What happens if gold explodes to the upside

[rickb]

Consider three little PP investors.  One builds his house out of paper and holds his gold in GLD (or IAU or SGOL).  One builds his house out of sticks and holds his gold in GTU or PHYS.  And one builds his house out of gold bricks (holds his gold in his grubby little hands).  Let's say the big bad wolf comes to town and the price of gold goes through the roof.

What happens to our three little PP investors?

Of course, the one in the brick house holding gold in his grubby little hands is fine.  He can sell his gold at the spot price, and possibly considerably more depending on how hard physical gold is to obtain.  He might want to be a wee bit careful on the way to the gold dealer (or on the way home).

What happens to the one in the stick house, holding GTU or PHYS?  These funds hold a known amount of gold and typically buy more creating additional shares when their premium over NAV gets high (to better track the actual price of gold) or (at least theoretically) sell some and in the process destroy shares if their premium goes negative.  If the price of gold explodes, I'd think the premium would likely go up and if these funds can't find any gold to buy they can't really control the premium.  The net effect would seem to be that holders of these funds would be fine.  The premium might go kind of haywire for a while, but there seems to be little risk that it would go negative in a scenario where the price of gold is exploding.

Hmmm.  What happens to our PP investor in the paper house who holds GLD, or IAU, or SGOL?  These ETFs are designed to track the spot price nearly exactly by giving their authorized participants an open invitation to create shares at "spot NAV" in exchange for actual gold (adding gold to the ETF) or destroy shares by exchanging shares at "market NAV" for actual gold (taking gold from the ETF) whenever they'd like.  This means if the ETF trades at a higher price than the NAV of the underlying gold, the authorized participants can make money by creating new shares (buy gold at spot, create ETF shares at "spot NAV", and then sell these shares on the open market presumably reducing the prevailing NAV), and conversely can make money by destroying shares if the ETF trades below the NAV of the underlying gold (buy shares and exchange them for gold).  Fundamentally this is how all ETFs work - they track the NAV of the underlying assets because the authorized participants can make money whenever it drifts (in either direction).  However, who are these authorized participants and might they have other interests?

Let's see.  GLD's custodian is HSBC.  IAU's and SGOL's custodian is JP Morgan.  Perhaps not coincidentally, these are two names that regularly show up as holders of massive shorts on gold.  If the price of gold is exploding, the short positions held by these firms are turning into hugely negative amounts of money on paper.  And, if the longs on the other side of the trade start demanding delivery, these firms are obligated to deliver gold.  Would they simply default (stiffing the longs), or would they possibly as an authorized participant take gold from an ETF to satisfy short positions in a short squeeze, destroying shares of the ETF and thereby reducing the NAV (but not because it was too high to start with)?  In either case, what impact would would there be on the price of both gold and the ETF?

I think it's worth thinking about what might happen here and whether being in a paper house is OK or if you might rather be in a stick or even brick house.

[murphy_p_t]

in my warped mind, GLD is only suitable for short term trades, at best. Gold, among other things, is for insurance against financial instability and collapse.

If / when GLD blows apart when you need to collect on your insurance...you're SOL.

Another way to look at this...GLD will be grand...until its not (that's when you need it most).

[stone]

rickb, as you describe it, HSBC destroying GLD shares so as to obtain gold sounds exactly like what they do all the time and is the mechanism that keeps GLD tracking gold. Even in every day circumstances, some large customers buy GLD as a convienient way to obtain gold. They employ an authorized participant to convert the GLD into good delivery bars. My remaining anxiety about holding gold etfs is that a period of sustained massive volatility where the physical and GLD prices got out of whack could lead to an errosion of the GLD value as the arbitrage bled it in a kind of death by a thousand cuts manner. As far as I know so far GLD has tracked gold very well. Silver has had a couple of days where it has zigged while the etfs have zagged. I just hope that the larger more liquid gold market prevents that from being a danger. 
For me the physical gold etfs are the only open ended funds I kind of trust (not stock or bond funds). Maybe I'm just naive and I'm failing to grasp something important. I take your point that gold is a sort of catastrophe insurance so we should be extra vigilant about it but catastrophes can also hit personal physical gold (eg theft, fire, fake coins etc etc). 

