What is Wealthfront's Direct Indexing and Will It Deliver in The Real World ?

I have noticed that many of the personal finance sites' reviews of the robo advisors like this one  from Nerd Wallet focus on low fees and simplicity But they are not sophisticated enough to understand the pitfalls in many of their strategies

Nerd wallet gives w\Wealthfront "five stars" but simply takes all their promotional material at face value not delving into the details of the strategy. Low cost, simplicity and what seems to be a sophisticated strategy seem to impress many observers without the inclination or skills to drill down.

There are many issues with understanding the allocation of Robo Advisors using traded ETFs. Wealthfront's strategy is even more complicated than other Robo Adviors. In fact the differences are so complicated that it will take more than one article to cover and took someone with expertise (me)  several hours to understand it.

This article will cover their "simple" direct indexing strategy, The next article will cover their strategy which is the next generation of indexing which they call "passive plus:

What is "Direct Indexing"

The most important difference between Wealthfront and other portfolios using ETFs is that most of Wealthfront''s US stock  assets are invested in "direct indexing". The strategy purports to offer a better alternative by owning a portfolio of constituent stock in the total US stock market instead of the Vanguard total index fund. Their strategy avoids s the wash sale rule with little impact on the total return and they claim adds to after tax return. A simple strategy of swapping ETF (explained below) avoids the wash sale rule much more simple.

By owning individual stocks Wealthfront claims the benefit of avoiding the management fee of the Vanguard ETF  tottl market ETF(. which is 05% ) is saved that there are more opportunities for tax loss harvesting through buys and sells of individual stocks instead of a single ETF.

Wealthfront claimed their strategy would generate an after tax alpha of 1.55% in return over 3-5 years vs the total stock market index. You can read through the assumptions used here on their whitepaper but it assumes the most favorable conditions possible strategy in terms of tax brackets and compares their strategy to the "straw man" of an alternative portfolio that does no tax law harvesting. This a 2014 "improvement over the .90% claimed in 2012. Of course there is no actual data for these strategies.


But do other hidden costs outweigh any benefits that come from tax harvesting and are the numbers given by Weathfront about the gains from tax loss harvesting realistic.

I will concentrate on other pitfalls of the strategy and how the purported results are presented

Is the strategy really that simple ?

Substituting stocks in the direct indexing portfolio

The Direct Indexing  strategy incorporates individual stocks and not ETFs.  There was  quite a bit of computer and human power used to define the strategy..but lots of things that could go wrong in the real world that will make it unlikely it meets its expectations.

Here is how Wealthfront describes its Direct Indexing

The next evolution in index investing

In contrast this is what Wealthfront  does with Direct Indexing

By replacing VTI with a combination of individual securities, you no longer pay the VTI expense ratio. This means the combined portfolio of individual securities and a completion ETF has a lower combined expense ratio.
More than 80% of the Advanced Indexing position is comprised of individual stocks and thus has no expense ratio. The remaining 20% or so is comprised of either the Vanguard Extended Market ETF (VXF) for $500,000 accounts or the Vanguard Small-Cap ETF (VB) for accounts over $1 million. The combined portfolio remains subject to Wealthfront’s advisory feeOur minimums for Advanced Indexing are based on dollar amounts required to hold a reasonable collection of individual US stocks while continuing to track the performance of the broad US market. We found that holding hundreds of stocks requires a US equity allocation of roughly $150,000 – thus necessitating an account minimum of about $500,000 (because approximately 30% of our portfolios are allocated to US Equities). This minimum increases again as you exceed 500 stocks and start holding some of the smaller mid-cap stocks in the position – thus necessitating a $1 million minimum for the higher level of Advanced Indexing

But is direct  indexing so easy to execute in the real world ?

Here is an example of a trade Wealthfront says it would execute in the direct indexing strategy:

stocks in the same industry tend to perform similarly over short periods of time. So if Coca-Cola misses its earnings guidance and trades down 10%, our Direct Indexing program will sell shares of the Coke and temporarily buy more shares of Pepsi to replace it. That way, our client would maintain exposure to the beverage industry, while still realizing tax losses and lowering her overall tax bill. After 30 days, the program would sell Pepsi and buy back Coke, but only if Pepsi doesn’t trade up because that would lead to a realized gain which would cause a tax liability

If this is how Wealthfront decides how to do its "paired trades it is in trouble even what seem to be similar stocks a close look shows how dissimilar they are Pepsi earns way over 305% of its income from non beverage products (Frito Lay) Coca Cola has only beverage products

Here is a chart of Pepsi (PEP) and Coca Cola . It's quite clear that even in the swap described moving between the stocks can be as much as a trade as it is a simple tax loss harvesting study.

