[sdnlafkjh34rw]

> Wealthfront, on the other hand, provides passively managed portfolios with financial advice based upon questions they ask their clients. The portfolios may be equivalent to Vanguard's funds, but the value-add is the financial advice. Maybe you don't value that advice, but it doesn't seem fair to characterize this as simply a repackaging.

So I worked at a competitor of Wealthfront and while this is how a robo-advisor is marketed, this isn't actually what Wealthfront does. There is no real personalization of the product; you merely fill out a survey and it comes up with a risk score that adjusts your asset allocation. I know how those surveys work (because I built one) and it is laughably close to a Buzzfeed survey that guesses which game of thrones character you are closest to. The survey gets a measure of risk, which is really just one of a million inputs that you need to give proper financial advice.

The truth is relevant financial advice requires inputs beyond a generic 15 question survey. You need to do a data dump of all your assets, your financial goals, and then you need a human (or robot) to do analysis to figure out the path forward. Wealthfront doesn't do that. It doesn't adjust my asset allocation in my wealthfront account because I already have existing funds in those areas in other brokerages. It doesn't tell me about the backdoor Roth IRA trick which will maximize my retirement savings. It doesn't analyze my 401k funds to tell me which funds make the most sense. It literally is a re-packaged target date fund that is customized based on one data point (client risk). For 99% of Americans it doesn't do any more than a target date fund or a mix of index funds do. Yes, I guess it works for the 1% of americans who's biggest financial issue is that their passive index funds don't exactly match their risk profile.

I'm a personal finance junkie, and in the last year I've talked to a lot of people about their financial goals and challenges and I've yet to speak to one where the main solution to their problems was a robo advisor. It's just not a real use case.

Here are real use cases I've encountered: * Figuring out how loan refinancing can reduce debt burdens long term * Figuring out how to contribute to their kids future education expenses * Figuring out the size of an emergency fund you need and where to put it (the answer is a high yield savings account) * Figuring out a reasonable living budget based on income and debt * Figuring out how to maximize tax burden through a combination of 401K, IRAs * Figure out how to get tax savings once you hit income limits on IRAs

The point I'm trying to make is this "variable mix of index ETFs based on a 15 question survey" is not as much of a gamechanger as the roboadvisor's tout. It's a very limited product that is actually quite easy to build, and doesn't really help most of Americans (but it sounds cool).

For me the sharedholder structure of a credit union and Vanguard is way more revolutionary than the current Wealthfront product.

[daxelrod]

I too have worked at a roboadvisor and to my understanding (as an engineer, not a financial advisor), its advice was more complex and useful than how you describe Wealthfront works. I haven't personally used Wealthfront so I'm not as familiar with what it does specifically.

> "variable mix of index ETFs based on a 15 question survey" is not as much of a gamechanger as the roboadvisor's tout.

I agree that if that's all your roboadvisor is doing, it's a bad roboadvisor. You point out a wide range of other financial advice people need. The thing is, quite a lot of that is possible algorithmically.

> Figuring out how loan refinancing can reduce debt burdens long term

You can absolutely do this algorithmically, and Credit Karma's offering is already moving in this direction, if it's not already there.

> Figuring out how to contribute to their kids future education expenses

If this gets into "how do I change my current spending and income to be able to contribute to my kids' education", you're right, that's not something that fits into the algorithmic model easily.

> Figuring out the size of an emergency fund you need

Not difficult get a reasonably good outcome here with a questionnaire.

> Figuring out a reasonable living budget based on income and debt

Agreed, this becomes similar to the education bullet point above. YNAB is a much better example of how to solve this with a software product.

> Figuring out how to maximize tax burden through a combination of 401K, IRAs

This fits into a roboadvisor's model reasonably well, right? (Although they likely can't make money on the 401K part without being the employer's 401K provider.)

> Figure out how to get tax savings once you hit income limits on IRAs

Is this so different from what TurboTax does algorithmically? (Depending on how fancy you want to get with the tax savings).

You've made a great argument about the original point you made that I challenged, and convinced me. Where we may differ is that I see a large number of corollary services that a roboadvisor could offer that together would be substantially more useful than a glorified target date fund.

[sdnlafkjh34rw]

I think I have a different perspective because even though I'm an engineer I worked in finance before and am a huge personal finance nerd. Some of the engineers on my team were impressed with the "sophistication" of our product, but I remember talking to our head of finance and thinking that's it? Even he admitted it was just basic asset allocation strategy. When I talked to our head of finance, he strongly believed Wealthfront (or Betterment) were doing the same thing. We actually back tested our competitor survey's and asset recommendations to ours and it was pretty obvious that all the roboadvisors in the space were building almost exactly the same risk model with very similar recommendations.

At the end of the day it's not a roboadvisor but it's a robo allocater. Frankly asset allocation isn't a hard problem, I've been doing it manually for years.

> Is this so different from what TurboTax does algorithmically? (Depending on how fancy you want to get with the tax savings). Turbotax does not do this at all. It just fills in your tax forms. It does not advise you on tax strategy. To give an example, I'm gonna assume that since you are/were a roboadvisor software engineer, you are most probably out of the income limits of a roth IRA or IRA. If so, did you ever have TurboTax suggest you to do a backdoor roth when filling out taxes? I've filed taxes with TurboTax and I've never them suggest that (not saying they couldn't). My human CPA did though :)

I think your other objections are that automation could do this work. I totally agree. In my original response, I called out that this work could be automated. So yes, I agree in theory that fintech could do this, and I wish they would. But in my viewpoint, I actually don't see this happening. Specifically a lot of the challenges for everyday Americans finance wise are not profitable to serve.

The point I'm making is that a lot of fintech players like to play this marketing game where they are helping out the "common" man against Wall Street with super "sophisticated" products. In my view thats largely marketing, the products are are not very sophisticated and they aren't interested in serving the common man since they can't make easy money off of them.

[iudqnolq]

What would you do without your knowledge? Would you use Vanguard?

[sdnlafkjh34rw]

For long term investments (i.e. 10+ years) I would recommend a Vanguard target date fund or one of these: https://www.bogleheads.org/wiki/Lazy_portfolios