Everyone from Main Street to Wall Street has been scrambling to deploy AI since ChatGPT dropped in late 2022.
By one estimate, 2023 saw AI-related sales jump 29% to hit a whopping $166 billion. By 2027, we will be staring at a colossal $400 billion market. And the financial sector will be responsible for a quarter of that spending, outpacing five other major industries.
It is easy to see why. From algorithm-driven trading systems that can parse vast real-time data to predictive analytics that forecast market trends, the pot of gold at the end of the AI rainbow is certainly enticing.
Here are three emerging trends we’re watching.
1. Sentiment Analysis
Much of the interest in finance centres on using natural language processing (NLP) to analyse sentiment. This process typically leverages FinLLMs – LLMs trained on a combination of general and domain-specific (financial) corpuses then finetuned with prompt engineering.
The aim is to analyse input text, such as news articles, policy reports, or earnings calls, to gauge sentiment – be it public sentiment towards a stock or the market, the sentiment of central bankers towards policymaking, or the sentiment of business leaders towards the economy.
In the central bank arena, the Federal Reserve has been the focus of sentiment analysis NLPs for a while. For example, the FedNLP dataset comprises documents sourced from various FOMC materials, which have been annotated with sentiment labels based on the Fed’s rate decision for the subsequent period.
At Macro Hive, we are developing an AI model that can understand central bank speeches. To finetune LLMs for hawkish-dovish classification, we have our domain-expert economists label text data from central bank communications to create datasets that clients can interact with. But rather than just focusing on the Fed, we are scaling this solution to cover every central bank in the world.
2. Asset Movement Prediction
AI is bringing high-level computational power to the analysis of technical indicators. These AI-driven algorithms delve into the complexities of various indicators like the exponential moving average (EMA), relative strength index (RSI), Bollinger Bands, Fibonacci retracement, stochastic oscillator, and average directional index. By automating the scrutiny of these technical charts, AI identifies trading opportunities with a precision previously unattainable using manual analysis.
For instance, an AI system analysing EMA trends can spot emerging patterns that suggest a stock’s upward or downward trajectory, while an RSI-focused algorithm might determine whether a stock is overbought or oversold, suggesting a potential reversal. AI models can generate these insights at a speed and accuracy that outstrips human capabilities, enabling traders to make informed decisions swiftly and stay ahead of market movements.
As these technologies mature, the integration of AI into financial strategies is becoming a critical tool for investors seeking to optimize their portfolios and capitalize on market inefficiencies.
The proof?
Research from Institutional Investor found AI-led hedge funds massively outperform traditional competitors, producing cumulative returns of 34% in three years.
3. Hyper-Personalised Insights
Investors face two fundamental problems: too little time, too much noisy information. They struggle to make decisions due to information overload, constantly changing market relationships and rapidly developing regimes.
And things are only getting worse.
That creates a massive opportunity for producing tools that generate hyper-personalised insights for clients. Typically conceptualised as a chatbot, these products can understand and respond to user inquiries in a conversational manner, providing real-time, customized financial guidance.
Critically, they are not only responsive to market conditions but also aligned with individual investment profiles. For instance, a chatbot can analyse an investor’s portfolio preferences, risk tolerance, and past financial behaviour to offer tailored advice on stock purchases, sales, and portfolio diversification.
At Macro Hive, we are developing a GPT that can extract only the most relevant information for the client’s persona from our broader research output, providing hyper-personalised, direct alpha to clients via signals that improve their portfolio returns.
For example, clients will be able to ask which markets are most correlated to their market, which quant signals are giving buy signals today, or what our latest view on the Fed is – all using a simple chat interface.
Where Next?
AI’s role in sentiment analysis, asset movement prediction, and hyper-personalized investment advice is the forefront of modern financial services. At Macro Hive, we’re leveraging these technologies to dissect central bank communications and automate complex trading indicators, setting the stage for a new era where financial insights are instantaneous and strategies are finely tuned. Our tools convert overwhelming data into actionable intelligence, empowering clients to outpace the fast-moving market trends. To find out more, book a demo now.
Matthew Tibble is Commissioning Editor at Macro Hive. He has worked as an editorial consultant and freelance editor for companies such as RiskThinking.AI, JDI Research, and FutureScape248.