[Larshus]

What will happen is simple. People will try to take their futures contracts and redeem them. The amount of gold that trades in futures dwarfs GLD every day. People fixate on GLD but that's not where the real players are for paper gold.

While conversion happens regularly (We do it for customers all the time), its happens as a very small portion of open contracts. Since futures contracts are a synthetic instrument there can be MANY more open than there are physical ounces at exchange approved warehouses. This is normally not a problem. However when a large run up and fear coincides with a futures expiration month, there can be a run on redemptions depleting the physical stock. This last became an issue in the beginning of 2007. Supply was tight but all the redemtions were made. Its worried ALOT of people though.

The problem is that if there is no physical stock left to redeem your futures are turned into cash, and not into gold. This is termed Force Majeure. It rarely happens, and when it does in other commodity complexes like the softs such as grain, its really not a big deal as you can take the cash and by more grain on the open market. The problem with gold is that when there is none left, the physical market will not let you buy any at the same price. Physical % over spot skyrocket and you just lost a large chunk of buying power. This scares the bajebus out of the wholesalers and many mistrust the futures gold redemption process because of this. Not only that, but the exchange only allows a select few warehouses to handle the redemption (old boys network) and they charge usurious rates for conversion and insurance/shipping... but that's another post entirely. (they do it to discourage conversion).

Anyways once the futures conversions fail, you will see a disconnect between "spot" price and what you can actually by gold for. Spot is calculated as the current forward month less the time value of money into that forward month. If the current forward month is not deliverable, then the spot price becomes a fictitious number and people start setting their own spot prices arbitrarily based on physical supply and demand. Obviously this is going to be higher than the old spot price

Hope this helps.

[AdamA]

While conversion happens regularly (We do it for customers all the time), its happens as a very small portion of open contracts.

Great post, Larshus.  What is your background?

[stone]

Larshus, I guess people on here are more concerned about "physical" etfs such as GLD than futures because (I hope I'm not not misrepresenting things) some of us hold "physical" etfs but fewer of us hold futures directly or etfs based on futures. Am I right in understanding that etfs such as GLD and PHAU (on the london exchange) link in to the price of actual gold bars because they hold actual gold bars and can be redeemed for actual gold bars by authorized participants and so are no more dependent on whatever the futures market does than is actual gold its-self? So if the price of actual gold bars diverges from that of futures, then the price of GLD will follow what the price of the gold bars does not the price of the futures? If GLD ever sells at a discount to real gold then people will buy GLD as a way to get hold of the real gold held in the GLD vault. If HSBC et al were to, for whatever reason, obstruct redemption of GLD for gold, then that could cause GLD to trade at a discount to physical gold but HSBC would only do that if they had some motive. They may have short positions for gold futures but I would only worry if HSBC etc were short GLD. I don't care whether they are short for gold futures. Am I missing something?

Like Adam I really welcome input from people like you who actually know about this stuff. I'm just a bewildered new customer.

[Larshus]

While conversion happens regularly (We do it for customers all the time), its happens as a very small portion of open contracts.

Great post, Larshus.  What is your background?

I work for a Precious Metals trading company that has been in business for 80 years. I've personally been trading physical and paper metals for 15+ years. I'd rather not say who as I do not want to appear biased in any way. I'd also prefer if the moderators, who have my work email, would keep it to themselves

[rickb]

The amount of gold that trades in futures dwarfs GLD every day. People fixate on GLD but that's not where the real players are for paper gold.

