Comment by hristov | Hacker News Reader

Comment by hristov | Hacker News Reader

Parent Full thread

I agree with most of this, except for the stuff on robo-advisors. You should be very careful about those.

First of all their fees are too high. Wealthfront's 0.25% fee seems rather small and it is smaller than what a lot of human advisers charge, but if you compute it over a lifetime of savings with the negative compounding effect it will cost you a lot.

Imagine you receive some money when you are 20 from a rich uncle and invest it for 40 years using the wealthfront fee structure. After 40 years you will have paid about 10% of your savings in fees. Or, in other words, you will have about 10% more savings if you had taken a couple of hours to sit down and decide which funds to invest in. Keep in mind that the wealthfront fees are in addition of any etf or mutual fund fees you have to pay to get into investment vehicles.

So yeah, compounding interest is a dangerous thing.

There is another problem with roboadvisers -- people put too much trust in them. In our society there is this implicit trust of the computer, probably bred from multiple sci-fi shows with all-wise computers. Well it is a very dangerous thing when it comes to your savings.

You may not be the best investor, but you should take responsibility in your investment choices. You should know what you are investing in and why. Even if the thing you are investing in is a boring simple S&P 500 fund (as it should be for most of you) you should know what it is and why you are investing in it. You shouldn't just blindly follow some algorithm programmed by god-knows who.

This. The typical portfolio drummed up by roboadvisors corresponds to basically stone age portfolio theory. With a tiny amount of personal responsibility and education, you can replicate the same and net the difference in fees without spending your life thinking about finance to try to do better.

If you just want to 'set it and forget it', consider its an approach you're taking with the fruits of decades of your life.

[edit: typo]

The impression I get is that the two areas where robo-advisors shine is UX and Tax Loss Harvesting.

The Betterment site is pretty, which makes me more inclined to put more money into it more often. This is irrational, but for me this makes it worth more than the 0.15% I end up paying them in fees.

I'm less sure about whether TLH is worth the 0.15% fee, though. The premise is that you do some "equivalent" (to you, but not to the IRS) transactions, report them on your tax return, and lower the taxes you pay today ("basis"). In exchange, you would have to pay those taxes later on your investments when you withdraw them. Is this always the right answer (because those later taxes are in compounded-inflation dollars)? What if I think my tax rate now is lower than the tax rate in ten, twenty, fifty years?

Sort of. Ideally you're trading future long-term capital gains taxes for lower income tax today. You assume the future cap gains rate will be lower than or equal to your income tax today.

The amount I've already saved in tax dollars thanks to TLH is orders of magnitude higher than what I've paid Betterment in fees.

That being said, you won't always benefit from it, and there are caveats you should read about. I've decided to take a slightly more hands on but simpler approach, and have moved all of my investments into a simple 4-fund portfolio at Vanguard.

As you reach the $3k max and your portfolio continues to grow, the benefit to fee ratio becomes worse.

> Or, in other words, you will have about 10% more savings if you had taken a couple of hours to sit down and decide which funds to invest in.

If and only if you would have done about as well as the robo-advisor in those couple of hours. (I'm not saying this would or wouldn't happen, but it's pretty big for an unstated assumption.)

Well the thing is that the roboadviser is not even intended to make particular security picks. It is only supposed to decide which broad areas to invest in and what percentage for each area. Thus, the roboadviser can say that for your particular risk profile you should have 60% stocks 40% bonds and within the stocks you should have 60% sp500, 20% foreign large cap ... etc.

My point is this portfolio distribution stuff is not an exact science and there really isn't a right answer. There are some broad accepted guidelines, but they are rather broad and simple and you definitely do not need computers to follow them.

maybe you'll do better, maybe you'll do worse. the prevailing theory is that the expected value is the same (minus the fees). adding up years of fees, and suddenly you're short a nontrivial amount of money

This is true to some extent, although I have found that the automated "Tax Loss Harvesting" being offered by some roboinvestors more than pays for their percentage fee.

Should I also be performing the work of checking several times a year if I should be making sales, recording the losses, buying equivalent securities, and bundling that into a form for the IRS?

You are absolutely right on the fees being a problem here. I opened a wealthfront account a couple of years ago and added some money because I really liked the interface and I was sold on tax loss harvesting. I recently wanted to compare how my wealthfront account did compared to the S&P500 and I was surprised to see a 0.98 beta (almost the same monthly returns as S&P500) and an alpha of (-0.15 almost the fees they are charging) with very little r-squared error. Atleast for my risk profile (10), I would have been better off investing directly in S&P500. I didn't get high returns nor did I get reduced volatility with the wealthfront portfolio.

If you posit that all a robo-advisor does is "decides which funds to invest in" for you, how are you putting too much trust into a computer? A target-date fund from Vanguard also makes allocation choices for you across US/international stocks/bonds; is the conclusion that you should always invest in the underlying asset class directly instead of trusting a fund/advisor/company to do it for you?

One thing I very much dislike about many of the roboadivisors inculding Wealthfront, is their subjective take on what risk means.

Ex: Wealthfront's high risk portfolio back in 2013 had significant exposure to commodities -- mainly oil and metals. That sector has done poorly to say the least, and many a retail investor would not have correctly understood what Wealthfront's definition of risk actually meant.

Reply on news.ycombinator.com