Robo Advisors

Lately, I’ve been thinking a little about robo-advisors. Really, these are just like those target date retirement funds that have been around for the last few years, except there are even fewer people working there. Or, they’re like a fund of funds. They take your money, invest it in the assets they pick, and use computer systems to monitor them, to make sure that asset allocations are within the bounds set by the firm (and for other features we’ll talk about later.)

The biggest feature of robo-advisors is that they’re not opinionated. The computer doesn’t have any idea if the stock market will do well this year or not. It doesn’t care; it just makes sure that $65\% \pm 2.5\%$, or whatever, of your assets are in stocks. On the one hand, this means that your assets aren’t going to be subject to the irrational, emotional reactions of some human money manager. On the other hand, it also means that you’re giving up on “beating the market.”

I’ll spend the rest of this post talking about some of the features of robo-advisors that are most interesting to me. I’m not going to make any recommendations, of course.¹

First, some assumptions:

1. You can’t, with high confidence, beat the market in the long run.²
2. To the extent that market inefficiencies do exist, transaction costs (especially the bid-ask spread) will make it impossible for retail investors to take advantage of them.
3. Investing in financial instruments is worth my time. While there are a number of scenarios in which financial instruments suffer large drawdowns, most of those would also prevent me from retiring comfortably anyway.³
4. Real estate might be a good thing to invest in eventually, but the high leverage, high transaction costs, large initial investment required, and low diversification make it impossible right now. So ignore it for the moment.

Passive investing

Given that I can’t figure out what money managers are going to be the best in the future any better than anyone else can, I should be investing passively. That is, I should be purchasing broadly across the asset classes I want to own. I shouldn’t be trusting some particular person or company to figure out exactly which are the right stocks to buy; I should just buy a little bit of everything. In other words, I should buy an index fund.

All the robo-advisors do this, and so do a whole lot of individual investors. In fact, it’s the default option in a lot of 401(k)’s these days. Target date funds are simply a special class of these that also do some asset allocation and rebalancing over time. This is probably where the big wins for the robo-advisors come in, but those wins are easy for people to get without paying an advisor. Just buy a decent target date mutual fund and move on.

Fees matter

One of the reasons that passive investing wins is because it’s usually cheaper. Over time, higher fees will eat away at earnings.⁴

The robo-advisors, though, will charge fees on top of the funds’ fees.⁵ So, that’s a point against the robo-advisors in general.

A crowded field

There sure are a bunch of these products springing up all over the place. Nerd Wallet puts the count at a couple hundred.

The fees that they charge, though, vary significantly. Betterment uses a sliding scale that’s on the low end. SoFi charges a flat fee (which is higher than Betterment’s for small account sizes). WiseBanyan is “free,” but you’ll have to pay (more than Betterment) if you want tax-loss harvesting, backdoor Roth IRA contributions, or a bunch of other stuff. Wealthfront charges more than Betterment for large accounts but has some cool features that might set it apart. Vanguard charges more but lets you talk to a person.

Schwab is also “free,” but uses its own ETFs and keeps a large share of assets in cash. They make money off the operating expenses of the funds and from lending out the cash. (Schwab also operates a bank.) Some people seem upset about this, but, though I don’t agree with a lot of what Heinlein said, TANSTAAFL.

There are a bunch of reviews out there, and even head-to-head comparisons. That’s great. Be cautious about the comparisons that put money into multiple accounts for a couple months and then see who “came out ahead.” If you’re in the market, you should probably be in it for the long term.

Exchange Traded Funds (ETFs)

Most of the robo-advisors use Exchange Traded Funds, or ETFs. There are some good reasons for this:

• ETFs don’t have as much turn-over as open-ended mutual funds.
• ETFs can be more tax efficient than mutual funds, too.
• ETFs often have lower operating expenses.
• ETFs can be traded intra-day.
• ETFs have a lot of assets under management, which means that the industry, as a whole, is unlikely to disappear.

