Monday, January 29, 2007
Working On A Theory
I have written many times that if you use investment products, like ETFs, you need to stay on top of new products that come out in case they are better than existing products. I have started to wonder if the DB Gold ETF (DGL) and the DB Silver ETF (DBS) might be better to use than streetTRACKS Gold (GLD) or iShares Silver (SLV). GLD is a client holding.
DGL and DBS own mostly treasury bills and enough futures contracts to create the exposure. The big plus is that the interest on the T-bills pays the fee and could pay out to shareholders after the fee is covered. The big negative is that contango could work against the performance of the fund. DB uses something call Optimum Yield which allows to cherry pick the best contract to roll to. In theory contango on any roll forward could be a money loser.
With GLD and SLV holding the actual metal contango is not an issue but both funds have a fee that has to be covered. They sell a little of the metal to cover the fee and in a few years the difference could matter, some would say it matters now.
It seems to me that if Optimum Yield works, it is too early to know now, then the total return from DBS and DGL would be higher than GLD and SLV. DBS and DGL would provide the returns of the metal plus treasury interest (no guarantee it will pay of course) whereas GLD and SLV provide the returns of the metal minus the fee.
I am not saying I am the first one to think of this, but I can't recall reading about this anywhere. I don't think it can be known for a while whether this can stand up but I wonder.
Further, although far from my skill set, I wonder if this, if it even exists, could create some sort of unintended consequence due to potential arbitrage as is going on with the MacroShares.
For its part, DB says that interest from treasuries will pay out once a year in a special dividend, that is if there is anything to pay.
DGL and DBS own mostly treasury bills and enough futures contracts to create the exposure. The big plus is that the interest on the T-bills pays the fee and could pay out to shareholders after the fee is covered. The big negative is that contango could work against the performance of the fund. DB uses something call Optimum Yield which allows to cherry pick the best contract to roll to. In theory contango on any roll forward could be a money loser.
With GLD and SLV holding the actual metal contango is not an issue but both funds have a fee that has to be covered. They sell a little of the metal to cover the fee and in a few years the difference could matter, some would say it matters now.
It seems to me that if Optimum Yield works, it is too early to know now, then the total return from DBS and DGL would be higher than GLD and SLV. DBS and DGL would provide the returns of the metal plus treasury interest (no guarantee it will pay of course) whereas GLD and SLV provide the returns of the metal minus the fee.
I am not saying I am the first one to think of this, but I can't recall reading about this anywhere. I don't think it can be known for a while whether this can stand up but I wonder.
Further, although far from my skill set, I wonder if this, if it even exists, could create some sort of unintended consequence due to potential arbitrage as is going on with the MacroShares.
For its part, DB says that interest from treasuries will pay out once a year in a special dividend, that is if there is anything to pay.
Labels:
commodity,
ETF,
investment products
Subscribe to:
Post Comments (Atom)





8 comments:
One other thing you may want to consider in your analysis are the various tax implications of ownership (for those holding it in taxable accounts). One little known fact about GLD is that holding it is considered holding a collectible and as such it is subject to the collectibles tax rate of 28%. For those holding it long term and expecting to get the LT cap gains rate they may be in for a surprise.
I have not looked into the DGL fund to see what its tax implications are but I would assume that holding treasuries and futures contracts simply subjects one to normal ordinary income rates on the interest and ST/LT cap gains where appropriate.
From pg 23 of the Streetracks GLD prospectus (http://www.streettracksgoldshares.com/pdf/streetTRACKS.pdf)
Under current law, gains recognized by individuals from the sale of ''collectibles,'' including gold bullion, held for more than one year are taxed at a maximum rate of 28%, rather than the 15% rate applicable to most other long-term capital gains. For these purposes, gain recognized by an individual upon the sale of an interest in a trust that holds collectibles is treated as gain recognized on the sale of collectibles, to the extent that the gain is attributable to unrealized appreciation in value of the collectibles held by the trust. Therefore, any gain recognized by an individual US Shareholder attributable to a sale of Shares held for more than one year, or attributable to the Trust's sale of any gold bullion which the Shareholder is treated (through its ownership of Shares) as having held for more than one year, generally will be taxed at a maximum rate of 28%. The tax rates for capital gains recognized upon the sale of assets held by an individual US Shareholder for one year or less or by a taxpayer other than an individual US taxpayer are generally the same as those at which ordinary income is taxed.
While it is always helpful to save on expenses, what are the odds that the difference would amount to as much as 30 basis points? If your basic goal is to establish a position in gold, my guess is that using one vehicle vs. the other will make relatively little difference.
Clearly,there are search costs. Even with your knowledge and experience, you can't readily evaluate the difference between products.
My guess is that the time may be better spent looking for undervalued companiesw. Which brings up another question I have had for some time. You never seem to discuss inestments in individual companies, or even styles - e.g., small-cap value. Is that because you don't want to talk about them,or because you only want to make macro bets?
anon@2:40,
DBS adn DGL are invested almost entirely in t-bils yielding 5%. The yield could possible work out to 4% for the year. If, big if, it works out that way the difference could be substantial.
The price of the future should include the carrying costs of gold, include storage and cost of capital. That 5% t-bill interest isn't a free lunch.
If you're of an apocalyptic turn of mind, the futures contracts are just more "fiat"paper, while bullion reigns supreme. I'm not of an apocalyptic turn of mind, and if DB feels they can always pick the right contract, more power to them.
i don't think it is so much "pick the right contract" as it is ok we have to roll today which of the next 13 months is the best one to roll to.
Roger, perhaps you might want to check out this article from the Financial Times called “Speculators profit from commodity investors” which discusses “date rape”. I’d like to know you’re thoughts on this. Yet another cost to consider with products that hold derivatives instead of the underlining product.
Article found here:
http://www.ft.com/cms/s/45804864-aa56-11db-83b0-0000779e2340,dwp_uuid=d8e9ac2a-30dc-11da-ac1b-00000e2511c8.html
RE the FT article. I tend to think that all products have pluses and minuses. One outweighs the other and this is a subjective call. The bottom line to any product (commodity or otherwise) is do you think it does a good enough job capturing the effect warts and all.
Post a Comment