We have heard plenty in the news about active fund managers underperforming which has led many investors to flock towards more passive investments such as ETFs. But another type of investment we have noticed is starting to gain some serious traction is algorithmic funds. Simply put algorithmic funds use an array of technical methods including AI and machine learning to process large quantities of both live and historical trading data to find trends and patterns which forms the basis for investment. Renaissance Technologies and Two Sigma currently each have $US45 billion assets under management and are tied as the second largest hedge fund firms in the US. Both these firms have increased rapidly in the last decade with Two Sigma with just $US5 billion in assets back in 2010 now has 9 times that amount. Meanwhile, Renaissance Technologies has made a staggering jump from $US27 billion at the start of 2016 to almost double that amount now. These high inflows of capital have come around a time when hedge funds are typically struggling and investors are more confident putting their money behind super computers and mathematical models.

Established in 2001, 2 Sigma places a significance on machine learning in which their technology scanning over 10,000 sources of public and proprietary data to not only analyse data but to perpetually learn from patterns and trends in the data. They have CPUs with over 1,695 terabytes of memory and can perform 1×1014 calculations per second. The firm invests large portions of its capital back into improving its technology. Amazingly, Two Sigma claims 72% of their employees have no financial background as they preference those with mathematical and science backgrounds stating that is what ‘helps them see what others in the industry don’t.’ Renaissance on the other hand has become famously known for being exceptionally secretive when it comes to their quantitative methods with many referring to them as the ‘black box’ of finance. The firm is run by 90 PhDs in either mathematics or science. One of the keys to the way Renaissance looks to invest is by looking at short-term movements with greater confidence than long term investing. By looking at observations and occurrences in a very short period of time the quantitative technology is more likely to identify patterns to predict short term trend. The firm believes they are more confident to pick short term movements other investors would consider to be randomness rather than investing in the long-term capital growth which introduces a multitude of factors and observations to be analysed.

The perception is rapidly becoming as technology continues to advance such algorithms will begin to outperform conventional hedge funds if they haven’t already. Why wouldn’t you invest in a sophisticated super computer that can make calculations trillions of times faster than the human brain? Why risk investing with a human fund manager that is subject to human error and emotional decisions? The short answer is that these algorithms lack one crucial element, common sense. The strides this technology has made is nothing short of remarkable and the results have shown. However, it is not without its own limitations. The first issue is that these algorithms and computer systems are purely based on past data and are unlikely to be able to predict outcomes that are very rare or have not occurred before. The other main issue relates to the data being analysed also known as Garbage In Garbage Out (GIGO). No matter the omnipotent power of the computer system being used, if the data input is of poor quality so too will the analysis and outcome.

The movement towards automated managed funds is becoming closer and closer to reality with many theorising that it will cost hundreds of thousands of jobs worldwide. In our opinion, while the number of algorithmic funds is set to increase as technology expands, there will always be a place for human managed funds. The skill and expertise of fund managers should not be undermined even when competing against a computer. The most successful funds in the future are likely to be those that are able to find the perfect balance of using technology such as machine learning coupled with the common sense and experience of fund managers.