My understanding is JP Morgan and HSBC combined have a significant portion of the open commercial short contracts in their house accounts (not on behalf of their brokerage clients), making them possibly the most real of players for paper gold.  My point is that these two firms are ALSO the custodians of all the major gold ETFs (it's hard to find one for which one of these two firms is not the primary custodian) and how they might balance their obligations to the futures market against their roles as authorized participants in the ETFs in the face of a sharply rising gold price is far from obvious.  Yes, they might default on their short contracts and deliver money instead, but it seems like they could also choose to obtain gold from the ETFs to honor the short deliveries (by legally buying up shares they'd redeem - or perhaps by illegal means if they're desperate enough) - and what happens in these scenarios is very different. 

[Larshus]

Larshus, I guess people on here are more concerned about "physical" etfs such as GLD than futures because (I hope I'm not not misrepresenting things) some of us hold "physical" etfs but fewer of us hold futures directly or etfs based on futures. Am I right in understanding that etfs such as GLD and PHAU (on the london exchange) link in to the price of actual gold bars because they hold actual gold bars and can be redeemed for actual gold bars by authorized participants and so are no more dependent on whatever the futures market does than is actual gold its-self? So if the price of actual gold bars diverges from that of futures, then the price of GLD will follow what the price of the gold bars does not the price of the futures? If GLD ever sells at a discount to real gold then people will buy GLD as a way to get hold of the real gold held in the GLD vault. If HSBC et al were to, for whatever reason, obstruct redemption of GLD for gold, then that could cause GLD to trade at a discount to physical gold but HSBC would only do that if they had some motive. They may have short positions for gold futures but I would only worry if HSBC etc were short GLD. I don't care whether they are short for gold futures. Am I missing something?

Like Adam I really welcome input from people like you who actually know about this stuff. I'm just a bewildered new customer.

Getting to point of Redemption and Risk from GLD. A couple things you must realize.

1) Shares are created in a basket 100,000 at a time. They must be redeemed 100,000 at a time. Thats alot of gold. Very few private investors will ever be able to redeem based on size. This turns GLD, in effect, into a paper fund. There is also a $2,000 fee to enact the transaction plus shipping and possibly overnight storage fees.
2) Gold Bars in the Trust do not have to meet the LBMA standards. In fact, in reality I'm not actually sure it even has to be in bar form. The Trust can elect to deliver gold in this non good form upon redemption. Even though the fund states that gold upon creation must be at least 99.5% fine and meet LBMA the redemption does not have to meet it. It has up to this point, but that part is very odd.
3) The GLD Trust can, at any time it chooses, for whatever reason, elect to stop redemptions. There are a myriad of reasons this may happen.

So.. the main point is that GLD will trade with spot price of the forward futures contract in regular markets, but in stressed markets it will trade at some "other amount" defined by the following variables:
1) If GLD stops redemptions to stop a run on pooled gold, it will then create a GLD pool of fixed size that will then be "bid up" in the market place.
2) If GLD continues redeptions and depletes its pools because large institutions want the physical, then the pool will shrink even though gold price is rising. ETF shares will become hard to buy. Will they trade at a premium? Its undefined. Remember to redeem you essentially just present the ETF shares in blocks of 100,000 and they give you the gold.
3) The small investor will be caught unable to redeem their shares and will go along with the wild volatility ride. Will they be able to sell their shares on the open market? Most likely. for what price though?

As far as rickb's statements. In essence HSBC and JPM would just call Force Majeure and deliver cash if their futures pool account ran out of gold. Remember, they are under no force by the exchange "to find gold" to convert. They simply run a depository where if someone brings in a gold bar in deliverable form they can deposit it for credit, and vice versa. If there is no gold in the pool, they do not have to deliver. The fact they they also TRADE gold really has nothing to do with the futures pools or the ETF holdings. They are separate.

BTW they can't "default" on their short contracts. If the depository, which they may or may not be running, calls Force Majeure they simply have to deliver cash... not gold. They could buy up ETF shares on the open market and convert, and then deliver the physical, but why bother with the hassle and expense when they can just deliver cash?

