by David Itzkovits
Risk management is a critical component in the investment management field. An investment manager views risk as volatility, or the standard deviation of investment returns. But clients more frequently interpret risk as the fear of losing money.
We not only face the risk of an asset value going down, but the risk of it staying down for long periods of time and not getting back into the markets quickly enough.
AI is a building block for solutions that react as quickly as required to today’s fast-moving markets.
The Sanlam Managed Risk SMR (UCITS) Fund is an example of an AI solution. It is a five-year-old international fund, not easily accessible to South Africa-based retail investors, that is 100% driven by AI.
We invest $90 of every 100 into an MCSI World index tracker and use the remaining $10 to take up short future positions, which protect investors against a fall in the underlying MCSI (in other words any fall in share prices). The AI ‘engine’ varies the number of short futures ‘held’ at any time to increase or decrease the ‘net’ equity exposure of the SMR fund to between 10% and 90%.
What the AI is doing on a weekly basis is deciding how many short futures it must enter into to reduce the net equity exposure in the fund. It moves in and out of the market aggressively, but without emotion. And it works. In the first four months of 2018 the AI-driven SMR advanced by almost 6% compared to the 0.15% growth in the MSCI World.
Humankind is riding the wave of a tech revolution that started with the development of the personal computer (1980s) and gathered momentum on the back of the Internet (1990s) and smartphones (2000s). The next stage of this revolution is being driven by artificial intelligence (AI), which already affects every aspect of our daily lives.
The highest level of AI would be for a computer to copy and mimic the conscious human mind, but there is a second level – referred to as narrow or specialised AI – where a computer performs a single task better and more efficiently than humans can. While we were some way from achieving the highest level of AI the second level was already upon us.
The Google search engine, the weather application on your mobile smartphone, the Skype communication service and the Tesla electric car are all examples of specialised AI in action. The latest version of Skype allows you to communicate in eight spoken languages and 50 written languages in real time while Tesla showcases machine learning in action.
You have a car with hundreds of cameras taking in around 1.8 million data points every second. This data is continuously analysed to decide one of three things: Should I brake, accelerate or turn and what combination of the three.
There are countless examples of the application of AI in the financial services environment. US-based bank, Wells Fargo, utilises a combination of its own AI-powered chatbot alongside Facebook Messenger to respond to natural language messages from its customers. Financial services firm JP Morgan, which will spend more than $10 billion on AI development this year, is using the technology to analyse legal contracts, among other applications.
Financial institutions – especially banks – are quite vocal about how they are using AI to gain competitive advantages, improve efficiencies and reduce costs, but we seldom hear about the application of this technology from investment managers. The use of AI in investment management is still incredibly secretive because institutions know that if they crack the code with AI they create the potential for better returns and lower risk.
Neither investment managers nor their clients need fear AI. AI is a diversifier to human risk management, it is a tool that people in the front office can leverage.
AI will invest with better performance and free up your time, allowing you to make better connections with your clients. It is not man versus machine, but man with machine versus man without.
- David Itzkovits, CEO of Sanlam Global Investment Solutions