Bron Suchecki's  Instablog

More »

On Sep 17 Jeff Nielson posted an article on SLV. I took issue with his belief that ETFs' management fees were unrealistically cheap and thus another indicator they were a scam. Below is the exchange between Jeff and I on the matter.

Bron: You say "custodians of the vast majority of all the world's bullion-ETFs – a service which they are providing free of charge" but SLV has an expense ratio of 0.50%, some of which if I remember the prospectus correctly, is paid to the custodian. If SLV holders pay 0.50% how can it be considered "free". By what do you mean free?

Jeff: Hi Bron. Just look at all that is SUPPOSEDLY covered by this 1/2% fee:

1) Transaction costs. Purchases must be made CONSTANTLY, all day long - in order to buy the actual silver for unit-holders at the same price they bought their units at. Given the huge volatility with silver, it's not even feasible to restrict buying to once a day - since silver has had MANY daily moves of 5% or more.
2) Insurance/delivery costs
3) Storage/security costs.

Obviously BILLIONS of dollars of silver require significant security to guard such a hoard. The U.S. government has an entire military battalion guarding Fort Knox - so no one can find out how much gold is NOT there.If you think these costs are minimal, then answer this question: why do the small number of companies who hold their own bullion need to charge MANY times that premium for their own security/storage costs?

Bron: Before I comment, just want to state upfront that I work for the Perth Mint, but I am speaking here in a personal capacity. While I’m speaking personally, obviously the ETFs are competitors to my employer’s business, both in respect of physical coins and bars as well as our own storage facility, so I’m not any apologist for the ETFs. Taking each of your points in turn.

1) Transaction costs. I note that SLV’s average Bid Ask Ratio is 0.08%. This is very tight but is not necessarily unprofitable for a market maker. You are right that the market maker must be purchasing (or selling) gold constantly as it sells (or buys) SLV shares. My experience with the Perth Mint’s ASX listed product (code: ZAUWBA) is that the market maker will simply set their stock exchange price for an ETF higher than their cost on the wholesale over-the-counter market and adjust this constantly during the trading day. This way they always make a profit on transactions, it is not a cost to them. If individuals bid prices under this than the market maker misses out on a trade. It is only where there are excessive buyers or sellers that the market maker’s prices will get hit.

2) Insurance/delivery costs. Delivery costs are effectively zero, as the metal is most likely already in the vaults as sellers of physical need to bring their metal to London to trade it. Insurance is a real cost, but are easily covered by 0.50%. Important to note that the metal is not fully insured, just the first couple of billion (I don’t think the prospectus says anything about the first loss limit of the insurance). Once you get to a certain size therefore, the insurance cost is a fixed cost, not variable.

3) Storage/security costs. These are fixed costs, once you have a vault and have secured it, every additional ounce does not result in any change in costs. Once you get to the point that you have covered these fixed costs, every ounce above that is pure profit and this is where custodianship can be highly profitable. At 280 million ounces, SLV is definitely there in my opinion. Storage business is a classic case of economies of scale, which is why smaller companies have to have higher storage charges (eg Perth Mint allocated silver is 2.5% pa).

I have been a bit brief on explaining the above, but my view is that they are making money with a 0.5% expense ratio. That is why I think the “free of charge” line of attack is not supported and you are better off focusing on your other criticisms.

Jeff: Bron, at the time that SLV was created, there was only 200 million oz's of silver in GLOBAL inventories. Now SLV and others hold close to 450 million oz's. Obviously there MUST be both delivery AND insurance charges for AT LEAST 250 million oz's of silver - which could NOT have "already been in vaults".

As for security/storage costs, I'll happily concede (for purposes of argument) that no new storage space was created. This brings me back to my point about the ludicrous idea of a BANKER (holding a massive short position) SUBSIDIZING "longs" by providing free storage/security.

Even if you subscribe to that ludicrous fantasy, there is still the issue of the "opportunity cost" to banks. Precious metals are not the ONLY items in the world for which there is a demand for high-security storage. Will ANYONE suggest that banks will provide a FREE service for precious metals longs - rather than charge someone a fee for storing other valuable assets? Try asking JP Morgan to store YOUR OWN precious metals for free - and listen to how hard they laugh at you.

Bron: "Obviously there MUST be both delivery AND insurance charges for AT LEAST 250 million oz's of silver - which could NOT have already been in vaults"

You've missed my point. Lets assume the additional 250moz is real and was bought by bullion banks to back SLV & others. In that case, the bullion banks would incur no delivery charges as the seller delivers metal to London at their cost to be able to sell it on the spot market in London. Secondly, the additional 250moz has no insurance charges - as I said, they only insure the first $1b of holdings, not the entire holdings.

"the ludicrous idea of a BANKER (holding a massive short position) SUBSIDIZING longs by providing free storage/security" & "Will ANYONE suggest that banks will provide a FREE service for precious metals longs - rather than charge someone a fee for storing other valuable assets?"

Jeff, you keep on saying they are doing it for free when SLV charges 0.5%. Some of that 0.5% goes to the custodian, they are being paid. That is not "for free" - I don't understand why you keep on saying they are providing free storage.

The question is whether the 0.5% charge is realistic, profitable assuming the volumes of metal SLV and others hold is physical. As explained in my previous reply it is. Saying this does not mean that they have physical, but nor does it mean they do not.

Jeff: Bron, your assumptions about delivery cost are only valid if you're implying that silver (and gold) goes straight from refineries into bankster vaults - rather than having to be PURCHASED by the banksters (first) on the open market, and then transferred to their vaults.

When you mention the 0.5% fee charged by SLV, my understanding is that this also (supposedly) covers their OWN administrative costs AS WELL AS all the shipping costs, transaction costs, insurance costs, and storage/security costs.

You would be hard-pressed to find any ONE bankster service (in ANY of their business activities) which they are willing to provide for a 0.5% fee. Suggesting that they are willing to REDUCE their fees (to close to ZERO) to SUBSIDIZE the entry of longs into the market is simply nonsense.

Bron: "your assumptions about delivery cost are only valid if you're implying that silver (and gold) goes straight from refineries into bankster vaults - rather than having to be PURCHASED by the banksters (first) on the open market, and then transferred to their vaults."

No it doesn't. There is no difference between purchasing from refineries or on the open market - refineries are all in different countries just like existing stocks. If market makers cannot acquire metal from investors or sellers already holding it in London, they will actually be able to acquire it at a discount to London spot (which is the usual state of the market), the discount equalling the shipment cost into London. Even if they have to pay a premium (or pay shipment costs into London), then they just factor this into their bid and ask prices quoted for SLV. This is why delivery is not a cost that comes out of the 0.5% fee.

"When you mention the 0.5% fee charged by SLV, my understanding is that this also (supposedly) covers their OWN administrative costs AS WELL AS all the shipping costs, transaction costs, insurance costs, and storage/security costs."

The 0.5% does cover their administrative and compliance costs, but as I have discussed above and in my previous replies, any shipping and transaction costs are recovered via market making activities, so these do not come out of the 0.5%. As I have also replied, insurance and storage/security are FIXED costs, not variable, whereas the revenue of 0.5% is variable. This means that once you cover you fixed costs, the 0.5% on any additional metal is pure profit.

"You would be hard-pressed to find any ONE bankster service (in ANY of their business activities) which they are willing to provide for a 0.5% fee. Suggesting that they are willing to REDUCE their fees (to close to ZERO) to SUBSIDIZE the entry of longs into the market is simply nonsense."

0.5% is not "close to zero". On 280moz, 0.5% = $24 million, that is not anywhere near zero. The fact is that in the wholesale market storage is offered for much less than 0.5%. Do you remember David Einhorn's Greenlight Capital exiting his GLD in favor of physical bullion? He did this because it was CHEAPER, in other words he could get storage for less than GLD’s 0.4%. In fact, quoting www.hardassetsinvestor/:

“By contrast, a $400 million player in the bullion market has substantial room to negotiate. You can be sure his [Einhorn] bullion holdings are being custodied for less than 12 basis points.”

