Imagine walking up to a salad bar, putting together a nutritious, well-balanced plate, and being told to put everything back, because the restaurant only sells each item as an entree. Seems crazy, yet that is exactly the way traditional investing has worked.

Exchange-traded funds, or ETFs, give us access to cheap, globally diversified portfolios, but traditional brokerages only allow you to buy and sell ETFs in whole shares.

So for example, if you want exposure to U.S. stocks via Vanguard’s VTI ETF, you must do so in multiples of $107 (the most recent price VTI traded at). Any extra cash will sit on the sidelines. If you are looking to put together a portfolio of a handful of ETFs, the problem compounds itself.

If this seems inconvenient and antiquated, that’s because it is. A smarter, modern investment service should abstract you from constraints such as the prices at which shares are trading, and it should let you invest precisely to your allocation, down to the penny.

Betterment was built from the ground up as a next-generation investing platform in order to democratize sophisticated portfolio management that has traditionally been available only to higher-balance investors. A key part of our design ensured that investors could buy and sell fractions of ETFs, down to six decimal places.

As a result, Betterment is more efficient at portfolio management for customers at every balance size—ensuring that every customer can fully benefit from features like tax loss harvesting , even if they are in the early accumulation phase.

Automated investing services that only use whole shares will be less efficient than Betterment at investing your money.

Here are several practical issues for investors’ portfolios when they are only invested in whole shares:

  • Your portfolio always carries a cash remainder because funds don’t line up perfectly—that leads to cash drag, which can hurt returns.
  • It is difficult to precisely achieve your target allocation, causing you to take too much or too little risk.
  • Automated tax loss harvesting is less effective, as there are fewer opportunities to buy and sell securities to harvest a loss.

Problem # 1: Cash Drag

Imagine an investment service called WholeShare, which cannot trade in fractional shares. A customer deposits $5,000 into an newly opened taxable account. For illustration, let’s assume a Betterment target allocation at 90% stocks. Using market prices from April 2, 2015, WholeShare might make the following purchases in an attempt to get to the target allocation:

WholeShare 90% Stock, Matching Target Allocation

Asset Target Value Share Price Whole Shares Purchased Actual Value
Cash $0 $331
MUB $274 $110.18 2 $220
LQD $28 $121.73 0 $0
BNDX $119 $54.02 2 $108
VWOB $79 $77.99 1 $78
VTI $808 $107.36 7 $751
VTV $808 $83.39 9 $751
VOE $259 $91.90 2 $184
VBR $226 $109.73 2 $219
VEA $1,874 $40.33 46 $1,855
VWO $525 $41.94 12 $503
Total $5,000

WholeShare would need to buy $808.14 of VTI to hit the target allocation. However, VTI is trading at $107.36, which means it can only buy seven whole shares, worth $751, leaving $57 sitting on the side in cash.

Across the entire $5,000 deposit, for each asset, this limitation leaves us with with $331 in cash, or 7% of the original deposit. That is 7% of the intended investment that’s not actually invested. The result is cash drag, which means a lower expected return than you could have achieved if your portfolio could hold fractional shares.

Problem # 2: Systematically Unbalanced Portfolio

WholeShare can mitigate the cash drag problem, but only at the expense of creating another one. For instance, it can use as much of the pooled cash as possible to over-purchase underweight assets. However, that would create substantial drift away from your desired allocation, because you would now be over-allocated to those portfolio components which were over-purchased.

Let’s assume one straightforward approach to use up as much cash as possible: start with the most expensive share that is underweight, buy one, and repeat until you can buy no more shares of anything.