Thats the real problem with this whole thing. When push comes to shove, you are more than likely getting cash, not gold... or you will be stuck in a position where you can't convert your ETF and you have paper shares  in an ETF that might become illiquid, or might go nuts in price.... which you will be afraid to sell and take fiat currency.

Both futures and GLD work very well in regular markets. They problem is one of the reasons you own gold is to insure against irregular markets. Both futures and GLD break down horribly in highly irregular markets. Hold physical costs more in time and sometimes in expense. However you take the irregular market risk out of the equation.

[hogtied]

And where does CEF stand in relation to the other etfs ?  Thanks

[stone]

I guess if the meme got about that GLD was "dodgy" and GLD consequently developed a slight discount then that might allow participants to obtain gold at a slight discount by buying up GLD and redeeming for gold. It would take collusion between the participants to hold off on the redemptions so as to maintain the discount and gradually milk it. To be honest, from my perspective, the whole gold market seems so manipulated (and perhaps needs to be??) that that "what if" possibility is neither here nor there. I only hold gold as a financial asset and then predominately hold it as a crude robust currency hedge rather than as some kind of "seed corn and ammo" style armegedon talisman. Gold itself has no use for me. It is only any use as a way of saving for nursing home accommodation when we're in our dotage and such like. I can envisage gold being a very useful financial asset as the world readjusts over the coming decades. BUT a "total collapse" in which reputable coin dealers nevertheless will securely exchange physical gold for safe passage/food seems remote. People normally just bury physical gold and never return for it whenever a "total collapse" occurs. The Romans just buried gold and fled in 400AD. Same with the Vikings in 800AD or whenever.

[Larshus]

I'm sorry if I came off all doom and gloom. The intelligent thing to do is to hold gold in GLD or futures for the normal account management reasons... re balancing et. al. You might want to hold a small amount of physical "just  in case". You may not. But you should be intimately aware of the underlying GLD and futures vehicles so if it ever DOES get to the Roman collapse as you brought up, or to put it better in perspective, into the Eastern Europe collapse that happened during World War 2 when Germany invaded, you'll know exactly what to do with your paper instruments.

[stone]

Larshus, its good to get warnings such as yours.

[AdamA]

I'm sorry if I came off all doom and gloom.

Not doom-and-gloom at all. 

You have some great insight. 

What do you think is going with GTU right now (up 4% yesterday, down 5% today)?

[rickb]

And where does CEF stand in relation to the other etfs ?  Thanks

CEF (a gold/silver mix), and its gold-only twin fund GTU, are closed end funds rather than ETFs.  Like the ETFs, when you buy/sell shares of these you're buying/selling them on the open market from/to someone else - and doing this does not cause the fund itself to buy or sell anything.  Also like the ETFs, ordinary shareholders of CEF/GTU have no chance whatsoever to redeem their shares for metal - so they're both "paper" instruments.

The biggest difference between CEF/GTU and the ETFs is that the ETFs are designed to closely track the spot price of the underlying metal by maintaining an open invitation to their authorized participants (not you and me) to exchange shares for gold (or vice versa).  If the market price drifts from the spot NAV, the authorized participants can make money through these exchanges.  If you watch, gold is coming into and going out of the ETFs all the time (sometimes in very large amounts).  This activity tends to keep the price close to the spot NAV, but the reasons the authorized participants are putting gold in or taking gold out of the ETFs are not at all obvious.

On the other hand, each share of CEF/GTU simply reflects the market value of one share's worth of the gold (or gold and silver) the fund owns.  This market value might be more or might be less than the spot price of the metal - it's whatever people are willing to pay for one share's worth of metal that is securely stored, insured, and audited.  These funds are set up as very straightforward trusts.  They buy metal.  They put it in a vault.  The metal sits there.  Very dull and boring.  CEF has a 40+ year history.

