Humans vs Robots: Which Is Better for Managing Investments?

By Anton Altement, CEO of Polybius and OSOM Finance,

In an era of technological advancement, innovation, and fear-mongering sci-fi programs, fears over a robot uprising and artificial intelligence coup are rife. While these two hyperbolic scenarios are likely a while off, trading bots used within financial dealing are starting to supersede their human researcher counterparts. On Wall Street, these infallible and emotionally-neutral trading automatons are gathering acclaim. And some propose that they’re going to change the face of finance forever.

Let’s face it, not many humans are cold-blooded and rational enough, which are essential qualities for long-term trading success. This means those few strong traders can ask for high fees for their services, which are often completely unpalatable for a small investor. Robo-advisors, on the other hand, do away with this hubris, epitomising financial inclusion and cost-efficiency. Moreover, they are much more scalable than a human trader, with the ability to trade multiple markets at once.

Major commercial banks are the first to see the potential in these robo researchers. In 2019, multinational investment bank Goldman Sachs announced its own robo-advisory service. While the launch is postponed until next year due to coronavirus-based disruption, the market for robo advisors is still booming, with trading bot usage has grown between 50% and 300% from December 2019 to January 2020.

Why? Because unlike human traders, robots aren’t restricted by the primal urges of the reptilian brain.

A Quantitative Solution to Irrationality

There are few triggers more powerful in electing an emotional response than money, power, and greed. Our internal struggle to satisfy any one of these desires can set us on a disastrous course for failure—particularly when it comes to trading. The fear of missing out, loss aversion, and even hubris present major obstacles for traders to overcome. And, historically, we have very little success in doing so.

There are a few techniques at a trader’s disposal when it comes to evaluating entry and exit points for a trade. For the most part, they can be categorised into two distinct approaches: qualitative and quantitative analysis.

A qualitative approach involves in-depth data analysis pertaining to subjective information, such as company management, earnings, and competitive advantage.

Quantitative analysis, meanwhile, examines the statistical attributes of an asset, including performance, liquidity, market cap, and volatility. For the data-driven cryptocurrency market, with its swathe of exchanges and bounty of information (total supply, transaction volumes, fees, and mining metrics, etc.), it’s the latter quantitative approach that is often favored.

This is reflected by the 2020 PwC–Elwood Crypto Hedge Fund Report, which details that nearly half of all crypto fund managers (48%) opt for a quantitative trading strategy. And there’s one clear reason as to why. A quantitative approach—in the main—aims to neutralise cognitive bias.

Still, try as they might, no human is capable of totally ignoring their primal instincts. And that can prove troublesome.

In a study into emotional reactivity on trading performance, researchers of the MIT Sloan School of Management found that excessive emotional responses can be extremely detrimental to trader returns, particularly during times of crisis and within high volatile markets.

But where humans fall down, the novel trading bot thrives.

The Rise of the Robo Advisor

Trading bots are much more nuanced than their all-encompassing moniker would suggest. These bots come in many different varieties. Two of the most common are the analyst and advisor bots. The latter advisors build portfolios based on the client’s risk profile. Robo analysts, meanwhile, probe data released in annual company records, as well as SEC filings, to provide buy and sell recommendations.

Despite their varying traits, both benefit from negating the cognitive biases inherent in human researchers, analysts, and traders.

As such, within volatile and high-pressure market conditions, trading bots have proven to surpass the performance of their human equivalents.

A 2019 study from Indiana University appraised over 76,000 research reports published over 15 years from various robo-analysts. Researchers found that the robo buy recommendations conferred 5% better returns than those of the human analysts.

But while bots may have the edge over humans, their results vary wildly when competing amongst themselves.

Between May 2019 and March 2020, researchers pitted 20 German B2C robo-advisors against each other, measuring their performance and calculating the differences. The variation among the bots was enormous. But most impressive of all was the bot that came in pole position. The top robo advisor managed to restrain losses to just -3.8%, beating the other bots by around 14 basis points. And decimating traditional hedge funds who were down approximately -10% across the board following March’s tumultuous marketwide crash.

As it turns out, the main difference between the top robo performer and the rest was its unique strategy. The robo advisor not only used quantitative analysis, but it leveraged the irrationality of the market to its advantage—measuring conventional risk metrics, such as loss aversion bias and recovery time, to ascertain illogical trades and position itself on the other side. In doing so, it was able to interpret the market better than both the determinedly quantitative-based bots and the human-operated hedge funds.