WholeShare 90% Stock, Minimizing Cash Allocation

Asset Target Value Share Price Whole Shares Purchased Actual Value
Cash $0 $2
MUB $274 $110.19 3 $331
LQD $28 $121.69 1 $122
BNDX $119 $54.02 2 $108
VWOB $79 $78.03 1 $78
VTI $808 $107.44 7 $752
VTV $808 $83.47 9 $751
VOE $259 $92.00 3 $276
VBR $226 $109.76 2 $220
VEA $1,874 $40.35 46 $1,856
VWO $525 $41.97 12 $504
Total $5,000

In this case, we’ve virtually eliminated the cash drag. Yet here is the resulting drift away from the desired allocation:

Allocation Drift Due To Whole Shares

When a portfolio’s actual allocation diverges from its target allocation, the result is “allocation drift”.  Allocation drift is undesirable because it can lead to an investor being exposed to a portfolio which no longer reflects the desired risk level. It usually occurs because of normal changes in the prices of the securities over time, and is addressed by rebalancing the portfolio.

In this case, the allocation is substantially off on day one, before any market changes, and rebalancing is of no help, because the size of the portfolio building blocks prevents a more precise allocation.

Problem # 3: Less effective tax loss harvesting

The problems above also come into play when combined with tax loss harvesting.

In our white paper, we explain in detail why regular deposits make tax loss harvesting more effective. The more frequently you purchase a volatile security, the more price points you have that can potentially present a harvesting opportunity. This is another case where the ability to purchase fractional shares offers an advantage.

To illustrate the concept, let’s look at a very simplified hypothetical.

Say that our $5,000 portfolio is generating $10 of dividends a month. Assume we are underweight VTI, so this is the asset we want to buy with any incoming cash flows. For illustrative purposes, we’ll assume that VTI, which can be quite volatile, is swinging back and forth in value each month. Let’s look at how WholeShare might make the purchase, as compared to a service that supports fractional share trading.

WholeShare Purchases Using Dividend Cash

Month Hypothetical price of VTI Shares of VTI bought Price paid (cost basis) Cash available
January $100 $10
February $95 $20
March $105 $30
April $95 $40
May $105 $50
June $95 $60
July $105 $70
August $95 $80
September $105 $90
October $100 1 $100 $0
Total 1 $100

Fractional Purchases Using Dividend Cash

Month Hypothetical price of VTI Shares of VTI bought Price paid (cost basis) Cash available
January $100 0.100 $10 $0
February $95 0.105 $10 $0
March $105 0.095 $10 $0
April $95 0.105 $10 $0
May $105 0.095 $10 $0
June $95 0.105 $10 $0
July $105 0.095 $10 $0
August $95 0.105 $10 $0
September $105 0.095 $10 $0
October $100 0.100 $10 $0
Total 1 $100

In both cases, we wind up with one share, with an aggregate basis of $100. However, with fractional shares, each $10 dividend was immediately invested, offering a variety of price points. Say we now check for harvesting opportunities, and the market price is still $100. With WholeShare, there are no built-in losses available—you are holding a single tax lot, with a basis equal to the current price.

But with a fractional portfolio, four of the lots you are holding (those bought in March, May, July and September, totaling .38 shares), are at a loss, with an aggregate basis of $40. A tax loss harvesting service can sell only these lots at the current price ($38 for .38 shares) and realize a $2 loss, which is 2% of the value of the position. In actuality, daily automated tax loss harvesting, such as Betterment’s Tax Loss Harvesting+, would have been operating throughout the year, potentially realizing bigger losses on each lot, each time the price dipped to $95.

These small benefits will add up over the years. Fractional shares squeeze every efficiency out of tax loss harvesting, utilizing every cash flow, no matter how small, keeping you in balance, always keeping you out of cash, and creating multiple tax lots at every opportunity across all positions in the portfolio.

Tax loss harvesting is not appropriate for everyone, which we discuss in detail in our white paper. In particular, those who are in a low enough income tax bracket to realize capital gains at a 0% rate should not use it. However, a smaller portfolio does not necessarily imply a lower tax bracket. Investors who are in the early accumulation phase are often earning enough to benefit from Tax Loss Harvesting+.

Fractional share technology is the foundation of a service that truly enables advanced portfolio management strategies at every balance size.

More from Betterment: