I’m often asked if we invest with Betterment, or if I would recommend them.
I assume this is because of their generous affiliate marketing program for personal finance bloggers, but we could generically ask, “Would I recommend using a robo-adviser?” But Betterment’s solution will make a fine reference point.
As with most things, there are pros and cons. But for the purpose of discretely foreshadowing my conclusions:
Betterment has zero of our dollars.
That isn’t going to change.
As with all things related to investing, I take the long view. Where I want to be in 30 or 50 years determines how I invest in the present.
For our dollars, robo-advisers don’t provide a good long term solution.
An Intro to Robo-advisers
I can understand the appeal of robo-advisers for new investors. They offer investment advice without the big fee, automatic portfolio rebalancing, and no trading fees. And in some cases, their tax-loss harvesting algorithms can save real dollars, enough to pay their fees and put a few bucks in your pocket.
For a new investor trying to get started, the sheer number of investment options can be overwhelming. Of the thousands of Mutual Funds and ETFs out there, which ones should we choose? With all of the smart people working on Wall Street, can’t they do better with my money then I can? Or should I just buy Apple or Facebook stock; my brother-in-law said those stocks are up?
The robo-adviser answers all of these questions and more. All you have to do is pay them a small fee of 0.35% to 0.15% of total assets, and their computer systems take care of everything. Compared to a traditional financial adviser charging 1.0%, this is a real bargain. Plus you can do everything online; no need to go into a stuffy office building ensconced in leather and mahogany, where everything seems to be designed to make you feel inferior.
Lower fees are great. Betterment shows how their lower fees result in superior performance over the medium term for a $100k+ IRA. From end of Jan 2004 to Dec 2015, cumulative returns for the Betterment portfolio are 115.6%. By contrast, the traditional adviser portfolio returned only 67.7%.
For taxable accounts, Betterment even offers a nifty algorithm that automagically harvests losses that occur as part of daily market fluctuations.
In many cases this can result in a tax deduction of up to $3,000/year, worth $750 in real money for someone in the 25% tax bracket. And this happens even as your account rises in value!
With a $100,000 portfolio, Betterment’s fees are 0.15% or $150. Last time I checked, $150 was less than $750 by 6 Benjamins. Sweet!
Yes, I can understand the appeal of Robo-advisers for new investors.
But these are expensive training wheels.
Critiquing the Robo-adviser
Looking past the shiny exterior, the robo-adviser offerings lose some appeal due to performance, quality of advice, fees, and the realization of tax savings.
No honest performance analysis can be done without comparing to the index. A comparison to a high fee adviser is a weak straw man argument.
Using total return data from Morningstar, during the period in which Betterment’s 12-ETF 100% equity portfolio returned 115.6%, the S&P500 returned 131%. An investment in VTI / VTSAX would have returned 141%.
While their use of “Nobel winning” Modern Portfolio Theory and back-tested data is impressive, and their selection of low-fee Vanguard funds as a base should be applauded, the results lag.
Henry Ford reportedly once said, “Any customer can have a car painted any color that he wants so long as it is black.”
Robo-advisers are the Ford Motor Company Model-T of the investment world.
Robo-adviser advice is in the form of target asset allocation; using Modern Portfolio Theory and back-tested data, the investment models tells you exactly which ETFs to buy and how much.
To someone (understandably) feeling overwhelmed with investing, this can relieve a great burden.
Or… the same advice is published on their websites. For free. If you want it.
Our own portfolio significantly outperformed the target Betterment portfolio, not because we have any investment prowess, but because of some questionable robo-choices.
The equity portfolio is invested heavily in International equities, even for people who never plan to step outside the US (54%? That seems an odd choice, especially when 1/3 or more of S&P500 revenue is already sourced Internationally.)
One example portfolio suggested we hold 44% bonds… nearly guaranteeing our senior years are spent in poverty (68% success rate in cFIREsim.) Yes I’m an outlier… Aren’t we all?
In return for their advice, Betterment charges a fee as a percentage of Assets Under Management.
Fees start at 0.35% for accounts with less than $10k, and drop to 0.15% after account values exceed $100k. This is in addition to the ETF loads (~0.09% for the 12-fund 100% Equity allocation) which brings total fees to 0.24%-0.44%.
By contrast, a traditional financial adviser may charge 1.0%, just for advice and the occasional lunch or parking validation. Lower is often better (generic Oreos perhaps being the exception.) On the other extreme, Vanguard charges 0.05% for VTI / VTSAX.
Naturally people are wary of fees. This is usually where someone mentions that we shouldn’t worry about fees, because “tax loss harvesting.” ;) ;)
(Potential) Tax Savings
If all works well, for portfolios less than $300k-$500k tax loss harvesting can save more in taxes than the robo-adviser charges in fees.
But fees are guaranteed. Tax savings are not.
What is Tax Loss Harvesting?
Tax Loss Harvesting is the process of selling an investment that has declined in value, in order to gain a deduction for income tax purposes.
In some cases up to $3,000 of capital loss per year can be used to offset other income. At the 25% marginal rate this is a real savings of $750.
This can be done manually, by selling an investment that has lost value. Or it can be done with an algorithm.
