2iQ Partner Community News: causaLens Releases Causal Portfolio Optimization Model

featured image

2iQ’s insider transaction data is used across the financial services, investment management and FinTech industries to improve analysis, quantify investments and enhance other technologies.

One of our close partners — causaLens — recently released a short paper covering causal portfolio optimization.

causaLens identified a common problem with many popular portfolio optimization theories. Traditional approaches like Modern Portfolio Theory and contemporary machine learning approaches often fail in the real world because they are based on fragile and spurious historical correlations.

This is especially problematic for portfolio managers in the current moment. Global markets are characterized by high uncertainty, amid divergent post-pandemic recovery scenarios. And long-standing correlations between asset classes are shifting and collapsing.

causaLens leveraged Causal AI — a new kind of AI that reasons about causes and effects — to build a portfolio optimizer that is not fragile to spurious correlations and concept drift. Causal AI intelligently defines asset classes that are relatively causally isolated from each other. As a result, Causal AI outperforms both traditional and machine learning-enabled approaches to portfolio optimization in terms of risk-adjusted returns.

In the paper, causaLens reports the results of an experiment conducted over the 2019-2020 period, showing that Causal AI achieved a 23% greater Sharpe ratio than state-of-the-art machine learning optimization approaches.

Figure. Causal AI produces superior Sharpe ratios and lower volatility than alternative approaches.
“Hierarchical Risk Parity” (“HRP”) is a prominent optimization algorithm that uses conventional machine learning. Modern Portfolio Theory or “MPT” is the standard approach to portfolio optimization, while the “naive” approach is a heuristic approach that assigns equal weights to all securities in the portfolio.

Leading asset managers and hedge funds are already benefiting from Causal AI. Algorithmic trading firm Jump Trading says, “The novel Causal AI techniques available on the causaLens platform have facilitated our joint effort of discovering valuable, profitable trading strategies.” Read the causaLens report here.

Outcomes like these are why 2iQ leverages Causal AI to generate insights. More on the relationship between our two companies can be found here.

Recent Articles