Gold ETF Fee War Gets Complicated

[This article appears in our May issue of ETF Report.]

In the ETF fee wars, the tickers may differ, but the story stays the same: Investors flock toward whichever fund on the market is cheapest, even if that ETF undercuts by a basis point or less—except, it seems, in the gold market.

Year-to-date, physical gold ETFs have brought in more than $1 billion in new net investment assets. But these flows have bucked precedent, in that the cheapest fund on the market isn’t the one raking in cash.

That’s because there’s more to the story in gold than just a headline sticker price. Nuance matters, and more expensive funds actually possess a subtle structural detail that, for buy-and-hold investors, can make all the difference.

When Cheap Isn’t Enough
With an expense ratio of 0.17%, the $904 million Aberdeen Standard Physical Swiss Gold Shares ETF (SGOL) is technically the cheapest physical gold ETF on the market. We say “technically,” because the $467 million GraniteShares Gold Trust (BAR) has an expense ratio of 0.1749%, which rounds down in most data services—including FactSet, which powers ETF.com’s fund reports—as 0.17%.

Last December, SGOL, the third-largest physical gold ETF, slashed its fees from 0.39% to 0.17%, an aggressive move that put the fund just a hair cheaper than BAR, which, up to that point, had been the cheapest gold ETF on the market.

Rather than usher in a wave of new investor cash, however, SGOL has limped along since its fee cut, bringing in just $45 million in new net assets since Dec. 1, 2018. Meanwhile, BAR has brought in $144 million over the same period.

2018: A Year Of Price Drops
SGOL’s fee drop was only the latest salvo in a months-long price war among gold ETFs. When BAR launched in August 2017, its expense ratio of 0.20%—half that of the segment’s existing leader, the $33 billion SPDR Gold Trust (GLD)—immediately made BAR the cheapest-in-class.

However, BAR was itself undercut last summer by two new ETFs: the $654 million SPDR Gold MiniShares Trust (GLDM) and the $109 million Perth Mint Physical Gold ETF (AAAU). Both funds launched with an expense ratio of 0.18%.

A few months later, BAR cut its expenses from 0.20% to 0.1749%, making it once again the cheapest gold ETF until SGOL’s move last December.

All this horse-jockeying makes it easy to forget that, once upon a time, the second-largest fund in the space, the $12.7 billion iShares Gold Trust (IAU), also launched as a low-cost competitor. However, with its expense ratio of 0.25%, IAU now seems downright pricey compared with the current generation of gold ETFs.

GLDM: Breakout Hit
That said, flows into BAR are being swamped by IAU and GLDM, two more expensive competitors (see Figure 1).

Year-to-date, IAU has brought in $849 million in new net inflows, while GLDM has brought in $248 million.

The numbers look even better over a one-year period: IAU has seen its assets rise by $1.3 billion, while GLDM has risen by $603 million. What’s more, in the nine months since its launch, GLDM has never seen a day of outflows.

Compare that with AAAU, which costs the same as GLDM, and which launched around the same time; it’s only seen about $17 million in new net inflows year-to-date.

Why would investors gravitate specifically to IAU and GLDM? What is it about these two funds that make them so attractive to investors looking to allocate to gold?

Gold-Per-Share Ratio Matters
The answer boils down to a nuance that many investors less familiar with gold might overlook: gold-per-share ratio.

GLDM is essentially a modified version of GLD: Its vaulting and custody arrangements remain the same, but whereas each share of GLD represents 1/10th of an ounce of gold metal, each share of GLDM represents 1/100th of an ounce (the same as IAU).

That smaller gold-per-share ratio especially appeals to retail investors, who generally seek to place smaller trades to fulfill a buy-and-hold asset allocation strategy. A smaller gold-per-share ratio brings down the ETF’s share price, making it possible for investors with smaller asset bases to allocate meaningfully to gold.

As a result, smaller gold-per-share ETFs often see wider distribution in robo platforms or in ETF portfolio models by strategists, who don’t necessarily want to switch between products depending on their clients’ existing asset base.

“Advisors need to cater to millennial clients who may only have $2,000 to invest, as well as to their parents, who might have $2 million,” said Matt Bartolini, head of SPDR Americas Research. “They just want to have one product, one model open to everybody.”

Larger Share Price Benefits Traders
While a smaller gold-per-share ratio benefits buy-and-hold investors, it does little for short-term or tactical traders, who prize liquidity over low expense ratios.

“The more active a trader you are, the larger a [gold-per-share ratio] you want to see,” said Greg Collett, director of investment products for the World Gold Council. “That’s because the more shares you have to buy for the same amount of money, the higher your trading costs become.”

Most traders likely won’t be swapping en masse from GLD, which trades with pennywide spreads and minimal premiums and discounts.

But BAR—which has the same gold-per-share ratio as GLD, but at half the cost—has been able to lure at least a few traders, given its inflows of $445 million over the past 12 months.

“A lot of our clients are entrepreneurs and business owners. They run their own practice,” said Will Rhind, CEO and founder of GraniteShares, and formerly of the World Gold Council and ETF Securities. “For them, price is important; but the insight from our team’s experience and the relationship they can have with us matters too.”

Liquidity = Versatility
For the most frequent traders, BAR is additionally attractive, because it offers a lower spread (0.06%) than GLDM, which has a spread of 0.08%.

BAR also possesses a much lower creation unit size of 10,000 shares versus GLDM’s 100,000 shares, meaning it’s easier to create bulk trades in discrete sizes.

That may also be part of the reason BAR has amassed assets at SGOL’s expense, as the two ETFs have the same gold-per-share ratio as GLD, but new shares of SGOL can only be created in blocks of 50,000 shares or more.

At this point, we’ll probably never see GLD dethroned from its position as the king trading vehicle for gold investors. But so long as low-cost competitors can devise a better mousetrap, we’ll continue to see them chip away at GLD’s dominance. Over the past 12 months, for example, GLD saw outflows of $2.7 billion. Meanwhile, all other gold ETFs combined saw net inflows of $2.3 billion.