Inside Robo Advisor Tax Loss Harvesting

Nobody’s an average Joe when it comes to taxes.

Director of Research
Reviewed by: Elisabeth Kashner
Edited by: Elisabeth Kashner

Nobody’s an average Joe when it comes to taxes.


This is the sixth blog in a multiple-blog series by’s Director of Research Elisabeth Kashner on the new “robo advisory” industry. The previous five, in the order in which they appeared, are as follows: 

“I rarely use oxygen myself, sir. It promotes rust.”

—Robby the Robot (Forbidden Planet)

Wealthfront, Betterment and FutureAdvisor trumpet their tax-loss harvesting and rebalancing services, claiming that their automated systems deliver hefty, seemingly risk-free returns.

If this is so, I’ll recommend one of them for managing my teenage son’s bar mitzvah money, or at least put a memo on my calendar for nine years from now, when, please God, he’ll begin paying taxes. But as always, I’ll have to make sure I understand the ins and outs, so that I can back up my recommendations with solid analysis.

So I invite you, taxpayer and investor, to come along with me on a deep dive into tax-loss harvesting.

We’ll survey estimates of tax-loss harvesting’s value, and then walk through a sample tax-loss trade. We’ll see how many variables are at play, so you can decide if harvesting would help you, with your current tax bracket, future earnings expectations, capital markets expectations and estate plans, if any.

How Much Is Tax-Loss Harvesting Worth?

Wealthfront’s home page claims 1 percentage point of annual tax-loss harvesting returns, plus 0.40 percentage points per year for rebalancing.

But Betterment goes even further.

Jon Stein, Betterment’s chief executive officer, told me that “Betterment’s aggressive tax-loss harvesting can be worth 2 percentage points per year, and that tax-aware rebalancing is worth as much as tax-loss harvesting in certain markets.”

FutureAdvisor provides a chart showing 3.03 percentage points year of tax savings, from single-stock switches around the S&P 500 Index, which is not the same as the firm’s ETF-switching service.

These are some big numbers—big enough to convince would-be robo-investors to demand tax-loss harvesting and rebalancing services, and to build a huge competitive advantage for the firms that do it well. Of the six robo-advisors who offer all-ETF portfolios, three don’t currently have tax-loss harvesting programs.

Should they?


Not so fast, says Michael Kitces, financial advisor and the author of the “Nerd’s Eye View” blog, who writes that tax-loss harvest yields could be as little as 2 basis points a year, and that the practice is far from riskless.

You read that right: Kitces thinks that the robo advisors’ estimates are up to 50,100 and even 150 times too high.

With my son’s bank account on the line, I’ll need to sort through these claims. I’m going to focus on tax-loss harvesting this time, and leave rebalancing for another day. It’s messy work, so please expect to get a little nerdy as we walk through the mechanics of a tax-loss harvest.

Tax-Loss Harvest Example

Tax-loss harvesting adds value by deferring tax payments, allowing clients to invest the tax savings. But the deferral isn’t forever, and it comes at a price: death, or taxes. Harvesting lowers your cost basis, increasing your future capital gains taxes, unless you manage to pass along the lowered-basis stocks to your heirs.

Let’s look at a simple hypothetical example. I’ve laid out the trades and taxes of my straw man, Joe D. Investor, in the table below. This model applies the top long-term federal capital gains rate, but excludes state taxes, for simplicity’s sake.

ActionInvestment ValueTax Savings ValueTotal ValueVariable
Buy VWO (primary fund)100,000 100,000 
After one year, VWO loses 10%90,000 90,000Time elapsed, % investment change
Sell VWO, buy IEMG (secondary fund)90,000 90,000Transaction costs
Calculate tax savings (20% of $10,000 = $2,000)   Your tax rate, holding period, availability & type of capital gains to offset
Invest the tax savings in IEMG (secondary fund)90,0002,00092,000Reinvestment returns
After another year, IEMG regains the initial loss. Sell IEMG100,0002,222102,222Time elapsed, % investment change
Pay Capital Gains Taxes of 2,000 (20% of $10,000) plus $44 (20% of 222)98,0002,178100,178Your tax rate, holding period
Net returns  0.18% 
Annualized net returns  0.09%Time elapsed, % net returns
Opportunity cost  0.78%Returns difference between primary and secondary fund


The story goes like this:



Joe D. Investor buys $100,000 of the Vanguard FTSE Emerging Markets (VWO). VWO is the “primary” emerging market ETF at both Wealthfront and Betterment. Poor Joe’s investment loses$10,000 in the first year, so Joe sells VWO and buys the iShares Core MSCI Emerging Markets (IEMG), Wealthfront and Betterment’s “secondary” emerging market ETF.

