Do Robo Advisors Use Artificial Intelligence or AI?


A word in any language should be used literally, especially when it’s a technical term with a precise, logical, or scientific definition, meaning, or explanation. However, our species has a penchant for the figurative, symbolic, metaphorical, analogical, and allegorical. Industries, in particular, get excitedly imaginative when they classify products or services, like robo advisors.

Most robo advisors don’t use artificial intelligence or AI in its truest sense. Most robo advisors are pre-written codes, executing predetermined functions based on inputs. There may be one or more algorithms, simple or complex, depending on the predefined purpose and a robo advisor program’s scope.

The word robo is a misnomer here. There’s no element of robotics in fin-tech unless an underlying software or algorithm initiates, regulates, or performs a mechanical function. Likewise, robo advisors can only be said to use AI only when they actually do so, not because they automate some strategy execution for traders. 

While our world is evolving fast, and AI might have all the potential to become a prominent force in the financial markets, let us take a first hand look at where things stand today. So, read on for an insightful discussion!

IMPORTANT SIDENOTE: I surveyed 1500+ traders to understand how social trading impacted their trading outcomes. The results shocked my belief system! Read my latest article: ‘Exploring Social Trading: Community, Profit, and Collaboration’ for my in-depth findings through the data collected from this survey!

All Robo Advisors Use Algorithms

All robo advisors rely on algorithms to function as intended. An algorithm is a computer program, a set of codes meant to perform a specific task or predefined functions. The tasks or functions depend on the inputs provided by a user. Hence, a typical algorithm is a finite loop with rigid rules and simple if-then and true/false consequences.

Let’s go back to C, the mother of all programming languages, to elaborate on the topic further. A single set of code written in C or any subsequent programming languages can add 2 numbers and return the sum as the outcome. However, this simple code can’t compute calculus.

An algorithm can solve calculus problems but only if the codes in the program include all the concepts, the integration & differentiation formulas, and the necessary if-then-else and on/off or true/false conditional statements.

For most part, a robo advisor program can do only those tasks that it’s developed for and intended to perform. The algorithm can’t serve any other purpose. Thus, it doesn’t learn, adapt, or grow by itself without new codes or inputs from the developer. By definition, such an algorithm isn’t artificial intelligence or AI.

Many Robo Advisors Use Machine Learning

The earliest robo advisors used only algorithms. In the early 2000s, such programs were exclusively available to asset management companies, wealth managers, and financial analysts. Contemporary robo advisors tend to use machine learning to varying extents.  

Machine learning is the ability of a set of codes or an algorithm to acquire insights through interactions with a user and adapt itself to compile and present more topical information. It’s machine learning that drives Google’s search engine result pages, YouTube’s and Amazon’s recommendations, and Alexa’s or Siri’s personalized experience.

When you ask a robo advisor to shortlist the best-performing index funds for a selected period, the algorithm refers to the database, compares the returns on investments, picks the top few, and presents them to you through the interface.


If this robo advisor program uses machine learning, it’ll use your inputs to improve its insights. Subsequently, the program will know what kind of topical information you seek. The same logic applies to how a robo advisor suggests actionable tips and makes passive investment decisions.

Still, machine learning isn’t necessarily artificial intelligence. By definition or theoretically and practically, machine learning is a precursor to artificial intelligence, a part of it, but in its most primitive form. This version is known as artificial narrow intelligence or ANI.

Some Robo Advisors May Use Deep Learning

Machine learning or ML predates artificial intelligence or AI. More than ML, deep learning has had and continues to have a more profound impact on artificial intelligence. Deep learning is exponentially more complex, insightful, and consequential in AI’s development and future.

Machine learning deals with data primarily, such as texts, numbers, and other written information. Deep learning can deal with voice, images or visuals, and sensory interactions such as tap or touch on apps and mobile interfaces.  

It drives the image and voice or speech recognition algorithms used by Google. Netflix uses deep learning to recommend its contents and also to assess viewer engagement internally.

A robo advisor leveraging machine learning uses the acquired insights restrictively to return similar outcomes and operate within the limited parameters. Unlike machine learning, deep learning can derive insightful and implied conclusions.

Deep learning enables a robo advisor algorithm to apply the acquired insights to aspects beyond the user’s limited interaction and inputs. These insights and assessments with their implications influence data processing, analytics, and actionable pointers for a particular user.

Despite being an advanced subset of machine learning, deep learning is not synonymous with artificial intelligence. The impact and relevance of deep learning remain within the ambit of artificial narrow intelligence.

Most Robo Advisors Don’t Use Artificial Intelligence

Real artificial intelligence has many fundamental prerequisites. Some quintessential attributes are autonomy, self-improvement, and auto-correction at the code level, rewriting old codes and writing new ones, continuous algorithmic evolution, and development of advanced & complex programs.

