Artificial intelligence is eminently an application for investors. Especially in a complex data-rich environment with a lot of uncertainty, artificial intelligence thrives.
Artificial intelligence was originally designed to simulate human intelligence using computers programmed to learn and think like humans. Various algorithms and techniques are used for this purpose. Currently, there are already numerous artificial intelligence applications, such as the virtual assistants Siri and Alexa. Netflix uses artificial intelligence in programming that makes people stay subscribers longer. Photos are classified with artificial intelligence on any iPhone or Mac. Google Maps uses artificial intelligence for 3D applications. Furthermore, it is remarkable how fast the adoption of Chat GPT is going, including concrete real-world applications.
The major advantage of an artificially intelligent machine compared to a human is that the machine is much faster at reading and analysing data than humans. Now, in the investment environment in particular, there is an extraordinary amount of data. It has been more than a decade since London traded in highly paid credit analysts for smart algorithms that could ‹en passant› also assess the creditworthiness of directors based on their LinkedIn network. You can bet that technology has not been idle in the meantime.
A big advantage for artificial intelligence is that investing is often a form of probability calculation. Absolute certainty about the future is scarce, then a good probability calculation based on all available data is worth a lot. Furthermore, artificial intelligence can combine structured data with unstructured data including media appearances and newspaper reports. Humans find it difficult to recognise a pattern amid all that noise, something artificial intelligence is actually very good at.
Investing based on quantitative factors
Before artificial intelligence was applied to investing, there were investment strategies that constructed a portfolio based solely on an algorithm. The most well-known of course is index investing and a variant of this is investing based on factors other than the weight in the index. In fact, with this variant of investing, there is already no human being making detailed decisions. A big advantage of these algorithms is that they are therefore not hindered by human biases. They can even take advantage of this to achieve greater returns. Furthermore, quantitative investors often arbitrage on small price deviations, with only high volumes and high frequencies paying off. Because computers can analyse data faster, they actually have inside information on things that many people do not experience until much later.
Investing based on sentiment
In the long term, the development of financial markets is determined by fundamental factors such as the economy, liquidity and, of course, valuation. In the short term, every news event seems to affect the stock market. This creates a lot of noise, making it difficult for human investors to see the wood for the trees. People then quickly switch to intuition and make decisions based on emotion, probably the only reason why technical analysis sometimes seems to work. Like a hunter on the savannah, man then often acts without thinking. Just read Kahnemann’s book, Thinking Fast & Slow.
An artificially intelligent investor, besides analysing the market, can also analyse how people will react to market developments. Especially predicting and recognising bubbles should be relatively easy for an artificially intelligent investor.
Help with portfolio construction
Many portfolios still consist of a combination of equities and bonds. These in turn are split into regions and sectors or categorised by creditworthiness and maturity. More and more investment opportunities are opening up, if only because of the strong development of private markets. An artificially intelligent investor can take into account different types of risk combined with other requirements and optimise the portfolio accordingly. Furthermore, portfolio managers regularly feel they are missing something in a portfolio, but don’t know what. This is where even a programme like ChatGPT can help.
AI for boring and mind-numbing work
Investing also involves a lot of boring and mind-numbing work. Checking portfolio for residual risks is usually subsumed under risk management. Often, that analysis is based on linear regression, while artificial intelligence is also good at non-linear relationships. Checking whether portfolios comply with the latest legal and regulatory insights is also not the most challenging function. In such tedious but complex environments, artificial intelligence can do an excellent job.Take further money laundering and terrorist financing control. That is where artificial intelligence can achieve spectacular results.
An artificially intelligent advisor
To answer the question of whether a computer can also give investment advice, we first need to define what exactly that is. Already, help desks and service staff appear to be being replaced by artificially intelligent machines at an ever-increasing rate. Where previously such computers still gave hilarious answers, today the only way to distinguish an artificially intelligent answer is the speed of answers. Computers are much faster than humans. With a slow answer, one is unlikely to be dealing with a machine.
At the same time, many advisers play an important role in terms of human confirmation that tells the investment client that he or she is doing the right thing. It will be difficult for a machine to match this human confirmation. This is probably why Robo-advisors soon found out that they also need human advisers. Probably the combination of human and machine is the ultimate winner here.
Han Dieperink is chief investment strategist at Auréus Asset Management. Earlier in his career, he was chief investment officer at Rabobank and Schretlen & Co. This column originally appeared in Dutch on InvestmentOfficer.nl.