Are All Robos The Same?

| About: Vanguard Total (VTI)


All the robo-advisors promote sophisticated models to generate investment portfolios for investors based on their risk profile.

Upon review, Wealthfront and Betterment produce very different investment allocations for the same hypothetical investor.

The differing proprietary models and inputs clearly differentiate the potential returns and risks you can expect from different robo-advisors.

There are plenty of online reviews of the burgeoning robo-advisor space. Most of the reviews cover basic info such as fees, minimum account size and extra value features such as tax-loss selling for which it is easy to do a comparison. Few of them, however, do a tough analysis of the core product that is actually being sold: the proprietary strategic investment models and the inputs.

The two main independent robos are Wealthfront and Betterment. They have comparable models, but upon review they provide significantly different investment portfolios for a given level of risk and, consequently, likely different expected investment performance.

Most robos employ the basic efficient frontier approach to portfolio creation based on the work of Harry Markowitz published in 1952. He shared the 1990 Nobel Prize in Economics with William Sharpe and Merton Miller for his work. The model evaluates the risk, return and correlations of asset classes and uses mean-variance optimization to create "efficient" portfolios of assets that produce the highest return for a given level of risk. Enhancements to the Markowitz model have been developed since its introduction to strive to improve the results and provide enhanced asset allocations.

Robust discussion of the model approaches are presented on the websites of both Wealthfront and Betterment. Both Wealthfront and Betterment use portfolios of ETFs such as VTI, VEA, BND etc., to create efficient portfolios. Wealthfront is very transparent and provides the actual correlations, standard deviations and expected real returns of all the asset classes used in its model. They disclose that they update these assumptions annually. Betterment provides similar information, but with much less detail. Both robos disclose that they use the core Markowitz Model to generate their efficient frontier with the Black-Litterman model and other models to generate better forward-looking expected returns.

Though both Wealthfront and Betterment use the same basic models, the investment allocations are significantly different. Based on a 65-year-old retiree with a tax-deferred moderate risk profile, the differences are summarized below:

  • Wealthfront uses 10% more equity (including dividend growth) than Betterment
  • Wealthfront uses broader and fewer asset classes (7) than Betterment (11)
  • Wealthfront uses less international bond exposure (9%) than Betterment (20%)
  • Both Wealthfront and Betterment make liberal use of Vanguard ETFs





























Total Equity


Total Equity


















Total Bond


Total Bond






Source: Wealthfront, Betterment

The differences in the investment portfolios could produce significantly different investment results due to the large 10% difference in equity allocations. It is interesting to note that Wealthfront makes a distinct allocation to the dividend growth sector through VIG that has performed well over the past five years and provides yields comparable to the bond space, while providing some upside potential.

Wealthfront provides a neat investment projection analyzer on its website that indicates its portfolio is projected to return 4.21% annually over a 10-year horizon with a 12% chance of loss. I was not able to find a comparable projected return metric from Betterment for its model on its website, but expect that its return could be lower with less risk given its 10% larger bond weighting.

Betterment does a much better job presenting hypothetical backtested charts and tables of historical returns showing 5.1% annual returns from June 2006 to June 2016, whereas Wealthfront only shows a static chart with limited interactivity. Backtested results could be instructive to see how the portfolios would have performed during the Dotcom Crash in 2002 and the Credit Crisis in 2008. It appears that both Wealthfront and Betterment suffered during 2008/2009, but without more information it is hard to do a direct and precise comparison. Betterment does not go back to 2002 to see how it would have performed during the Dotcom Crash.

The focus on broader asset classes from Wealthfront compared to the more refined equity and bond asset classes is harder to critique. The overweight to value and small cap equities and international and emerging market bonds in Betterment could work out to its benefit, compared to the different bets on dividend growth and REITs (through VNQ) from Wealthfront. These different bets expose the portfolios to more scenario-dependent risk for each portfolio.

Interestingly, for taxable accounts not shown here, both Wealthfront and Betterment have tilted their bond allocations into the municipal bond space, given the observation that the persisting "bond bubble" has priced municipals very attractively on a pre-tax basis relative to Treasuries, corporates and high-yield bonds for a given level of risk. It is thus clear that both of these robos are sensitive to current market conditions while creating their portfolio allocations. Hopefully, however, they do not get whipsawed from rebalancing to and from these anomalies that could be shorter-term in nature and could reverse badly.

Not surprisingly, both Wealthfront and Betterment make liberal use of Vanguard as the low-cost leader in ETFs. They both support a low-cost approach.

Both Wealthfront and Betterment do a good job promoting themselves as quant shops for investors looking for inexpensive online sophisticated investment management solutions. However, even though the underlying models are similar and promise the same level of academic rigor, the final portfolio allocations and investment performance results are likely to be quite different.

Investors using robo-advisors should be warned that the robo-advisors are not all alike and they have biases and idiosyncrasies just like human advisors. It will be very interesting to see some actual standardized performance metrics, like those provided by mutual funds and ETF providers complying with CFA Institute Global Investment Performance Standards ("GIPS"), so that we can have a better sense of how they are doing and do a fair comparison.

So for now, buyer beware!

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Tagged: Financial Advisors, Wealth Management,
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here