29 June 2008

The Gold Value Chain Part III - Manufacturing

There are two types of gold manufacturing – big margin/fixed price or low margin/variable price. An example of the former is jewellery, of the latter, coins and bars. For this blog I will focus on the latter because this business is more exposed to changes in the gold price – when you are dealing in products with single digit margins there isn’t a lot of room for error.

So let’s say you want to help all the goldbugs out there by manufacturing a range of good quality bars at reasonable prices. You have invested a bit of your own cash and borrowed some from your friendly banker. This helps pay for rent, equipment, wages etc but your biggest cost (and most risky) is the raw material for your products – gold. Problem with gold bars and coins is that the gold value of the product is a substantial part of the overall price – for example a 1oz coin sells for gold value + 6%. I doubt if you looked at the raw material costs in normal products like a car or table you would find that the steel and plastic or wood was 94% of the retail price. This is an issue on two fronts: firstly, you have to tie up a lot of capital in inventory and secondly, you have to manage the volatility of the gold price.

In normal manufacturing the process is: borrow cash, buy raw material, value add to it (add some staff costs and your intellectual property of how to transform the raw materials into a product), sell the product, use the cash to pay back the loan and hopefully has some left over cash (ie profit). Easy, makes you wonder why anyone can go bankrupt.

What happens if you try and do this with gold. OK, say you borrow $1010. Buy 1oz @ $1000/oz. Pay staff $10 to beat it into a bar shape and stamp it. You think $10 profit is reasonable so offer to sell it for $1020. However, while you were making the bar the gold price has dropped to $950/oz. Funny thing about goldbugs is they will only pay current market price for gold. They say to you “that is a nice bar, I agree there is $20 worth of fabrication in it, so I’m happy to pay $970. “But”, you say, “it cost me $1010.” “I don’t care, it is only worth $970” says the goldbug. So unreasonable of them, isn’t it?

Sure, the gold price could have went up, but if it dropped while you were making your first batch, you’re out of business before you have started. You could wait until the price moves above your purchase cost, but this could take a while and you have to pay rent and staff costs in the meantime. This isn’t really the way to run a business. You put your thinking cap on and come up with the idea of asking people to buy from your first, and then you’ll make it. But you find out that goldbugs are very trusting and don’t want to wait a couple of weeks while you make their bar – they want to exchange their cash for a gold bar now. You also find out that other bar manufacturers seem to have bars ready to sell immediately, so you solution isn’t very attractive compared to your competitors.

Thankfully you’ve found this blog and ask me how your competitors get around this problem and I tell you there are two ways – buy and hedge or lease.

Buy and Hedge

This business “model” works like this:

1. Borrow $1010 cash (ignoring interest for the moment)
2. Buy 1oz gold @ $1000/oz
3. Forward sell 1oz gold @ 1000/oz (or do the same on a futures market) ignoring the time value of money and any margin deposits required
4. Spend $10 making your product
5. Price drops to $950oz
6. Sell the bar for $950 gold value & $20 fabrication cost
7. Close out the forward sale or futures contract. As you have a contract to sell in the future @ $1000 and the current price is $950, you get paid $50 profit on the contract
8. Repay the $1010 loan

Cash flows in this example are: +1010 (step 1) -$1000 (2) -$10 (4) +$970 (6) +$50 (7) -$1010 (8) = $10 profit, yippe. This seems pretty good and it is the legitimate use of futures markets. There are a few negatives, however. Firstly, there are transactions costs involved with the hedging. Secondly, you have to estimate how long to hedge for, that is, estimate when you think you are going to sell your product. You might know exactly how long it will take to make your bar, but you can’t be sure about whether there will be demand for it when you’ve finished making it, as that is determined by the gold price itself (e.g. if the price is falling, there may not be much investment interest in gold bars). Thirdly, you have to estimate how much you are going to sell at that future time. Again, this depends upon demand which is influenced by the gold price.

You might think these uncertainties are manageable, and they are to an extent. You can run small production batches, hedge for short time periods going forward and roll the contracts if necessary. Problem is that demand for gold is as volatile as the gold price. Take my word for it, you goldbugs are fickle. If gold is hot and moving up, demand can double, triple instantly. Same on the downside. This means that you won’t get your estimates right all the time and either lose money, be stuck with stock, or missing opportunities to sell more product. In the end you make up for these by having to raise your profit margin, which in the end means goldbugs pay through higher fabrication prices. But, better than not having any bars to buy at all, right?


The buy and hedge/forward sell/futures method does eliminate most of the risk dealing with a raw material like gold whose price fluctuates. There is a better way, however. The leasing business “model” works like this:

1. Borrow $10 cash (ignoring interest for the moment)
2. Lease 1oz of gold
3. Spend $10 making your product
4. Price drops to $950oz
5. Sell the bar for $950 gold value & $20 fabrication cost
6. Immediately buy 1oz of gold @ $950
7. Repay the 1oz lease
8. Repay the $10 loan

Cash flows in this example are: +10 (step 1) -$10 (3) +$970 (5) -$950 (6) -$10 (8) = $10 profit, yippe.
Gold flows in this example are: +1 (step 2) -1 (5) +1 (6) -1 (7) = 0oz.

