Even Within the Same Company, Returns Vary Drastically... Alternative Data and Machine Learning Capabilities Prove Decisive

Essential to Check Downside Protection and Total Costs Over Short-Term Returns... Careful Selection Aligned With Personal Risk Profile Required

[CBC News] The Financial Services Commission recently disclosed for the first time the complete performance records accumulated since the launch of its robo-advisor testbed. By compiling and comparing the cumulative returns of numerous algorithms that have been in operation, this release offered a rare opportunity to survey the entire robo-advisor market at a glance.

The results were far more dramatic than expected. The top-performing algorithms recorded overwhelming performance, opening up a clear gap from other algorithms.

Performance Varies Even Within the Same Company

What drew particular attention in this release was not simply the performance of the top-ranked algorithms. It also revealed that the performance variance between algorithms within the same company and on the same platform was substantial. While some algorithms recorded notable returns even during periods when the broader market was shaken, other algorithms operated by the same company fell short of market averages. Although they are grouped under the umbrella of robo-advisor asset management, they are in effect a mix of products built on entirely different strategies and yielding completely different results.

This carries important implications for investors looking to use robo-advisors. It has become clear that a casual approach — assuming that "if you leave it to a robo-advisor, it will manage things well on its own" — is insufficient, and that outcomes diverge significantly depending on which algorithm is chosen.

Technology Gaps Translate Into Return Gaps

Notably, the top-ranked algorithms were those that actively leveraged alternative data and machine learning-based analysis. This demonstrates that even within the robo-advisor space, gaps in technological capability translate directly into differences in actual returns.

Three Things to Check When Choosing a Robo-Advisor

1. Cumulative Performance and Downside Protection

The first thing to check is not an algorithm's short-term return, but rather its cumulative performance built across various market conditions and its ability to defend against downturns. There is no guarantee that an algorithm that posts a brief flash of strong performance in a bull market will deliver the same results in a bear market.

In practice, since robo-advisors are structured to optimize diversification and risk management, there were cases where they failed to keep pace with human-managed funds or index gains during periods of sharp, one-directional market surges. Conversely, when unexpected sharp declines occur, emotion-free mechanical rebalancing can serve as a strength.

2. Total Cost Calculation

It is also important to scrutinize the fee structure. Comparing only management fees makes it difficult to fully grasp actual costs. One must look at the total cost — including the embedded fees of the ETFs themselves, currency exchange fees, and withdrawal fees — to understand the true burden.

3. Note That Principal Is Not Guaranteed

It should also be kept in mind that robo-advisors are not deposits, meaning the principal is not guaranteed. A practical approach is to start with a small amount, observe the rebalancing process and performance, and then gradually increase the investment.

They Are Not Homogeneous Products

Ultimately, the release of the testbed performance records reaffirmed that robo-advisors are not a single homogeneous product but rather a diverse group of products, each carrying entirely different performance outcomes and risks depending on the algorithm. For investors, more caution than ever is needed — rather than being drawn to flashy top-return figures, they should carefully compare and select algorithms that match their own investment profiles and goals.

[※ This article is an investment reference produced with AI assistance. It does not constitute investment solicitation, and investment decisions and their consequences are the sole responsibility of the investor. This publication bears no legal or financial responsibility.]