Genuine AI is artificial general intelligence (AGI) or artificial superintelligence (ASI). 

AGI is the equivalent of a human mind. Artificial general intelligence has consciousness or self-awareness. Its algorithms empower the entire program to think and act like a person.

ASI is smarter than the average human mind.

Artificial superintelligence has advanced consciousness and cognitive abilities, making it capable of real-time computing far beyond our mortal prowess.

No known robo advisor uses either AGI or ASI. Some may use ANI or artificial narrow intelligence to perform a single or a few selected tasks. Let us consider the examples of a few popular robo advisors:

Betterment

Betterment has a few strengths, such as goal-based tools, automatic rebalancing, and fractional shares. Hence, the robo advisor is suitable for beginners aspiring for well-customized portfolios, risk mitigation in a volatile market, and maximized cash management.

Yet, all these strengths are powered by rigid algorithms, significantly driven by machine learning but without any deep learning, program autonomy, or code consciousness. Hence, Betterment doesn’t use AI.

Wealthfront

Wealthfront has a few strengths, such as automatic rebalancing and tax-loss harvesting. Like other robo advisors, this one also harnesses the power of modern portfolio theory. The tax-loss harvesting feature is irrefutably helpful, especially for beginners and those without dedicated accountants.

Yet, the strengths are essentially singular tasks. These features are per predetermined and predefined functions of Wealthfront’s algorithms. Thus, this robo advisor doesn’t use AI.

Interactive Advisors

Interactive Advisors, the robo advisor version of Interactive Brokers Group, has its strengths and weaknesses. It empowers passive investors to explore active investing without much effort, albeit at a cost. A relatively unique feature of this service is socially responsible or sustainable investment recommendations.

Yet, Interactive Advisors’ services don’t transcend the limitations pre-set by the developers. The algorithm can mirror the market, indexes, and portfolios of stocks and exchange-traded funds operated by active wealth managers.

However, the codes aren’t self-aware, proactively comparing anything beyond its parameters or insightfully recommending something not already permitted by the algorithm.

Even if a robo advisor offers a bouquet of all available services, the algorithm is still a predetermined set of rigid codes programmed to serve predefined purposes.

Genuine autonomy, consciousness, auto-evolution, and self-awareness-driven changes are nonexistent. Such attributes are the quintessential criteria for a program to be classified as AI.

Author’s Recommendations: Top Trading and Investment Resources To Consider

Before concluding this article, I wanted to share few trading and investment resources that I have vetted, with the help of 50+ consistently profitable traders, for you. I am confident that you will greatly benefit in your trading journey by considering one or more of these resources.

Conclusion

There are innumerable endorsers and critics of algorithms, machine learning, and other code-driven analytics in investing, trading, and financial planning. True AI is an untested technology in this context.

Artificial intelligence may offer unprecedented benefits to investors and traders. On the flip side, there could be unforeseen risks and tragic consequences, too.  

BEFORE YOU GO: Don’t forget to check out my latest article – ‘Exploring Social Trading: Community, Profit, and Collaboration. I surveyed 1500+ traders to identify the impact social trading can have on your trading performance, and shared all my findings in this article. No matter where you are in your trading journey today, I am confident that you will find this article helpful!

Affiliate Disclosure: We participate in several affiliate programs and may be compensated if you make a purchase using our referral link, at no additional cost to you. You can, however, trust the integrity of our recommendation. Affiliate programs exist even for products that we are not recommending. We only choose to recommend you the products that we actually believe in.

Subscribe To Our Mailing List

We send no more than 1 newsletter every month

and, you can unsubscribe at any time

    We respect your privacy. Unsubscribe at any time.

    1. Artificial intelligence | Definition, examples, and applications. (n.d.). Encyclopedia Britannica. https://www.britannica.com/technology/artificial-intelligence
    2. Gravier, E. (2021, September 1). A robo-advisor can invest on your behalf — here’s how they work. CNBC. https://www.cnbc.com/select/what-is-a-robo-advisor-how-they-work/
    3. Personalized Robo-Advising: Enhancing Investment through Client Interaction. (n.d.). Department of Mathematics, University of Texas at Austin. https://web.ma.utexas.edu/users/zariphop/pdfs/Robo-advising.pdf
    4. What is machine learning? (2020, July 15). IBM – United States. https://www.ibm.com/cloud/learn/machine-learning
    5. What is tax-loss harvesting? (2021, March 15). Morgan Stanley. https://www.morganstanley.com/articles/tax-loss-harvesting

    Navdeep Singh

    Navdeep has been an avid trader/investor for the last 10 years and loves to share what he has learned about trading and investments here on TradeVeda. When not managing his personal portfolio or writing for TradeVeda, Navdeep loves to go outdoors on long hikes.

    Recent Posts