Is There a Hard Limit to Investing Algorithms' Progress?
JUL 06, 2017 21:07 PM
A+ A A-

Is There a Hard Limit to Investing Algorithms' Progress?

by Anna Johansson
Investment algorithms have become increasingly popular, attracting cautious and technologically enthusiastic investors who want a safe, reliable way to make investing decisions. So far, they've been successful in securing their investors a return on their investments, and they keep getting better as automated trading companies like Wealthfront and Betterment continue investing in better technology.
But is there a limit to how far these algorithms can progress? Or a limit to how much we can rely on them for the future of our investments?
How Algorithms Are Effective
First, let’s take a look at how algorithmic trading works and how these companies have managed to succeed thus far. It sounds futuristic and complex, but the realities of algorithmic trading are no more complex than Google search or similar algorithmic processes.
Though each company uses a proprietary algorithm, with different specific processes, all algorithms operate by following a core set of rules. These rules may include things like “buy a stock when its 30 day moving average moves beyond its 100 day moving average,” or “sell a stock when its 30 day moving averages falls below its 100 day moving average.”
Accordingly, algorithms are effective for helping investors secure the best possible price for equity investments, and curtailing losses as proactively as possible. Because these algorithms operate automatically, they also reduce the amount of errors made by human brokers—and decrease the total costs for overseeing such a system.
There are some variants to how algorithms specifically make decisions. Most focus on a combination of trend monitoring and intelligent rebalancing, selecting different types of investments based on the specific investor’s tolerance for risk, and basing most decisions based on how they compare with broader market trends.
Potential Downsides of Investing Algorithms
However, there are some potential downsides to algorithmic trading—and some major problems that algorithm producers need to resolve:
  • Investing options. Most investment algorithms focus on simple investment options like stocks, bonds, and index funds. However, there are far more complex securities to exchange, including futures and commodities, which require far more knowledge and experience to achieve success. You’ll need a thorough understanding of risk, trading strategies, and other variables here that algorithms just aren’t ready for.
  • The limits of our economic understanding. Even our best economists don’t fully understand the complex inner workings of our global economy. Markets boom and crash, seemingly without warning, and experts frequently contradict each other on what we should expect from the coming months and years. Computer models can’t perfectly predict how the market will evolve from here; all we can do is use past data and make assumptions about how that past data may be applied to future happenings.
  • Consumer faith. Excessive consumer trust in algorithmic trading may also become problematic. Investors who see algorithmic trading as a catch-all investment strategy (i.e., they assume they don’t need to do research or make decisions on their own) will see worse returns than their avid researching counterparts. Should there be an unexpected downward fluctuation in consumer profiles due to an algorithmic trading error or glitch, it could trigger a miniature cascade of negative effects, including algorithmic investors pulling out of the market, potentially collapsing the entire industry. We don’t have enough information on algorithm performance to rule this possibility out.
  • Algorithmic direction. Algorithmic trading has also been known to trigger “flash crashes” and other peculiar market behaviors, in response to their automated, unified trading decisions. Developers have been trying to solve this problem in recent years, but it’s nearly impossible to tell how an algorithm could affect the broader economy—especially when working in unison with other trading algorithms.
Limits Today and Limits Tomorrow
Currently, algorithmic trading is limited by how well we understand the economy, and how well we understand algorithms’ roles in that economy. However, both of these problems are surmountable with sufficient research and further development. It may be decades before algorithms are so polished that they work with markets like well-oiled machines, but for the time being, they’re relatively safe investment vehicles.
The only hard limit for algorithms’ progress is their potential return. Though it may be possible, through deep learning, to eventually mimic human intuition, the emotional, risky decisions made by human investors still have a higher potential return than a safe, long-term investment algorithm could allow.
This doesn’t reflect the limits of algorithm design potential, but sets the stage for how far we’ll push trading algorithms—and how the market will change, accordingly. 
Anna is a freelance writer, researcher, and business consultant. A columnist for, and more, Anna specializes in entrepreneurship, technology, and social media trends. Follow her on Twitter and LinkedIn.
[%= name %]
[%= createDate %]
[%= comment %]
Share this:
Please login to enter a comment:

Computing Now Blogs
Business Intelligence
by Keith Peterson
Cloud Computing
A Cloud Blog: by Irena Bojanova
The Clear Cloud: by STC Cloud Computing
Computing Careers: by Lori Cameron
Display Technologies
Enterprise Solutions
Enterprise Thinking: by Josh Greenbaum
Healthcare Technologies
The Doctor Is In: Dr. Keith W. Vrbicky
Heterogeneous Systems
Hot Topics
NealNotes: by Neal Leavitt
Industry Trends
The Robotics Report: by Jeff Debrosse
Internet Of Things
Sensing IoT: by Irena Bojanova