Joe can now claim the $10,000 as a long-term capital loss. If Joe has $10,000 of realized long-term capital gains to offset, then he just saved himself $2,000 in taxes, assuming a 20 percent capital gains tax rate. As I mentioned earlier, the savings will be greater once you factor in state capital gains taxes.

Joe is on the ball, so he actually invests the $2,000 savings. To keep things simple, we’ll say he doubles down and buys more IEMG.

Then Joe gets lucky; IEMG rebounds. His initial investment is once again worth $100,000, and the $2,000 harvest is now worth $2,222.22. Joe sells the whole thing.

Now Joe owes taxes on his capital gains. He gained $10,000 on his first tranche of IEMG, and $222.22 on the second. At the same 20 percent tax rate, he now owes $2,044.44 in capital gains taxes (assuming at least a 12-month holding period). He withdraws the doubled-down $2,000, and takes the $44.44 out of his $222.22 gains, netting a total of $177.78, assuming no trading costs along the way.

So, over a two-year period when Joe’s emerging market investments went nowhere, Joe used tax-loss harvesting to earn 0.18 percent, before trading costs. That’s 0.09 percent per year.

But is that 0.09 percent pure profit?

VWO and IEMG track different indexes. The difference is South Korea and small-caps—IEMG includes them, while VWO does not. From Jan. 2, 2008 through Aug. 7, 2014, this has led to a median rolling one-year index returns differential of 0.78 percent. Sometimes VWO outperformed; other times IEMG prevailed.

But one thing is for certain: They’re not the same. The trade most likely will not run afoul of the Internal Revenue Service’s “substantially identical” rule, but will come with opportunity cost. In fact, for 95 percent of the rolling one-year periods measured, the performance gap between the two indices was larger than our hypothetical 0.09 percent tax-loss harvest return.

Conditions Matter

Joe D. Investor’s story stared with a randomly unlucky trade. Did you notice other sources of randomness? Plenty of situations could have changed Joe’s overall returns, making them more—or less.

The biggest variable is probably tax rates—both state and federal. The higher your tax rates at the time of the harvest, the larger your tax savings. The opposite holds at the time of liquidation, when you face your lowered cost basis. Joe’s liquidation rates could be high, if, for example, he’s in his peak earnings years, needs to use taxable savings to send his kids to college, and also happens to be in a tax-the-rich political climate. On the other hand, if Joe were a retiree with no income and a mortgage, he’d beat the taxman.

Betterment’s and Wealthfront’s tax-loss harvesting white papers point out that Joe can get out of the liquidation tax by dying or giving his investments to charity. Congratulations?


At the extreme, if Joe were in the 15 percent federal tax bracket, he would pay zero percent on capital gains. In that case, Joe should harvest gains, and reset his basis, lowering future tax bills. He should never harvest losses.

The type of gains offset matters too. If Joe happened to face short-term capital gains taxes, and were in a high tax bracket, he’d save real cash by harvesting losses and avoid paying ordinary income rates, a 3.8 percent Medicare surcharge, and state taxes.

If not, Joe might have long-term gains to offset. He can also offset up to $3,000 of ordinary income. Otherwise, Joe would just carry the loss forward. It has no immediate cash value, and therefore produces no tax savings to invest.

The second-most-important variable is time, because the longer Joe can invest the tax savings, the greater the power of compounding. Both Wealthfront and Betterment use time to their advantage in their tax-loss harvesting white papers. By starting their analysis in 2000, with the tech crash, they produce most of their tax savings at the outset of their backtests, and allow it to compound through 2013.

