Not all roboadvisors are created equally.

Investors driven by the allure of asset allocation and rebalancing services at minimal costs embrace the roboadvisor trend, but the choice of digital advice platform has major implications for their experience, according to a recent study by Martinsville, N.J.-based Condor Capital Management.

“Any single given investor might be better off in one roboadvisor or another, but with the current amount of information out there, it would be difficult for an individual investor to determine which one is best,” says Ken Schapiro, president of Condor Capital Management. “That’s the problem our report is trying to solve.”

Earlier this month, Condor Capital, via its sister firm BackenDBenchmarking, released the fourth edition of “The Robo Report,” a quarterly analysis of the investment returns, portfolio composition and trades offered by leading roboadvisors.

According to the report’s authors, the top performing roboadvisors of the second quarter of 2017 were E*Trade’s hybrid service, where a balanced portfolio in a taxable account returned up to 3.22 percent, and WiseBanyan, where an aggressive portfolio returned up to 3.88 percent.

The worst performing roboadvisors in the second quarter, according to the study, were Wealthfront and Personal Capital, where a balanced portfolio in a taxable account returned as little as 2.13 percent, and Hedgeable, where an aggressive portfolio in an IRA returned 2.71 percent.

The best performers in terms of one-year trailing returns were Schwab, which posted an 11.94 percent return over the past 12 months, and E*Trade’s hyvrid service, which returned 11.92 percent. The worst performers were Acorns, posting a 7.3 percent return, and FutureAdvisor, which posted 8.34 percent returns.

For the report, Condor opens up its own accounts with roboadvisors and tracks the holdings and performance of those accounts. In creating the accounts, Condor’s advisors provide similar answers to the risk tolerance and account opening questionnaires.

“We don’t publicly publish our own performance, but we do show it to clients and prospects on a one-on-one basis,” says Schapiro. “People are putting their money into these roboadvisors and it’s kind of like a black box in that nobody really knew how these tools allocated and how they had performed over time. So two years ago I started opening roboadvisor accounts in hopes of finding out.”

In taxable accounts, the Condor methodology opens up accounts with a moderate, balanced allocation of 60 percent stocks and 40 percent bonds to reflect a standard demand for investors in high tax brackets. In IRAs, Schapiro opens the most aggressive account available, defined as accounts with the highest allocation to stocks. While the baseline allocations are not identical across all study subjects, they’re similar enough to create a basis of comparison.

Schapiro argues that roboadvisors are often overhyped to investors, leading to misconceptions about what a digital advice service actually offers to its clients.

“Roboadvisors suffer from the same problem as helicopter skiing: When people hear I go helicopter skiing, they think that I jump out of a helicopter, but that’s not true,” says Schapiro. “With robo-investing, people think there are a tremendous amount of trades being done and hidden algorithms that the human eye can’t figure out, but the vast majority of these platforms are very passive and the most dramatic thing that they do is the actual signup process.”

Thus far, Condor is tracking portfolios from Acorns, Betterment, eTrade, Fidelity Go, Future Advisor, Personal Capital, Schwab, SigFig, Tradeking, Vanguard, WiseBanyan, TD Ameritrade, Ellevest, Hedgeable and Merrill Edge. Since many of the accounts in the study have been open for fewer than two years, the most recent results should be taken with a grain of salt.

The study’s subjects varied greatly in performance and allocation over a very short time period, yet roboadvisors don’t talk about their differences in their marketing materials, says Schapiro, electing instead to emphasize their commonalities: low fees, ease of use, automation and all-digital interactions with clients.

“Also, the Department of Labor’s fiduciary rule is forcing retirement accounts to be fee accounts,” says Schapiro. “Young investors with small accounts aren’t going to garner the attention of traditional advisors because they won’t generate enough money on a fee basis to be profitable as clients. Roboadvisors are a good solution.”

According to Condor, approximately $150 billion in assets are now contained in roboadvisor portfolios. That number is expected to grow to $489 billion in 2020, according to a recent report by Cerulli Associates, accounting for more than one-fifth of total RIA AUM.

“We think people are still getting warmed up to the idea of allowing a digital platform to manage their investments,” says Schapiro. “I think we’ll find that more investors are okay with limited discretion moving forward.”

In the second quarter, total portfolio returns for roboadvisors ranged from 2.13 percent to 3.22 percent in taxable accounts, and from 2.71 percent to 3.88 percent in IRAs. The top three roboadvisors for total portfolio returns were E*Trade’s hybrid service, Vanguard and Betterment. In the first quarter, the top three roboadvisors based on trailing 12-month portfolio returns included Schwab, Sigfig and Personal Capital.

In the second quarter, the top three roboadvisors for equity returns were FutureAdvisor, E*Trade’s hybrid service and WiseBanyan. By comparison, SigFig, TD Ameritrade and Vanguard posted the best equity returns during the first quarter.