Both have limitations.
Manually Harvesting Losses
In a down market, manually harvesting losses is an easy process.
I did it myself as part of year-end portfolio rebalancing in December 2015. It took less than 7 minutes and cost $15.90 in trade fees. Compared to having Betterment do it for me, I earned at least $25,000 per hour (Multi millions * 0.15% >= $3,000 / 7 minutes >= $25k+)
With hourly rates like that, it pays to get a basic investment education.
What about in a bull market?
Substantial New Investment Requirements
Due to the relentless upward march of the stock market, it becomes increasingly difficult to harvest losses without significant new investment.
If we purchased Investment A for $100 and it is now worth $200, the market must drop 50% before a loss can be realized. With new investment at the $200 price point, some loss opportunity is available.
For a $3,000 tax loss, we would require either a large investment or a large stock market correction. ($30,000 * 10% = $10,000 * 33% = $3,000.)
Since we already contribute to a 401k and IRA to minimize today’s taxes, contributing these substantial sums to a taxable account requires a fairly high income and aggressive savings rates.
When we sell Investment A to realize a loss, we can’t just repurchase Investment A or any substantially identical investment within 30 days. In IRS parlance, this is called a wash sale. Instead, we must seek a different investment that we hope has similar performance, Investment B.
In the event of a wash sale, which can occur from something as simple as a dividend reinvestment, the loss is disallowed for income tax purposes until we resell at a future date.
However, if the wash sale occurs in an IRA, the loss is disallowed permanently. (Ouch!)
A wash sale doesn’t have to be in our own investment account. Even a spouse purchasing a similar investment is enough to trigger a wash sale.
To avoid wash sales with a Tax Loss Harvesting Algorithm requires one of two things.
We must either:
a) allow Betterment to manage 100% of our assets, including our IRAs
b) ensure our non-Betterment portfolio holds none of the ETFs, or ETFs that track the same index, as the Betterment portfolio
Option a comes with additional fees.
Option b comes with more complexity (which usually means more fees.)
Capital Gains are First in Line
If we ever realize a capital gain, capital losses are first applied to capital gains before they can offset other income. Selling assets that have increase in value is a recommended best practice for readjusting asset allocation, so capital gains are a norm.
At the 25% marginal rate, a $3k loss was worth $750.
But long term capital gains are taxed at a lower rate, from 0% to 15%.
At 15% a $3k loss is worth $450. At 0% it is worth zero.
Long Term Implications
Long term, my goal for investing is to build wealth and establish a financially secure retirement.
Life long fees and low basis don’t help.
With no earned income and tax free long term capital gains, tax loss harvesting has minimal to zero advantage. All fees must be paid with cold hard cash.
That means forking over $1,500 per year for each million in investments. Or put another way, it is 4% of total spending for someone following the 4% Rule.
cFIREsim is a powerful tool for analyzing portfolio longevity, and is very helpful for determining the long term impact of a 0.15% increase in fees.
A $1 million portfolio with $40k annual inflation-adjusted withdrawals and 0.09% annual fees has a success rate of 95.7%. Increasing the fees to 0.24% reduces the success rate to 94.8%. A 1% reduction in success rate isn’t substantial.
But the impact to median terminal value is massive! Over 30 years, assuming equal performance the higher fees reduce the average terminal value by ~$200,000! That is 20% of the initial portfolio value!
This, based on a decision we made as kids some 30 years ago, when we didn’t know anything.
Actively harvesting tax losses over an extended period reduces overall basis.
During retirement we want the opposite. This is why we are actively raising our basis by harvesting capital gains, all part of our plan to never pay taxes again.
When basis is low, a greater percentage of our wealth is considered income when we spend it. As part of total income, it is subject to taxation and may impact healthcare subsidies and taxation of Social Security.
All things considered… the prevalence of free high quality asset allocation advice, lower performance than our own simple portfolio, questionable robo-choices, the substantial impact of higher fees to portfolio long term value, the ease of do-it-yourself tax loss harvesting while working, the limited to zero value of tax loss harvesting in retirement, and the advantages of high basis, I don’t see any advantage to having a robo-adviser manage our portfolio.
This is why Betterment has zero of our dollars.
This is only half of the question, of course. Also to be fair, Betterment would say I’m not their target customer. Would I recommend robo-advisers for someone else?
There is a certain customer that should consider adding their assets to the $3 billion that Betterment already manages.
- Suffering from Paralysis of Analysis
- Plans to work to traditional retirement age
- Zero interest in learning about $; even 2 hours is too much
- Earns enough to save $35k+/year/person while still falling in the 25% tax bracket ($50k+/single, $100k+ for a married couple.)
If you match this profile, that’s cool. Investing via a robo-adviser is better than not investing at all.
I think I’ll apply for one of those affiliate links and paste it right here.
Update: GCC now has an affiliate relationship with Betterment. If you sign up for their services with this link, we will receive a commission. Thank you.
Update #2: I’ve been asked via email if there are good Asset Allocation books. Here are two that are popular and probably in your local library. Feel free to recommend others in the comments.
The Bogleheads’ Guide to Investing (Simple DIY portfolios and a great resource for all investors)