Headlines last week were about GLD's largest holder making the switch to the hard stuff. Why the move?
- Bullion vs. GLD by the numbers
- It's all about assumptions
- Einhorn's math
GLD's largest shareholder, until recently, was noted hedge fund manager David Einhorn's Greenlight Capital hedge fund. Last week, we learned that he'd sold all of his GLD (SPDR Gold Shares (NSYE Arca)) GLD holdings in favor of physical bullion. How much? About 4.2 million shares of GLD, or roughly $390 million.
So does Einhorn know something we don't know? Is the 40 basis points charged by GLD for the privilege of owning their shares a rip-off? Or a great deal?
Let's put some of the moving parts on the table.
The Challenge Of Owning Physical Gold
If you are going to buy physical gold, you first have to decide where you're going to store it. Sticking a few hundred thousand dollars of gold in a safe deposit vault is a bad idea, for one big reason: If you're buying large quantities of gold, the bars you buy will be serialized. They have provenance, and their acceptance as a thing worth a few hundred thousand dollars is based on an unbroken chain of ownership. Most bars never leave the big gold vaults; ownership simply moves from one buyer to another.
How much does this custody cost? Well, that varies, but on the cheaper end of the scale, you can go to someone like BullionVault.com, an online gold exchange that custodies just over 18 tonnes at the moment. BullionVault charges 12 basis points per year.
Once you decide where custody of your gold will be, you actually have to buy it, which means dealing with spreads and commissions in the variable precious metals market.
Spreads in particular can vary massively in the precious metals market. If you're a primary dealer on the OTC market, trading in 10,000 ounce blocks, your spread could be as low as 50 cents an ounce. For retail investors looking at pooled gold (where you are buying an allocated portion of a larger bar), a quick survey of available quotes suggests a spread of about 25 basis points is the norm.
Commissions are easy to figure out. If you're buying gold from a dealer, like Kitco.com, you won't pay a commission - Kitco's profit is built into the spread they charge. In the case of something like a pooled account, the spread varies, but a spot check last week saw gold offered at 936.20 on the bid and 941.20 on the ask, putting the spread at $5/ounce, or 50 basis points.
If instead you choose to do the buying and selling yourself on an exchange like BullionVault, your commission works out to be about 10 basis points for a million dollars of trading (more for smaller trades, less for larger).
So your round-trip cost for physical gold is the width of the spread, plus commissions in and out, plus a custody fee.
In comparison, buying GLD is easy. You enter the ticker into your brokerage account and, whammo presto, you own gold. You pay an expense ratio of 0.40% per year and the good folks at SSgA and (more specifically) their custodian (JPMorgan) take care of things for you, storing numbered gold bars in a vault in London.
There are a few additional expenses, of course. You have to pay a commission to buy and sell shares, which can range from $5 for an ultra-discount broker to as high as 4-5 cents a share for a full-service firm. And you have to pay the available spread, which Morningstar says averages about 11 cents per share.
For the sake of argument, I'm going to assume that over time, premium or discount to NAV isn't an issue, and indeed, the evidence is that the premium/discount is pretty normally distributed around zero.
With these ideas in place, let's walk through a thought experiment that quite a few investors might be considering right now.
Let's suppose I have $1 million, and I want to buy gold. What's the cheapest way?
My costs will be:
GLD: 40 bps expense ratio, 2 cents per share commission, 11 bps spread
Bullion: 12 bps custody, 10 bps commission, 25 bps spread