The world is still reeling from ChatGPT’s implications: that there is a machine that for all intents and purposes, could emulate human sentience purely through language. ChatGPT is now used worldwide to generate project proposals, draft emails, write articles, social media posts, indeed all forms of written content, and generally make life easier in the home and in the workplace. The speed and accuracy with which AI can be trained to learn means that the new technology can also be utilised to make decisions based on given data – even financial. Neither the idea or the technology is new of course, using information technology to make financial decisions has been around since the earliest computers – though this is the first time we’re trying to get the machines to make reliable decisions on their own.
Using AI to make investment strategies over human decision-making offers many advantages. Not only can AI analyse vast amounts of data over a shorter period of time, it can also utilise machine learning technologies to accurately learn and extrapolate from past trends. AI can be taught to consider various types of data, without being limited by experience as most of us are – it can study scholarly articles written on market analysis, financial statements, news articles, and even social media content generated hour by hour, minute by minute, and second by second. This puts more information at the AI’s disposal to make informed decisions. The potential of AI and machine learning to inform investment strategies however still remains an untapped potential.
Machine learning and investment strategies
Machine learning is just one component of AI – one that elevates AI from being just a sophisticated computer. Machine learning enables AI to learn and adapt for better decision making without prior programming. Algorithmic innovations behind machine learning technology enables AI to synthesise data, learn from them, and continuously improve its own capabilities. Deep learning especially is a recent innovation that helps AI process large, unlabelled data sets into abstract forms, making extrapolation faster and more accurate. The technology is crucial in automating the process of strategy innovation. Machine learning’s growing prominence in investment management however, does not mean that humans have been replaced in selecting the framework for said investment strategy or managing risk just yet.
Another important aspect of AI and machine learning in relation to investment planning is what AI cannot learn. AI naturally eliminates human bias in investment-making decisions. Human investors on the other hand can be prompted by emotion, prejudice, or even superstition to make irrational decisions that may well prove to be disastrous. AI on the other hand bases its analysis solely on the data available, and can objectively process even information such as public sentiment without being affected by it.
How AI and machine learning can be utilised in forming investment strategies
Risk assessment
The vast amount of data that AI is capable of processing allows it to take more business variables into consideration in assessing the market for risk. Its capabilities also allow AI to implement various statistical formulae to this data with increased speed and accuracy to better assess the values of various business scenarios. This helps investors optimise their portfolios better, based on their risk appetite.
Business forecasting
Machine learning allows AI to make business forecasts by extrapolating from historical data. These forecasts, or predictions will help investors decide on the optimal times to buy, sell, and to hold on to their investments. The forecasts will also help set investment strategy moving forward.
Is AI and machine learning all it’s said to be?
The power of AI can certainly be a powerful ally to have when making investment decisions. From reviewing investment vehicles to giving customised advice, the power of AI can be an invaluable ally to have in a competitive market. However, it is important to remember that any and all markets are extremely volatile contexts, as any context with a human element involved tends to be. While AI may be far better at utilising the best formulae with the best accuracy rates, human investors or investment advisors can still have an edge over machine learning. Lived experience, gut feeling, or institution, there is still an unknown factor to decision making that AI cannot yet fully grasp.
There are also some dangers to basing investment decisions solely off of the capabilities of AI as well. Leaving human intuition and judgement aside to rely solely on pattern-based (algorithmic) thinking is one of the most critical. AI also depends on historical data for the predictions it makes, so there is a very real possibility that this over-reliance leads it to miss out on unforeseen events for which it has no reference points to fall back on.
Accountability is also an ethical debate that is frequently bought up in relation to the role AI plays in investment-related decision making. The objectivity with which AI makes its decision can call into doubt whether it is suitable to place it in charge of the hard-earned wealth of others. This is because emotion-based decisions enable humans to capture a dimension of human endeavour that AI are not equipped to perceive. In the event of a mistake, it is also not possible to hold an AI accountable in a court of law, leaving all investors with little legal recourse.
There are also other factors muddying the waters when it comes to evaluating whether AI is a viable tool to utilise in setting investment strategy. For one there is the dubious study carried out by the German publication WirtschaftsWoche, where a dog picked the more lucrative investments over its ‘opponent’, ChatGPT. ChatGPT is hardly an AI designed for an investment advisory role, but it is impossible to argue that it will not be the first one that most would turn to.
There’s always been debate on whether luck plays a role in investment decisions such as picking stocks on the stock market. In experiments where students and even animals were tasked with choosing the better stock option by throwing darts at a wall, it was found that the ones that relied on pure chance performed better than ones chosen by the best financial analysis. Were AI to be involved, it is possible that it would iron out the interference of Lady Luck altogether.
When it comes to investment strategies at least, AI and machine learning still has to be viewed as a tool that complements the strengths of humans. The combination allows investors to make use of AI’s capabilities while retaining the unique addition of human experience and intuition. The objectivity of AI means that it also helps eliminate human bias from investment-related decision making. The use of machine learning enables investors to ‘outsource’ the dog work of market analysis to better focus on evaluating the information that AI provides. AI and its machine learning capabilities would effectively maximise investment outcomes while minimising risks as much as possible.
The future of AI-assisted business is bright and full of potential , though it’s not quite capable of taking the driving seat just yet.
(Theruni Liyanage)