Among the many too-good-to-be-true financial stories is the Alpha Architect 1-3 Month Box ETF, known by its ticker BOXX, that offers Treasury bill yields taxed at capital gains rates. For some investors that could mean paying a 20% tax on returns rather than 37% — a “saving” that’s helped draw some $1 billion in assets to the fund.
There have also been a series of warnings. Steven Rosenthal, a senior fellow in the Urban-Brookings Tax Policy Center, thinks the tax benefits will not stand up. Elm Partners Management’s Victor Haghani and James White point out that the tax benefits will be smaller for most investors, and can even be negative. Moreover, they say that even with the full tax benefit the risk-adjusted return of BOXX is inferior to holding Treasury bills or money market funds.
I have a different objection: it doesn’t make sense to want an investment that beats the long-term total return of Treasury bills on an after-tax basis, because Treasury bills are a terrible long-term investment in the first place. The only reasons to hold liquid, short-term, low-risk investments like Treasury bills and money market funds are for cash you might need suddenly, or sitting out risk temporarily.
You pay a tremendous cost in yield for liquidity and safety, greater than the difference between 37% and 20% tax rates.
Consider the strategy of buying five-year Treasury notes versus rolling 13-week Treasury bills. Going back to Sept. 1, 1981, the five-year note has almost always delivered a better pre-tax total return. At a yield of 4.06% (the level early Monday), a five-year note will deliver a minimum 20.3% total pre-tax return over five years, compared with a historical total return of between 2% to 18% for rolling 13-week Treasuries.

I admit this is not entirely fair. I didn’t cherry-pick my data, but the Federal Reserve numbers only go back to 1981. In some years in the 1970s when interest rates were going up rapidly, rolling 13-week bills beat the five-year. Also, if you bought the note at the pandemic low yield of 0.19% on Aug. 4, 2020, you would have lost to Treasury bills by Aug. 4, 2025 — something that’s not in the chart above.
Still, it’s fair to say that an investment in rolling 13-week bills almost always loses pre-tax compared to buying a five-year note, and it often loses by quite a bit.
What if you pay 37% income tax on the coupons from the five-year note as they are received but only 20% at the end for rolling Treasury bills (as the BOXX ETF tries to emulate)?
As shown in the chart, that makes things a little more even, but the five-year note still looks like a considerably better bet. You’ll earn around 15% if you hold the five-year to maturity after 37% tax on the coupons, which historically has been associated with a 2% to 10% return from rolling 13-week Treasury bills and paying taxes at the end.
Of course, all this is for an investor who never touches the money, in which case she should probably be in riskier, higher-average-return assets than either Treasury bills or Treasury notes. I assume any money in the BOXX ETF is either temporarily sitting out risky investments or is a cash reserve for potential emergencies.
If the BOXX ETF is sold in less than one year, the tax advantage disappears, so it’s not suitable for temporary positions, nor for money likely to be used within a year. The only remaining investment case is a cash reserve that is unlikely to be used. Only a fairly wealthy person can afford this — it’s a luxury for peace of mind, not an investment that can be justified quantitatively.
But even for this scenario, I don’t think BOXX makes sense. If the money is needed within a year, or is never needed within five years, BOXX loses to alternatives. If the money is needed in more than one but fewer than five years, I think it’s pretty much a coin flip whether BOXX outperforms holding a five-year Treasury.
The Treasury note can be bought with zero commission or bid/ask spread from the Treasury Department’s Treasury Direct platform and sold instantly with very little transaction cost. The downside is it can have appreciable price volatility, especially in the earlier years. So you need to hold more of it to be confident of having enough when it is needed.
BOXX offers more certainty of value but comes with a lower expected return — even after tax — and you pay commissions and bid/ask spreads on both the purchase and sale. Its price volatility is lower than that of a five-year note, but considerably more than a Treasury bill due to its complex positions and changing market appetite. It can move as much as 0.25% in value in a day without moves in Treasury bill prices. It would likely revert to value in a few days, but if you had to sell on a bad day, that’s no comfort.
Even if BOXX were completely riskless, and an investor were entirely certain of a tax advantage of 20% versus 37%, I can’t see a place for BOXX in a sensible portfolio with a rational liquidity allocation. It strikes me as a product designed to avoid taxes as if that were an end in itself without considering whether an owner is better off after the savings.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out our full schedule of upcoming CE-approved virtual events.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
More ETF Topics >