PEP (Pepsico) Brown KO (Coca Cola) black

But this isn't how tax loss harvesting is generally done with ETFs. The more widely used strategy doesn't "hope" for two stocks in the same industry to move as described. It is designed to immediately realize losses in one etf and trade into one that is essentially the same without changing the long term investment strategy..and little need a scenario like the one described above. Swapping between ITOT and VTI would keep the investor's allocation in the total US market  while creating a tax loss.that could be deductible with the position purchased ultimately liquidated at the lower long term capital gains rate (although the bais on the position has gone down)It isn't dependent on one stock going down while the paired stock goes up..and then the paired stock recovering.

It Gets Worse: Pairing Stocks in Present Day Markets

In today's markets the stocks with the highest weights in the total stock market index don't have even the correlation Coca Cola and Pepsico might have

What are the industries in the most highly weighted stocks in today's market and what could be the other part of a paired trade in tax loss harvesting  ?

Apple has just introduced its newest line of Iphones a much awaited event. The announcement may trigger a significant move in the stock after the announcement or when the product comes to market or if the introduction is delayed. If the move is significantly to the downside what stock could be used as a pair in tax loss harvesting

Apple is the largest holding in VTI at just under 3%. other top holdings in VTI are Amazon and Facebook both around 1.5% and Google/Alphabet at 2.2%. and Berkshire Hathaway is 1.34 That total is  8% of the index. What stocks in the "same industry" would be used for tax swaps akin to the one in Pepsi/Coke described above. ?

 And even within the same industry nothing is as simple as the Coke Pepsi trade example.Wells Fargo and JP Morgan  are the two largest banks in VTI. In mid August Wells Fargo announced it had discovered 1.4 million more fake accounts putting the total at 2.5 million..and says it may find more. Obviously despite being in the same industry JP Morgan(JPM) and Wells Fargo(WFC) are not trading similarly as can be seen in the chart below:

Wells Fargo (black) JP Morgan (brown)

More complications

Transaction costs. As Robert Arnott notes in his excellent paper Live from Newport Beach It's Smart Betat factor that can explain disappointing live performance is implementation cost. Implementation costs are contributing to an ever larger portion of the gap between the expected performance of a smart beta index and its live record as the total amount of assets managed by these strategies rapidly grows.. 

The costs associated with executing a strategy are both explicit and implicit. The explicit costs, such as brokerage commissions and settlement/clearing charges, are directly observable, and explain a significant part of performance slippage, or the amount a fund’s return underperforms the index it is tracking. The implicit costs, referred to as market impact costs, are the changes in a stock’s price around index rebalancing dates, especially when the strategy’s assets under management are large; that is, the prices of stocks being purchased are temporarily inflated, and those being sold are temporarily depressed. As prices revert in the days following the rebalancing, the strategy loses money. This outcome is not easily observable in smart beta strategies because the impact is embedded in the return of the underlying index, whose value is calculated on the basis of closing prices.

For strategy implementers, whose primary goal is reducing tracking error, a rational response is to lump all trading around the market close so the portfolio can perfectly track the index. These clustered trades also happen to be the most costly because they are reducing, within a very short time span, already-limited liquidity. Chow et al. (2017), after studying various portfolio characteristics related to implementation, recommend spreading trades over several days around the rebalancing, if possible.

Another way to lower market impact costs is to avoid .. strategies that invest in stocks with low liquidity. Screening out micro-cap and thinly traded companies’ stocks is an important step in ensuring a strategy is “tradable,” even before considering the market impact of trades. Alphas produced “on paper” cannot successfully be reproduced when, for example, $10 million in buying power is attempting to take advantage of a mispricing opportunity in a stock of a company with a total market capitalization of $5 million. When they conduct simulations, thoughtful researchers will consider the trading volume of a stock, as well as set up proper constraints on the trades required by the strategy.

Investors who allocate to strategies, such as high dividend-growth, that typically require holding a relatively illiquid subset of the universe (Chow et al.) can apply a haircut to backtest results when setting their forward-looking return expectations. Illiquid stocks do offer more mispricing, and thus profit, opportunities, on average, because the price discovery process for these stocks is generally slower. Being attentive to the potential that the paper alphas of these strategies will likely be lower when they are live can shield investors from unpleasant surprises.

Strategies with high turnover rates, or when turnover occurs only for a few stocks rather than across the entire portfolio, also tend to experience high implementation costs. If this product feature is necessary to deliver the outcome investors seek, and no product design changes can address it effectively, sophisticated implementers can use algorithms to tactically take advantage of available liquidity. A momentum strategy falls into this category. For this reason, incorporating momentum in a passive smart-beta index strategy is very challenging

With $4.65 billion under management and growing and portfolios including hundeds of stocks and a strategy designed to execute a large number of trades in volatile markets,,it's hard to belive that transactions wont be an issue in matching real world results to those in the research paper. Although Wealthfront includes an estimate for transaction costs for their passive plus strategy I couldn't find an assumption for this.

