibility to take into consideration is: in a future scenario where everyone uses AI-driven systems and assistants to trade, will this remove the emotional factor from the picture? Or will the data used by each system provide some kind of subjectivity in their decision-making? Will biases still permeate the systems that move the markets?

The future of regulation and accountability

Nowadays, regulators are still figuring out how to deal with regulation and accountability in the context of AI-powered trades. Regulating new systems takes time - as we have seen with algorithmic and high-frequency trading - and AI develops at a very high speed.

In addition, the “black box problem” can create issues where not even developers may fully comprehend how they arrive at specific decisions. 

Nonetheless, we can imagine a future where AI could be held accountable for its trading decisions. AI might be judged as humans would, but according to different guidelines that consider the benefits and risks of machine learning. This might create demand for specialists in AI and AI regulation, for example.

Alternatives to AI with City Index

City Index offers two alternatives to AI: SMART signals and MetaTrader4.

How to trade with SMART signals

SMART signals are an algorithmic system which generates short-term trading signals. It is not a trading robot, but it is developed to run according to precise instructions, and each signal provides a suggested trade entry level and a suggested stop level based on historical price patterns

Follow these four steps to open your first position using SMART Signals :

1. Open your City Index account, or log in if you already have one
2. Navigate to the Signals tab on Web Trader, or tap ‘More’ and then ‘SMART Signals’ on mobile
3. View the list of the current trade ideas from the SMART Signals engine. You’ll see their previous win rate, details of the trade and how far the market is from the suggested entry
4. Choose a signal to trade. You can set up your own trade parameters, or go with those suggested by the signals engine.

How do I use MetaTrader 4?

With MetaTrader4, you can explore algorithmic trading strategies with Expert Advisors from FX Blue.

Follow these three steps to download MT4 to your desktop today:

1. Open your City Index MT4 account
2. We’ll send you an email with a link to download MT4
3. Click the link and follow the instructions

¹https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-in-2022-and-a-half-decade-in-review#talent