According to the Eurekahedge index, hedge funds that utilize machine learning and artificial intelligence in their trading process recorded the worst month ever in February. In two years, the first equity correction toppled their strategies while cross-asset correlations changed. Eurekahedge has been tracking the hedge fund industry since 2011.
Although computerized programs are often feared for their potential to phase out human traders by taking over their jobs, the artificial intelligence (AI) quants recently trailed behind their counterparts.
In fact, the AI index declined 7.3 percent in February in comparison to a 2.4 percent drop for the more extensive Hedge Fund Research index. Even so, the decline even exceeded a more conventional category of commodity trading advisers and quantitative analysts, which presented near-record losses while the equity reversal outdid the automated trend – following plans.
The magnitude to which quant funds can worsen selloffs has elicited heated debates whereby some managers even argue that they are too small to trigger such as an impact. However, according to JPMorgan Chase &Co., the situation experienced in February was an exception considering the bank’s torrid performance in recent times.
MORE – 10 Applications of Machine Learning in Finance
MORE – Pioneer AI Hedge Fund – DE Shaw
MORE – RPA – 10 Powerful Examples in Enterprise
Strategists at JPMorgan Chase & Co., under the leadership of Nikolas Panigirtzoglou, gave their analysis of last month’s situation.
They think that AI funds similar to commodity trading advisors may have played a considerable role in last month’s correction in response to being forced to eliminate risk, bearing in mind the extraordinary 7.3 percent decline experienced in February. In addition, they indicated in their Friday note that adoption rates have gone up, causing AI strategies to become increasingly crowded.
According to Nigol Koulajian of Quest Partners, the chief investment officer of a $1.4 billion quant fund, strategies made for one-direction based markets may have misled hedge fund managers.
In the same interview, he also told Bloomberg News that practitioners, on the other hand, likely turned complacent upon optimizing models that were intended to restore order or stabilize the bull market. In turn, this undertaking led to the creation of strategies that are not ideal for market shifts.
With all that said, the Eurekahedge index remains only but a mere representation of the hedge fund industry. Since both machine learning and artificial intelligence technologies are broad categories, the funds may require using considerably different methods.
In this case, some of the methods may be drawn from traditional statistics that can evaluate more sophisticated data sets. The other techniques, on the other hand, such as deep learning can break down data by using multiple layers of analysis. According to the theory, such layers are similar to the functions of the human brain.
Despite the situation that happened in February, there is uncertain evidence that AI funds, currently being tracked by the Eurekahedge index, are anchored to enjoy the trend of growing markets. Eurekahedge tracks approximately 15 hedge funds presently. Aside from that, JPMorgan Chase & Co. pointed out that artificial intelligence (AI) funds have become more correlated, especially to trend-following commodity trading advisors or CTAs in the last 12 months.
MORE – Top 50 RPA Tools – A Comprehensive Guide
Source Bloomberg