CEF/GTU funds occasionally buy more of (and can at least theoretically sell) the underlying assets, creating (or destroying) a proportional number of shares.  Usually they buy when the shares are selling at a fairly high premium which allows them to create shares at a premium, but less than the prevailing premium.  GTU just announced such an event today.  Unlike the ETFs, the fund (not the market makers) decides when these events occur, they're announced ahead of time, and both the fund and the market makers who are involved profit (the profit comes from the difference between the spot price of the additional metal and the market price of the shares).  The fund uses the profit to pay ongoing expenses.  In contrast, only the authorized participants profit by creating/destorying shares of the ETFs and the ETF expenses are paid by essentially taking a small amount of the gold the fund oversees (about 0.4% per year).

The bottom line is CEF/GTU and the ETFs are similar in many ways and, while the ETFs track the spot price much closer, they are far more complicated entities with many more moving parts (each of which is a potential point of failure under stress).

[Storm]

Thank you for the insight into the way gold futures and GLD works.  This is one of the great things I love about this board - you can get real, insightful analysis from experts in the industry.  I shudder to read a lot of comments on zerohedge from gold bugs that think they know it all.

[stone]

rickb "the reasons the authorized participants are putting gold in or taking gold out of the ETFs are not at all obvious"

Why do you say that? I thought that it was obvious. Whenever a small premium or discount creeps in, the authorized participants redeem shares for gold or vice versa. So if $1 worth of gold can be exchanged for 99c worth of GLD then they annul the shares and remove the gold and if $1 worth of GLD can be exchanged for 99c worth of gold, then they create more shares and add gold to the vault. Am I missing something or in a muddle?

[rickb]

rickb "the reasons the authorized participants are putting gold in or taking gold out of the ETFs are not at all obvious"

Why do you say that? I thought that it was obvious. Whenever a small premium or discount creeps in, the authorized participants redeem shares for gold or vice versa. So if $1 worth of gold can be exchanged for 99c worth of GLD then they annul the shares and remove the gold and if $1 worth of GLD can be exchanged for 99c worth of gold, then they create more shares and add gold to the vault. Am I missing something or in a muddle?

Here's the daily difference between the closing price and NAV (in dollars), and the next day's difference in reported ounces held by GLD since 9/1/2011 (from http://www.spdrgoldshares.com/sites/us/ ... l_archive/):

09/01/11 0.58 0
09/02/11 0.66 0
09/06/11 -1.57 0
09/07/11 0.87 29205.27
09/08/11 1.19 -337962.71
09/09/11 0.51 0
09/12/11 -1.86 19469.04
09/13/11 1.37 0
09/14/11 0.19 0
09/15/11 0.93 -340696.02
09/16/11 1.4 0
09/19/11 -1.36 -9733.75
09/20/11 0.67 0
09/21/11 -0.93 0
09/22/11 1.44 0
09/23/11 -4.59 175189.41
09/26/11 2.05 155724.35
09/27/11 -0.84 0
09/28/11 -3.71 321170.95
09/29/11 0.72 0
09/30/11 0.42 -9732.13
10/03/11 -0.13 87587.85
10/04/11 -1.78 0
10/05/11 2.13 0
10/06/11 1.38 0
10/07/11 -1.59 48657.07
10/10/11 1.6 0
10/11/11 0.27 0
10/12/11 -0.42 15519.07
10/13/11 1.15 0
10/14/11 0.11 0
10/17/11 -1.05 0
10/18/11 3.14 0
10/19/11 -1.05 0
10/20/11 0.14 0
10/21/11 -0.3 -194598.04
10/24/11 0.28 -340543.07
10/25/11 4.46 0
10/26/11 0.44 19459.38
10/27/11 2.78 0

If the closing price and NAV difference is positive the authorized participants can make money by depositing gold (I'd think the next day's reported ounces difference should be positive) and if the price difference is negative the authorized participants can make money by withdrawing gold (the next day's reported ounces difference should be negative).

Hmmmm.

If anyone would like to explain these numbers to me I'm all ears.  As far as I can tell, they look kind of random - which would actually tend to keep GLD's price in line with its NAV as well as anything else. 

My point is that unless the price vs. NAV difference is truly ridiculous (say, greater than 5% in either direction, which at GLD's current price would be between $7 and $8) the daily variability in the ETF's share price and the spot price of gold (which both vary due to market factors) makes the net outcome of exchanging gold for shares a complete crap shoot.  So, why are the AP's buying and selling so often?  It certainly looks like when they do this they're not always making money on the trade.  Perhaps they're making enough money on some other part of whatever total deal they're doing that this part of the trade is irrelevant - i.e. keeping GLD's share price close to its NAV is not the primary factor in what they're doing but as long as it's close it's good enough.

I'm not saying this is necessarily bad, just that there seems to be a lot more going on here than meets the eye and that whatever it is isn't necessarily good for GLD's shareholders.  Again, it might not be bad but if you don't have any idea what's going on or why it seems like a pretty curious place to keep 25% of your assets.

[stone]

rickb, thanks for posting that up. I only knew the idealized sales blurb not the murky detail. To be fair to them though, GLD has not drifted away from the spot price to any great extent -so far..... Perhaps intra day price movements account for the apparent weirdness of what they do?? Perhaps you are correct and they do borrow gold from the vault even it it means they make a loss and the fund makes a gain. That would in effect be like the security lending done by bond etfs? Thanks again for all of this.

[rickb]

rickb, thanks for posting that up. I only knew the idealized sales blurb not the murky detail. To be fair to them though, GLD has not drifted away from the spot price to any great extent -so far..... Perhaps intra day price movements account for the apparent weirdness of what they do?? Perhaps you are correct and they do borrow gold from the vault even it it means they make a loss and the fund makes a gain. That would in effect be like the security lending done by bond etfs? Thanks again for all of this.

The APs are certainly not supposed to "borrow" gold from the vault - although this is perhaps only a technicality since they can exchange ETF shares for the gold (or vice versa) and can presumably make this exchange with borrowed ETF shares.  Note that the fund doesn't make any money from this in either direction (they charge a fee for the transaction, but it is definitely not a function of the difference in value between the gold and the shares).  GLD has certainly tracked the price of gold quite well - no doubt because its shares and the underlying gold are effectively fungible.  But this also means if the APs can borrow shares they can effectively borrow the gold.

[stone]

rickb, on reflection, I still don't think this poses any potential problem. Because GLD shares get destroyed whenever an AP takes gold out, each GLD share actually corresponds to MORE real gold if an AP does this for some motive other than simply to get the GLD price and NAV aligned. Perhaps the seemingly irrational flows of gold into and out of the GLD vault are motivated by say jewelry manufacturers using GLD a stop gap supplier of gold. A jewelry manufacturer (or even a short seller in a pickle) could hire the AP to exchange GLD for gold and so ensure that they got the gold they needed. If the jewelry manufacturer was unlucky and GLD was selling at a slight premium over NAV, then the consequence would be that each GLD share was caused to subsequently hold 0.095 rather than 0.094 ounces or whatever. This is a benefit to everyone who owns GLD shares not a risk. Is my thinking on track??

[rickb]

rickb, on reflection, I still don't think this poses any potential problem.

We've already established GLD tracks the price of gold during normal times.  The question is what could happen during abnormal times?  Say an AP borrows a few million shares of GLD and exchanges these for the underlying gold intending sometime later to exchange gold back for the borrowed shares.  On GLD's books everything looks good.  This transaction has no effect on the NAV (gold was withdrawn, but a corresponding number of shares were destroyed).  So far, so good.  And if things go as planned, the gold gets returned and the number of shares goes back up (again, with no effect on the NAV). 

But what if something bad happens before the gold is returned causing the price of gold (and, hence, GLD) to spike up by 15% in a day?  Whoever the AP borrowed the GLD from demands an increase in their collateral (securities lenders mark to market every day).  If whatever bad thing that has caused the price spike also affects the AP in a very bad way, the AP simply might not have the money.  If it's bad enough (think LTCM or Bears Stearns) the AP defaults on the loan leaving the lender of the shares holding the bag (they have cash equivalent to the previous day's market price of the shares).  The lender in this case is a fund of some sort, or a brokerage who has lent shares out of its customers' margin accounts.  What has happened is the lent shares have simply vanished (replaced with cash, but not as much cash as today's closing price of GLD).  GLD's books look fine.  No NAV change.  But some poor patsy suddenly doesn't have his shares of GLD and has only as much cash as his shares were worth yesterday (i.e. misses the 15% spike). 

To be fair, there's nothing in this scenario specific to GLD - the same problem could occur with any security that can be lent including CEF or GTU.  However, in the case of GTU borrowing shares doesn't let you get your hands on the underlying gold, all you get is shares of GTU which you can't convert to gold or anything else.  You can borrow them to sell them short, which could conceivably have the same net effect (borrower defaults, lender gets screwed) - but open short interest is something you can determine if you look hard enough and the motivation of the borrower (to short GTU) is pretty clear.  In the GLD scenario, the fact that the shares are borrowed is totally obscured and the AP has gotten their hands on a pile of gold for completely unknown purposes.

Scenarios where the market price of GLD and its NAV decouple aren't too hard to come up with either - most of the ones I can think of off hand involve outright fraud on the part of an AP (e.g. an AP deposits tungsten filled gold bars in exchange for shares - the custodian finds out later and freezes all AP exchanges while they assay all the bars during which interval the market price plummets).  Like I said before - lots of moving parts, each of which is a potential point of failure.

[stone]

rickb, GLD is an open ended fund. It is supposed to get larger and smaller. Its just like TLT in that regard. Just as TLT adds bonds  when money flows in (and vice versa) so GLD adds gold when money flows in. The difference seems to be that with GLD sometimes rather than it being investors taking money out and so causing shares to be annuled, it is sometimes AP taking gold out and so causing shares to be annuled. The gold that is taken out in that way never need be returned. Because the GLD shares have been annuled, it is not owed to the fund. It can be turned into wedding rings etc and never seen again. I'm certain that is how it is supposed to work??
Your scenario seems to revolve around AP being short GLD. I completely agree that if AP were ever short GLD then that would be a massive hazard BUT is there any evidence that they ever have been? I thought they bought (not borrowed) GLD to annul it in return for getting gold.  The are apparently massively short gold futures but that doesn't seem relavent to me. I hope I'm not in too much of a muddle.

[Larshus]

And where does CEF stand in relation to the other etfs ?  Thanks

CEF (a gold/silver mix), and its gold-only twin fund GTU, are closed end funds rather than ETFs.  Like the ETFs, when you buy/sell shares of these you're buying/selling them on the open market from/to someone else - and doing this does not cause the fund itself to buy or sell anything.  Also like the ETFs, ordinary shareholders of CEF/GTU have no chance whatsoever to redeem their shares for metal - so they're both "paper" instruments.

The biggest difference between CEF/GTU and the ETFs is that the ETFs are designed to closely track the spot price of the underlying metal by maintaining an open invitation to their authorized participants (not you and me) to exchange shares for gold (or vice versa).  If the market price drifts from the spot NAV, the authorized participants can make money through these exchanges.  If you watch, gold is coming into and going out of the ETFs all the time (sometimes in very large amounts).  This activity tends to keep the price close to the spot NAV, but the reasons the authorized participants are putting gold in or taking gold out of the ETFs are not at all obvious.

