About 2 years ago, I tested Wealthfront as well as ETrade's roboadvisor (https://us.etrade.com/what-we-offer/our-accounts/core-portfo...), mostly to see their UX. ETrade did a much better job of exposing the predicted 99th percentile VaR and predicted peak-to-trough drop (though, understandably, I don't think it used either phrase).

Basically, "Can you handle a 30% drop and not lose sleep (and certainly not consider liquidating)? 40%? 20%?"

For most investors, 99% VaR, peak-to-trough, and maybe 95% VaR will lead to a more accurate risk tolerance than their self-assessment of risk tolerance (low, medium, etc), desired returns/goal (the cute graph of possible asset trajectories), or tolerance for volatility in an average year. Their failure case is experiencing a big drop and reactively selling, all because they took too much risk. This is also true of investors who work with human advisors (https://www.vanguard.com/pdf/ISGQVAA.pdf pages 4 and 16 - "Behavioral Coaching"). A great roboadvisor should clearly and loudly predict what an awful year (99% and 95% VaR), and an awful multi-year bear market (peak to trough) looks like.