If you believe that 0.5% is an unrealistic fee, a subsidised fee and therefore proof that SLV is a scam, then logically you must also believe that Bullion Vault, with a 0.12% storage fee, is also a scam. This puts you in a bit of a spot, because Bullion Vault is one of the most transparent operations in the market, and favoured by many goldbugs and commentators. Your stepping out on a limb here.

The post above was on Sep 21, Jeff replied to another post on Sep 22 but ignored mine. I posted the comment below on Sep 27. No response by Jeff as at Oct 4.

Bron: You have replied to someone else's comment which appear after mine, but ignored mine. Does this mean you conceed on the issue of the reasonableness of the storage fee?

The article Scotiabank and the Real Silver prompted me to have a closer look at their 2008 Annual Report. Two interesting quotes (note all dollars are Canadian):

“In Scotia Capital, revenue declined by 25%, due mainly to charges relating to the Lehman Brothers bankruptcy, valuation adjustments and generally weak capital markets. These were partially offset by record foreign exchange and precious metals trading revenues, and strong growth in corporate lending.” (p28)
and
“Precious metals trading revenue was a record $160 million, an increase of $44 million or 38% over last year, with higher revenues recorded in each of our major centres.” (p30).

Not surprising to see strong precious metal results from Oct 2007 to Oct 2008 (Scotiabank's reporting year). What I did find interesting is the observation by ispeakofpeak that Scotia's gold and silver certificates declined from $5,986m to $5,619m (p122), a 6.1% drop. This drop would reflect both changes in precious metal prices as well as changes in ounces held.

Unforunately, Scotia do not provide a breakdown of how many gold or silver ounces made up the certificate dollar total. But we do know that Canadian dollar gold prices were up 18% from 1 Nov 07 to 31 Oct 08 and that silver was down 14%.

If you think about it, assuming all the certificates were gold, then if the price was up 18% but Scotia's value dropped 6.1%, they must have lost a lot of ounces. On the other hand, if it was all silver, then as the silver price dropped 14% yet Scotia's value dropped only 6.1%, then they must have had an increase in ounces of silver.

Either of these would not be correct - there must be a mix of gold and silver. For sake of example and to put some numbers to it, lets assume for every $1000, $500 was gold and the other $500 silver. This is not unreasonable, I have seen many clients make this sort of "portfolio allocation" when buying precious metals. A 50:50 split by value works out as:

Gold Oz 2007: 3,996,336
Gold Oz 2008: 3,179,160
Change: -817,176

Silver Oz 2007: 220,173,903
Silver Oz 2008: 240,380,913
Change: 20,207,010

First off, some pretty impressive ounce totals, that would put them up there in my gold and silver league tables, if they were prepared to publish their actual ounce numbers.

What I do find interesting is that they lost gold at a time when everyone else (ETFs, GoldMoney, etc) were gaining. And it does not matter what you assume the split at. If you chose 75:25 gold:silver, or 25:75 it may change the amounts of gold and silver, but it still results in a loss of gold and a gain of silver.

Another interesting observation is that on their balance sheet they list Precious Metals at only $2,426m ($4,046m for 2007, p106). So dollar value precious metal liabilities only down $365m, but precious metal assets down $1,620m. This means that in 2007 they had 68% of their liabilities covered by physical but in 2008 only 43% cover.

If we look to their derivatives, p150 shows that “Foreign exchange and gold contracts, futures” with 1 year or less maturity were $2,602m out of a total of $4,239m. Gap between 2008 precious metal liabilities and physical assets was $3,193m. Conclusion: remaining 57% covered by COMEX futures and/or over-the-counter forwards.

Yesterday I was dismissive of the recall of Hong Kong's gold as significant, but it is another bit of evidence of a shift in central bank attitudes towards gold. Far more significant indicators include (see this MineWeb article):