On the other hand, ETFs are relatively new, and that means there’s a lot of uncertainty if you’re investing for the long term. Because ETFs trade independently of their underlying indices, there can be tracking error: the market cap of the ETF can differ from the net asset value of the fund. This is impossible for a regular, open-ended mutual fund. The tracking error is likely to be particularly severe when markets are unsettled, such as 2009, exactly when you’re most worried about your money.

Some ETFs are also very thinly traded, leading to larger bid-ask spreads and larger tracking errors. The good news is that most of the robo-advisors seem to stay away from these smaller ETFs and stick mostly to the big, well-known players in large, liquid asset classes. This is one reason, by the way, why you might not want to go with the advisor who offers you the most asset classes: some of them will be thinly traded.

But, there’s another consideration. If you’re a long-term investor, saving for retirement, a long way away, you might be able to ride out any market volatility. In other words, part of the risk control you might want to undertake as your investment horizon comes closer is moving out of ETFs and into more proven investment vehicles that don’t have quite as many unknowns. So, maybe ETFs are a good vehicle for retirement savings?

Taxes

Many robo-advisors claim to save you money on your taxes by performing “tax-loss harvesting.” There are some opportunities for this, since the market moves up and down every day and the bid-ask spread on highly liquid securities can be quite small.

I don’t know how useful this is. Some firms claim it can add up to 200bps of return per year, but that sounds a bit extravagant. Because of the wash sale rule, the robo-advisors move money from one ETF to an ETF that tracks a similar, but non-identical index. So, essentially what you’re doing is taking a tax deferral on your capital gains, assuming you had some that year. That’s a good thing, but it’s not like the tax will never have to be paid. Assuming you have a long time horizon, though, that can be a big help.

Of course, in a tax-deferred or tax-exempt account like an IRA, none of this matters at all. You pay for the capability, but can’t take advantage of it.

Finally, there are direct indexing services such as Personal Capital and Wealthfront. They don’t use ETFs for part of larger accounts and instead directly purchase a representative sample of the underlying stocks. There might be more tracking error here, but there’s also a lot more opportunities for tax loss harvesting. This seems particularly attractive if you have a large enough taxable account to qualify, since it also reduces the chances of one of your ETFs trading at a significant discount to NAV when you need to sell.⁶

Style and size weighting

Fama and French argued that size and price-book ratio were important measures of risk, besides just the market beta ($\beta$).

Be that as it may, it often pays to load up on risk in a retirement account. A higher risk premium should lead, over, say, 30 years, to higher total returns. So, I’m interested that some of the robo-advisors are offering portfolios that over-weight small cap and value stocks.

I’m also a little suspicious of having so much of one’s personal worth tied up in Apple, Exxon, and, if that IPO ever happens, Saudi Aramco. Maybe I have an irrational belief in mean-reversion, but it seems unlikely to me that such successful firms can continue to be successful forever.⁷ That’s not to say Exxon will suddenly disappear, but it’s been enormous for over a century.

Of course, ETFs tracking indices that are weighted by something other than market cap are more expensive: their operating expenses can be ten times as much as the least expensive ETFs.

Going it alone?

I could always do something similar without paying an advisor. There are plenty of discount brokerages that would be happy to take my money and put them in indexed mutual funds. I could rebalance, say, twice a year, and be careful about avoiding excess trading.

Of course, I could have been doing those things for years and yet I haven’t. The rebalancing by itself might be worth some amount of fees, though, as I mentioned before, I’m still a little wary of ETFs.

Conclusion

Ideally, I was hoping to have come to some sort of conclusion by now. A real conclusion, I mean, in which I resolve on a particular plan and then go execute it.

I haven’t. The difficulty with all of this is that it’s unclear what will happen in the future. Will Roth IRAs stay tax exempt, or will Social Security be means-tested based on them? Will ETFs remain a fixture of the retail investing world, or do they have undiscovered problems?

That said, there are some features of some of the robo-advisors that I think are especially interesting. First, direct indexing. The rebalancing algorithms worry me a lot less than the financial instruments, so direct indexing seems like a great feature.

Second, fundamental indexing. It seems like there’s a real opportunity there to add some risk for a higher return without hitting the discontinuities that leverage can give you.

It’ll sure be interesting to see how this shakes out.