Lastly, there’s the pattern of investment returns to consider. Tax-loss harvesting only works if you can sell investments at a loss. Investing itself works under the opposite conditions. For both to work, you need volatility and a strong stomach, and you also need your losses to precede the gains.

The takeaway here is that an awful lot has to go right for investors to realize sizable benefits from tax-loss harvesting. And some of it isn’t really predictable ahead of time.

The robo advisors acknowledge this—kind of.

Betterment publishes a “who benefits most/least” analysis, plus a substantial disclosure section in its tax-loss harvesting white paper, laying out the many variables that affect its returns. Wealthfront's is a bit coy, emphasizing cases where tax-loss harvesting works well, but burying mentions of variability in the fine print disclosure section.

Better Estimates?

Wealthfront’s tax-loss harvesting white paper shows actual client results of 0.40 percent “tax alpha” over a 13-month period from October 2012-2013, not the 1.00 percentage point that Wealthfront posts on its home page.

And still, there’s a catch: Wealthfront’s tax alpha examples ignore the eventual sale of the assets, for which Michael Kitces, the blogger behind “Nerd’s Eye View,” has taken them to task. Presumably, Kitces would do the same to FutureAdvisor, which also presents tax alphas rather than internal rates of return that capture terminal tax liabilities.

Tax alpha measures the savings from the first stages of the tax-loss harvest, but not the increased tax liabilities from the eventual sale of any lowered-basis securities. Wealthfront switches back to the primary after 30 days. Unless the secondary regained all of the primary’s losses during that 30 days, the primary’s cost basis will be lower than it was before the harvest.

Betterment’s tax-loss harvesting white paper runs a range of liquidation examples, giving more information than Wealthfront’s. However, Betterment did not supply actual client results, but instead presented a backtest.


As it happens, Betterment’s look-back period started just months before the 2000 tech crash and ran through 2013, providing a huge initial loss, and lots of time to compound any invested tax savings. Wealthfront presented backtested results from a similar time period, along with actuals.

When I asked Betterment for the results of their Monte Carlo simulations that aren’t tied to the 2000-2013 time period, its spokesman demurred, explaining in an email that: “Ultimately, the range of benefits varies tremendously depending on time horizon, and deposit schedule. In the white paper, we chose to focus on real data, but we will publish results based on forward-looking projections as well, and can share at that time.“

In the end, tax-loss harvesting doesn’t have an absolute value, not the way that expense ratios, trading costs or even asset allocation does. There are too many variables involved, many of which are unknowable at the time of the harvest.

Going forward, I would advise that prospective robo investors think carefully about their current and future tax brackets, their time horizons and their volatility expectations. Also weigh the tax-loss-harvest expectations against the opportunity costs of switching to the second-best, non-“substantially identical” ETF.

I can’t resist one final irony.

Wealthfront, Betterment and FutureAdvisor all chose VWO as their primary emerging market ETF, with IEMG as their secondary choice. If I were to invest in either, I’d be happier post-harvest (except for the bit about taking a loss), because, as I wrote blog No. 5 of this series, IEMG is a broader-based, better-run fund.

My happiness would be short-lived at Wealthfront, where they switch back to the primary after waiting out the 30-day wash-sale window, but potentially infinite at Betterment, where they only switch back if there are further losses to be harvested. But let’s not think about that part.

The Bar Mitzvah Boy

As my husband and I wrap up our process of helping our son choose a robo advisor, we will ignore tax-loss harvesting, because our 13-year-old doesn’t owe any taxes, and probably won’t for at least nine years. Moreover, he doesn’t yet need bonds, so most of his assets will be highly correlated, and therefore offer fewer harvesting opportunities.

We’ll counsel him to think hard about the asset allocation bets he wants to make, and the risks they entail, showing him the costs and philosophies of each robo advisor. In my seventh and final blog of this robo advisor series, I’ll let you know what he chooses, or if he decides to go it alone.

At the time this article was written, the author held no positions in the securities mentioned. Contact Elisabeth Kashner, CFA, at [email protected].



Elisabeth Kashner is FactSet's director of ETF research. She is responsible for the methodology powering FactSet's Analytics system, providing leadership in data quality, investment analysis and ETF classification. Kashner also serves as co-head of the San Francisco chapter of Women in ETFs.