On the other hand, each share of CEF/GTU simply reflects the market value of one share's worth of the gold (or gold and silver) the fund owns.  This market value might be more or might be less than the spot price of the metal - it's whatever people are willing to pay for one share's worth of metal that is securely stored, insured, and audited.  These funds are set up as very straightforward trusts.  They buy metal.  They put it in a vault.  The metal sits there.  Very dull and boring.  CEF has a 40+ year history.

CEF/GTU funds occasionally buy more of (and can at least theoretically sell) the underlying assets, creating (or destroying) a proportional number of shares.  Usually they buy when the shares are selling at a fairly high premium which allows them to create shares at a premium, but less than the prevailing premium.  GTU just announced such an event today.  Unlike the ETFs, the fund (not the market makers) decides when these events occur, they're announced ahead of time, and both the fund and the market makers who are involved profit (the profit comes from the difference between the spot price of the additional metal and the market price of the shares).  The fund uses the profit to pay ongoing expenses.  In contrast, only the authorized participants profit by creating/destorying shares of the ETFs and the ETF expenses are paid by essentially taking a small amount of the gold the fund oversees (about 0.4% per year).

The bottom line is CEF/GTU and the ETFs are similar in many ways and, while the ETFs track the spot price much closer, they are far more complicated entities with many more moving parts (each of which is a potential point of failure under stress).

What he said. rickb nailed it.

[rhymenocerous]

And where does CEF stand in relation to the other etfs ?  Thanks

CEF (a gold/silver mix), and its gold-only twin fund GTU, are closed end funds rather than ETFs.  Like the ETFs, when you buy/sell shares of these you're buying/selling them on the open market from/to someone else - and doing this does not cause the fund itself to buy or sell anything.  Also like the ETFs, ordinary shareholders of CEF/GTU have no chance whatsoever to redeem their shares for metal - so they're both "paper" instruments.

The biggest difference between CEF/GTU and the ETFs is that the ETFs are designed to closely track the spot price of the underlying metal by maintaining an open invitation to their authorized participants (not you and me) to exchange shares for gold (or vice versa).  If the market price drifts from the spot NAV, the authorized participants can make money through these exchanges.  If you watch, gold is coming into and going out of the ETFs all the time (sometimes in very large amounts).  This activity tends to keep the price close to the spot NAV, but the reasons the authorized participants are putting gold in or taking gold out of the ETFs are not at all obvious.

On the other hand, each share of CEF/GTU simply reflects the market value of one share's worth of the gold (or gold and silver) the fund owns.  This market value might be more or might be less than the spot price of the metal - it's whatever people are willing to pay for one share's worth of metal that is securely stored, insured, and audited.  These funds are set up as very straightforward trusts.  They buy metal.  They put it in a vault.  The metal sits there.  Very dull and boring.  CEF has a 40+ year history.

CEF/GTU funds occasionally buy more of (and can at least theoretically sell) the underlying assets, creating (or destroying) a proportional number of shares.  Usually they buy when the shares are selling at a fairly high premium which allows them to create shares at a premium, but less than the prevailing premium.  GTU just announced such an event today.  Unlike the ETFs, the fund (not the market makers) decides when these events occur, they're announced ahead of time, and both the fund and the market makers who are involved profit (the profit comes from the difference between the spot price of the additional metal and the market price of the shares).  The fund uses the profit to pay ongoing expenses.  In contrast, only the authorized participants profit by creating/destorying shares of the ETFs and the ETF expenses are paid by essentially taking a small amount of the gold the fund oversees (about 0.4% per year).

The bottom line is CEF/GTU and the ETFs are similar in many ways and, while the ETFs track the spot price much closer, they are far more complicated entities with many more moving parts (each of which is a potential point of failure under stress).

What he said. rickb nailed it.

So, excluding tax treatment, are you saying that it is preferable to hold GTU over GLD/IAU because the way the fund is set up is more straightforward and less prone to manipulation?