Finding the balance between risk and reward
by Aaron Cirksena
Aaron Cirksena, Founder and CEO of MDRN Capital, is a 2011 graduate of the University of Maryland, College Park, where he studied economics. Since then, he has devoted his entire career to financial planning, distribution planning, and managing client money. He first worked with multiple $1 billion teams at Morgan Stanley and independent firms and eventually created his own independent services firm in MDRN Capital, which is revolutionizing retirement planning by offering a comprehensive range of services, including income planning, investment management, tax planning, healthcare, and estate planning, all with a greater degree of effectiveness compared to traditional providers. As a fully digital firm, MDRN prioritizes efficiency and convenience by providing remote consultations and opening a digital account.On the surface, it would appear that artificial intelligence and investing are the perfect match. AI excels at data analysis, providing the capability to crunch numbers from a wide range of sources and provide insights in seconds. AI also excels at automation, meaning it can execute trades for investors based on predefined conditions at high speeds and frequencies. Through predictive modeling, AI can give investors a forecast of the movements markets are most likely to take.
With AI offering investors tools such as these, it’s hard not to see AI’s integration into the investment space as an overall boon. In fact, AI has already made significant inroads in investing, as it is now being used to develop strategies, power a new generation of robo-advisors, and provide a foundation for algorithm-driven funds.
However, those who look below the surface will see the potential downsides of AI-powered investing, not the least of which is algorithmic bias. Additionally, because AI’s inherent properties make it difficult to trace its decision-making process, those who rely on it often must accept an answer without an explanation, which means investors who partner with it must usually be willing to move ahead with blind faith.
Nevertheless, AI’s shortcomings don’t necessarily mean it should be passed over by investors. When used responsibly, it can amplify human efforts.
The following are some thoughts on how a judicious balance can be achieved.
Automation Isn’t Always The Smartest Play
It’s hard to argue against the benefits automation brings to the investment process. Using pre-set rules and conditions, AI can automate the trading process to ensure opportunities to secure optimal entry and exit prices aren’t missed.
AI can also automate data analytics to rapidly provide investors with insights that allow them to gain an advantage over those relying on manual assessments. AI’s speed lets investors incorporate more data sources in their analysis, increasing the chances of identifying hidden trends and risks. Overall, it makes it much easier for complex data to be transformed into actionable insights.
AI-driven automation also helps with accuracy. Manual processes present risks when it comes to precision, but AI guards against the costly errors that can crop up when figures are misread, or entries are miskeyed.
Still, automation isn’t always the best approach. As advisors know, investing involves a lot of nuance and context that can be challenging for AI to incorporate.
One issue is that AI commonly relies on quantitative data and historical patterns rather than qualitative factors that can affect investing. For example, a brand crisis emerging in media channels may not make an AI platform’s radar, despite its potential to impact investor activity. AI is also incapable of comprehending the impact of black swan events and responding to them appropriately.
Algorithmic biases are another factor that adds risk to over-reliance on AI automations. Biases built on past market inefficiencies or other factors can creep into AI training, impacting models and the actions they take.
The risk of AI biases was a top concern voiced by investors in a recent Morningstar survey. Investors are aware that any planning that taps into the power of AI could cause biases to surface. Advisors who can’t ensure AI-driven insights are bias-free may have trouble maintaining their clients’ trust.
Leveraging AI Automations Requires Balancing Speed And Effectiveness
To get the most benefits from AI automations, investors must find the right balance between speed and effectiveness. For example, AI automations can help advisors with onboarding and reporting processes, streamlining the process and cutting costs. Using AI in those capacities frees up time for advisors to help clients develop and deploy strategies. It can also help to ensure processes are compliant with relevant regulations.
In data analytics, investors and advisors can find balance by ensuring they have the final say, as human oversight is crucial to ensuring AI-driven systems provide what they promise. It is also vital to consider AI-driven recommendations in light of any changes to an investor’s situation or goals.
Algorithm-Driven Portfolios Have Risks And Limitations
Advisors turn to data to develop portfolios to guard against investing becoming emotionally driven. AI takes that practice to the next level, allowing data-driven portfolios to be developed and maintained with unprecedented speed and accuracy. With AI, investing can incorporate far more data inputs, allowing for more comprehensive strategies.
Choosing algorithm-driven portfolios, however, essentially makes AI the “trusted advisor” that guides investments, which raises some important questions. For example, the “black box” problem associated with AI processing means investors may not be able to know the thinking behind AI’s actions. Whether or not all critical factors are being considered could be a question that can’t be answered.
Overfitting — which occurs when an AI model learns its training too well, resulting in poor performance when new data is introduced — is another problem that can affect the performance of algorithm-driven portfolios. Model complexity is one factor that can contribute to over-fitting, but swings in market conditions can also cause significant problems for models affected by overfitting. Rather than using new data to develop new strategies and adjust portfolios, overfitting causes models to over-optimize based on historical data. While those who produce the models can attest to their effectiveness in the lab, they can’t say how they will perform with new data that emerges from markets.
Algorithm-driven portfolios can also fail investors looking for customized strategies. Once core algorithms are developed, it can be impossible to adjust parameters like asset allocation or rebalancing schedules. In many cases, algorithm-driven platforms are built around standardized risk profiles that lock investors into certain strategies.
Herding behavior is another risk that algorithm-driven portfolios bring to the investing space and occurs when large numbers of investors make decisions based on similar strategies. If not tempered by adjustments that consider a broader context, herding can lead to increased market volatility and unanticipated losses.
AI Tends To Downplay The Value Of Investor/Advisor Interactions
The value of the relationship between investors and their human advisors can be lost when algorithm-driven portfolios are used. Advisors do more than simply offer options and activate strategies, though these are the primary functions of algorithm-driven portfolios. Advisors help investors to lay the groundwork for a solid strategy based on their goals.
For example, detailed conversations between investors and advisors are often needed to define aspirations, limitations, and expectations. Advisors can help their clients understand the factors in play and how they affect each other, as well as help investors understand and define their risk tolerance.
The investor/advisor relationship is also important when portfolios don’t perform as expected. Algorithm-driven portfolios can make changes needed to redirect investing, but won’t engage with investors in a comforting and reassuring way. Human advisors bring the emotional intelligence required to steer investors away from emotional responses triggered by market movements and point them back to solid, data-driven decision-making.
Artificial intelligence is changing the investment landscape, providing investors and advisors with a variety of new tools for enhancing efficiency and uncovering insights, but how trustworthy AI is as an advisor remains to be seen. For now, investors are best advised to strike a balance that leverages the efficiency AI can provide while not abandoning the stability and understanding that experienced advisors bring to the process.