The top four roboadvisors for fixed-income returns in the second quarter were E*Trade’s hybrid service, Schwab and a tie between Fidelity and E*Trade’s ETF platform. In the first quarter, Schwab, SigFig and TradeKing Momentum had the best performing fixed-income allocations.

Exposure to international investments explained some of the divergence in returns. Digital advice platforms like FutureAdvisor, Schwab and Betterment kept more than 40 percent of their equity allocation in international stocks, while robos like Hedgeable, Acorns and E*Trade’s ETF service dedicate less than 30 percent of their equity allocations to international equities.

In past years, exposure to international securities hurt the relative performance of some roboadvisors, notes Schapiro, with many globally oriented digital platforms underperforming due to China’s slowing growth and one-off international events like 2016’s Brexit vote.

The outperformance of growth stocks relative to value stocks has also shaken up the roboadvisory world, according to the report, because many robo portfolio managers weight towards value to capture the historical outperformance of inexpensive stocks. Betterment, Schwab, FutureAdvisor, and to a lesser extent Ellevest and Merrill Edge were all negatively impacted by their preference for value investing, according to the report.

“There are no roboadvisors overweighted in growth-style investing,” says Schapiro. “At the same time, they’re varying in a wide degree in the risks they’re taking on the stock and the bond sides.”

The inclusion of alternatives and real estate within roboadvisor portfolios should have an impact on their performance, notes Schapiro, but few roboadvisors currently offer alternative exposures beyond an allocation to REITs.

While many roboadvisors use domestic REITs within their allocations or allocate to international REIT ETFs, only three hold gold assets: Personal Capital, Schwab and Hedgeable. Personal Capital is also the only roboadvisor in the study with commodity exposure, via a small allocation to an ETF.

Finally, the use of active fixed-income managers has helped some roboadvisors generate alpha for their users, in particular, E*Trade’s hybrid offering harnessed active management within two municipal bond funds, allowing it to outperform the E*Trade ETF strategy and its roboadvisor peers.

Schapiro says the results thus far have been measured over two to three years of a bull market, and thus additional services like tax-loss harvesting have little impact on the performance of roboadvisors over the study period. Investment trends are also difficult to distinguish over a short time period encompassing a bull market.

The idea, says Schapiro, is providing greater transparency to how roboadvisor algorithms work in different market environments. By tracking the real performance of roboadvisors, investors and human advisors can make comparisons and more informed decisions about where to allocate assets.

Roboadvisors comprise a diverse assortment of client prospecting, onboarding, account opening and asset allocation tools—not every roboadvisor contains the same functionalities, and not every platform uses the same automated and algorithmically guided processes. Some robos include cash flow management and savings tools, while others do not. Some roboadvisors are actually “hybrids” of digital platforms and traditional advisors or phone banks of financial professionals.

“It appears that roboadvisors are starting to expand their market or are realizing that many people do want human advice, someone they can chat with,” says Schapiro. “That’s a trend we’ve identified since the start of the study.”

Condor will consider any roboadvisor directed towards retail investors and advisors for its study, but does not include tools that an investor cannot use without intermediation from an advisor like most professional planning software.

Condor also tracks news around the roboadvisory industry, publishing a short digest as part of its quarterly reports.

“I don’t feel like the market is going to support a lot of different robo-only firms,” says Schapiro. “Generally, these firms are not profitable, but losing money on a roboadvisor is not a big deal for Vanguard or Schwab, where a roboadvisor can serve as a loss leader for them. They can also use roboadvisors to move assets from self-directed accounts to managed accounts, which means they can garner more assets, produce more revenue from those assets, see more consistent performance in those assets, and keep those assets for a longer period of time.”

The report doesn’t offer investors any insight into the effects of advisory services offered around digital advice providers, whether they be automated recommendations and prompts delivered by a roboadvisor algorithm, or financial guidance provided by call centers full of advisors.

The report also doesn’t consider other value adds offered by digital advice providers, such as Acorns’ automatic deduction and saving tools, Wealthfront’s direct indexing for large accounts or Personal Capital’s holistic do-it-yourself planning offerings.

In the past, Condor has received backlash from companies controlling some of the roboadvisors it is studying, says Schapiro. Most often, negative responses criticize the time horizon or research methods of the study. However, in some cases,companies have attempted to cancel or block Condor’s access to roboadvisory accounts.

In the past, providers like Wealthfront have closed Condor’s roboadvisory accounts, while Vanguard moved a Condor account from its Personal Advisor Services platform to a self-directed account. Both companies have since reversed course and permitted Condor to continue its research.

“We’ve seen a new trend of roboadvisors contacting us directly and wanting us to include them in our study,” says David Goldstone, a Condor Capital research analyst. “Since we’ve become established as a go-to source for this kind of information, we’ve also seen less push back from companies trying to close our accounts.”

Schapiro says that the BackenDBenchmarking techniques applied to roboadvisors will eventually be applied to any investment available in the retail segment.