As in many presentations of the expected success of a strategy there is no consideration about how markets work in the real world . as Arnott notes These studies are based on closing prices and make no consideration of bid/ask spreads/transaction costs. If only the world traded that way !. But of course it doesn't every trade involves a bid/ask and direct indexing involves hundreds of trades each year. 

More Complications Tax Loss Harvesting in Emerging Markets

 The Wealthfront strategy has many flaws when Wealthfront uses it in Emerging Markets where it swaps between VWO to IEMG.

The difference in holdings between the two ETFS is quite significant  VWO has no holdings in South Korea IEMG has 15% of its holdings in South Korea

Following the links above you can find other significant differences in the portfolios

Not surprisingly there is a gap in performance. Below are the twelve month numbers:. Returns on top volatility below.

The gap in performance means that it is likely not to be able to do a tax swap : realizing a loss in IEMG and then purchasing VWO one is not purchasing identical portfolios so the swap will not be made without what is essentially a change in strategy (because of the allocation)

VWO green IEMG Blue Returns top volatility below

How did Wealthfront develop its strategy (improved in 2014) ?

 The Dangers of Backtesting and Data Snooping

As Wealthfront explains its claims were

...based on both back testing our algorithms, as well as through detailed Monte Carlo simulations of thousands of possible market scenarios

Put in simple terms for the layman they tested against past data found the algorithm for that model combined it with a statistical model for future stock returns called a monte carlo simulation..which itself can be based on a wide variety of assumptions  and many flaws which I I won't go into here there are many sources on the web on the pitfalls of such simulations.

In developing their model they used a frequent (and flawed) technique: data snooping: They kept running calculations based on past data until they found one that optimized their objective and made that the model..they used in sample data. As Robert Arnott of Reserach Affikiates notes in the article below no one ever presents a backtested strategy that doesn't work.

Furthermore as their whitepaper shows virtually all the benefits from tax harvesting over 2000-2014 came during two periods 2000- 2003 and 2008 2009 both periods when the us market dropped around 50%. And this was the period used by Wealthfront design its tax loss harvesting strategy. The likelihood of a market pattern just like that is next  10 years is next to zero and the utility of using a time series with a monte carlo simulation with two 50% drops in 10 years followed by one of the greatest market recoveries in history just as problematic.

Wealthfron "Tax Alpha" by Year

S+P 500 index

Since 2014 the market has been basically straight up meaning the opportunities for tax loss havesting (large losses) would have been quite minimal...so much for making a forward looking strategy based on past data.

Despite all the caveats raised in this article Wealthfront confidently presents its strategy is based on backtesting to produce the next generation of indexing. Any strategy based on backtesting is problematic be it a mix of ETFs or Wealthfronts direct indexing. But at least on the major asset classes upon which the broad category ETFs are based have a longer track record and there is a data series going back far before the ETF was created.

As Ben Carlson writes about backtesting in his great blog A Wealth of Common Sense:

How many bad backtests came before the good ones? I wonder how many millions of deceased backtests there right now are sitting in a recycle bin graveyard on computer desktops all across the globe? No one ever shows you a bad backtest because it’s much easier to date mine the past than the future.

Data availability at the time. The fact that we now have data that wasn’t available in the past changes the nature of that past data. There would have been ripple effects if investors knew then what we know now. Hindsight changes perception.

 What the frictions were. It’s almost impossible in a backtest to completely account for costs and frictions such as taxes, commissions, market impact from trading, market liquidity, etc. Sure, you can estimate these frictions, but you never truly understand how these things will affect your bottom line until you actually have to execute buy and sell orders.

What’s going to happen in the future. As my colleague, Michael Batnick, put it recently, “Unfortunately there is no such thing as a front-test.” Every market environment is different than the last so you have to be able to accept that the future will never look exactly like your time-tested strategy.

Finally: How Do You Get Money Out of Your Direct Indexing Portfolio ?

Wealthfront direct indexing portfolios have more in common with the seperately managed accounts (SMAs) offered by many brokerage firms..albeit at far lower cost. Those portfolios seldom hold more than a few dozen stocks (as opposed to the hundreds held by wealthfront)  and the amount of assets allocated to the strategy are far less than the assets under management.

Investors in individual stock portfolios can choose the method of sales they want to use to maximize tax savings their brokerage firm will have a record of their "tax lots " created by reinvesting dividends and buys and sells. The investor can give the brokerage firm standing instructions  average cost , last in first out, or first in first out  or highest or lowest cost or tax minimization at many brokerage firms.  The proper choice of designating tax lots to minimize taxes is complicated.However nowhere on their website is there an explanation of how Wealthfront accounts for tax lots when it sells shares when investors liquidate all or some of their portfolios.

Wealthfront may look simple and cheap to folk like those at nerd wallet and many of its customers. But a look under the hood shows a very complicated strategy based on questionable assumptions. It will likely be difficult for real world execution of their strategies will produce the resuts presented in their whitepaper.