According to recent statistics, the global value of the asset management industry stands at just over $503.3 billion dollars (1). Asset and portfolio oversight have always represented a combination of art and science. They require expertise, intuition and the ability to grasp market fundamentals in order to remain ahead of the curve. However, an increasingly interconnected fiscal ecosystem has challenged many traditional methodologies. This is why a growing number of successful investors are leveraging the benefits associated with AI-driven portfolio optimization.
What does this concept entail, what are some of its core tenets and what benefits can it provide? Before we address these questions, it's a good idea to review the basic principles of portfolio management.
Portfolio Management: Appreciating the Fundamentals
Simply stated, overseeing a lucrative portfolio involves choosing the right investments at the appropriate times. This could involve active (short-term) management or passive oversight intended to provide long-term returns. Either way, some of the basics include (2):
- Alignment with specific goals.
- Governance and compliance.
- The ability to make the correct strategic investment decisions.
- Buying and selling when warranted.
- Risk mitigation.
- Diversification.
Having said this, automation has become key within today's financial marketplace. The liquidity associated with specific sectors such as cryptocurrencies and ETFs are two common examples. On-the-fly decisions may need to be made at a moment's notice while the collation of big data has become commonplace. Unfortunately, standard approaches to asset management may no longer be sufficient.
This is when the power of AI asset management becomes relevant. This approach involves the use of artificial intelligence in order to predict price fluctuations, to analyze the performance of specific holdings and to optimize the overall return on investment (ROI). While employing AI for portfolio construction may appear somewhat complicated at first glance, it can be better understood by highlighting a handful of key tenets.
The Relationship Between AI Asset Management and Genetic Algorithms
A genetic algorithm is a term used to describe the ways in which artificial intelligence finds optimized solutions related to predetermined problems (such as determining the appropriate investment). As the phrase suggests, these algorithms use models based upon seemingly unrelated concepts such as evolutionary biology and natural selection (3).
The main takeaway point here is that genetic algorithms essentially employ the principle of natural selection in order to determine the "fittest" investment opportunities at any given time (such as a high rate of return alongside relatively low volatility).
So, how can genetic algorithms be directly applied to portfolio management? These approaches embrace the notion of "survival of the fittest" to identify which assets or asset classes should provide the highest rate of return that is aligned with the overall intentions of the portfolio. While the algorithms are highly technical in nature, their reliance upon the principles originally proposed by Charles Darwin represent their core foundation.
Variances and Covariances
Before delving into this next topic, it's wise to briefly review the principles of variance and covariance in relation to asset management (4):
- Variance represents the distribution of data around its mean (average) value. In finance, this is often used to gauge the volatility of a specific asset.
- Covariance instead involves comparing the ROI of two individual investments over time. Positive covariance indicates that the assets will simultaneously increase or decrease in value. Negative covariance signifies that their values tend to move in opposite directions (such as the relationship between yields and bond prices).
Variance and covariance also play crucial roles when used in tandem with AI asset management. Calculating these parameters would often require a significant knowledge of statistics and the application of "brute force" models in the past.
Not only would this take a great deal of time, but mistakes could (and did) often prove costly. The use of AI algorithms aims to take the guesswork out of the equation while providing stakeholders with user-friendly results that can be easily interpreted.
Relevant Case Studies in AI Portfolio Construction
Artificial intelligence has been employed within the financial sector since the early 1980s (5). However, this concept is still relatively new in relation to the world of portfolio management. Let's therefore examine a handful of relevant case studies to highlight its growing presence.
AI-empowered hedge funds
Hedge funds are often embedded within portfolios as a means to offset open-market volatility. In this case, AI can be used to accurately predict how the value of a certain asset will be impacted by future events. For example, more stringent carbon regulations could lead to a rise in the value of lithium-ion technology and the further growth of green hydrogen batteries (6). Investors can therefore make strategic decisions well ahead of the curve; helping to cement long-term returns.
R.E. Lee Capital
R.E. Lee Capital is a financial firm primarily associated with supervising the portfolios of high-net-worth individuals. Their discretionary use of AI portfolio construction helps to simplify data-driven analyses while ensuring that bias does not affect decision-making policies. As stated by Investment Director Bao Vu in a recent interview, R.E. Lee Capital now uses AI and machine learning to compliment existing management strategies while ensuring higher rates of return (7).
American Century Investments
American Century Investments has been providing bespoke portfolio management solutions for nearly five decades and they have recently included AI within their arsenal of tactics. Through a combination of genetic algorithms and price analyses, American Century Investments is now able to generate lucrative investment decisions and to more accurately estimate future asset values (7). These variables are then incorporated into their overall approach to portfolio management with the intention of streamlining in-house processes (such as sales and customer relations).
OPTrust
OPTrust is a Canadian firm primarily associated with pension fund management. In this case, their team deployed artificial intelligence to obtain greater insight in regard to non-linear factors that may impact the value of an underlying asset such as movements within seemingly unrelated markets (8). OPTrust also uses AI portfolio optimization in relation to risk management; employing price data in order to better appreciate how past and present market conditions may lead to increased volatility and therefore, a higher degree of perceived risk.
The Tip of the (Digital) Iceberg
Please keep in mind that this is only a basic overview of how AI is transforming the portfolio management industry. We encourage those who are curious to learn more to subscribe to our weekly newsletter. After all, it pays to work smart as opposed to hard.
Discord
Telegram
Twitter