Now you might say, well, that isn’t much different to the Buy and Hedge method, same profit, same issues with estimating when and how much will be sold and thus how long to lease for. Yep, but there is one big difference – the funding cost. With Buy and Hedge you have to borrow the $1000 to buy the gold. Currently in Australia the interest rate would be, say 9%. With leasing the “interest” rate is more like 0.3%. Hmm, lets see what the difference in interest would be if you held 100,000oz of work in progress and finished product inventory over a year:

Borrow cash: $1000 x 9% x 100,000oz = $9,000,000
Borrow gold: 1oz x 0.3% x 100,000oz = 300oz, which at $1000/oz = $300,000

I don’t think you have to be a great business brain to realise that and extra $8,700,000 is a freaking big difference. But you are too worldly so ask me “too good to be true, what is the catch?” Yes, well you see bullion banks are as trusting as goldbugs. They aren’t going to just give you some gold without some sort of surety, some collateral. The fact that you are running a gold manufacturing business is some comfort, but how do they know you won’t just run off with the gold? So they ask for $1000 cash to cover them in case you can’t repay the lease. But you have to borrow that $1000 @ 9%, so not much point then with the leasing method. Well yes, except if you have a credit rating acceptable to the bullion bank, then they will just lease you the gold without any need for cash collateral.

A little unfair, I suppose, for you as you are just starting up and haven’t got a S&P rating, but then in business, those with the credentials get advantages that smaller competitors can’t. Advantage for the gold bugs is that businesses using the leasing method can produce products cheaper than others.

Next week I’ll expand upon the leasing method and some of the mechanics involved.

27 June 2008

Precious Metal ETF Holdings

I found this short but interesting comment by Tim Iacono on Seeking Alpha about whether changes in GLD's holding have anything to do with the price of gold: http://seekingalpha.com/article/82626-does-gld-inventory-affect-the-price-of-gold

This was a topic I was planning to cover in a future blog because I have seen other commentators analysing GLD creations and redemptions. I feel some caution needs to be exercised with such interpretations. I'll expand upon this in the future, but in the meantime here is my reply to Tim that briefly explains my caution:

I'm a bit wary of reading too much into changes in GLD holdings over the short term because of the inherent lack of transparency in the gold market. As most gold trading is over the counter (OTC) and not all done on a nice visible stock exchange, you can't be sure that positions in GLD are not offset in other markets.

The GLD (or any ETF) redemption/creation process involves costs, so it is more profitable for market makers in GLD to avoid this where possible. For example, where retail investors are selling GLD, the normal (ideal) process is for the market maker to buy GLD from them and sell gold on the OTC spot market. They redeem GLD for physical gold and use this physical to settle their OTC spot sale.

However, if the market maker feels that the sell off in GLD is temporary and that retail investors will come back in the future, then they can make more profit by holding GLD and avoiding redemption/creation cost. They still have to buy GLD and still sell gold OTC so that they do not have a trading position and any exposure to the gold price, but instead of redeeming GLD, they lease gold in the OTC market and use that leased gold to settle their OTC spot sale. Their long GLD position (asset) is offset by a lease (liability). Considering that gold lease rates are 0.2% there isn't much holding cost with this strategy.

The only time a market maker would then redeem GLD for physical gold is if there is sustained selling over a period of time. In this situation the market maker's holding of GLD would continue to grow. They then redeem and use the gold to repay the lease.

As a result, I feel that analysis of GLD's (or any other gold or silver ETF) redemption/creation flows against the gold price is only realiable if done in time period blocks of a month or more.

24 June 2008

The Gold Value Chain Part II - Refining

Refiners exist because the gold that comes out of the mining process varies in purity and size but trading is more efficient if the units of trading (i.e. gold bars) are standardised. Their basic job is to convert dore (the stuff from mines) into a tradable form. Organised markets (e.g. COMEX) or trading associations (e.g. LBMA www.lbma.org.uk) set standards and rules for acceptable forms for gold and accredit refiners or their products so that market participants know exactly what they will get (or have to deliver) when they deal with each other. Very sensible.

What is that tradable form? The LBMA standard for gold is 400oz+/- @ 99.5%+ purity. Two interesting features here: first the weight is not exact and second the purity is not exact or 99.99% (the purity retail investors usually get with coins and small bars). Why? Because it is cheaper this way. We are dealing with wholesale markets here, so they don’t want to pay unnecessarily for higher purity or exact weight, especially if the physical bars will be alloyed down to 9ct, 14ct or 18ct for jewellery anyway.

With weight, it is cheaper to just pour molten gold into a mould and as long as you fill it up to the correct height, you can get a bar within +/-10% of your target weight. To get to exact weight, you need to precisely weigh out small gold granules then melt them and put into a mould (or vice versa). This process can be mechanised/automated, but there is still an additional cost involved.

With purity, 99.5% purity gold can be produced using a chlorination process (impure gold is melted and gaseous chlorine is blown through it with the impurities joining with the chlorine). It is rapid and simple and therefore cheap but only goes up to 99.5%. To get to 99.99% you need an electrolytic process (stick impure gold into a solution of hydrochloric acid and gold chloride, pass an electric current through it, the gold dissolves and pure gold moves to a negatively charged electrode). Problem is that this process costs more, mainly due to the need to keep an inventory of gold chloride on hand. It is also slower than chlorination, so you tie up gold in the process for longer.

Whether a miner has pre sold their gold or wants to sell it for spot (current cash) price, do to so they need to deliver it in a tradable form. The mining process gets gold up to a reasonable purity, but it is more efficient to give this to a refiner to get it to the acceptable 99.5% purity than for the miner to do it themselves. It is also easier for the gold industry as a whole to accredit the production processes of a few refineries than those of many miners.

So your miner (or prospector, or someone with scrap) goes to their local refinery and negotiates a refining contract. If they don’t have much bargaining power or history with the refinery, the refinery will say “look, I don’t know what the purity is of this stuff you’ve given me, so I’ll pop it through my processes and in a couple of weeks I’ll tell you how much pure gold there was in it." It is worth noting that in this example, the refinery doesn’t actually own the gold, the miner still has title to it and the refinery is just processing it for them – sometimes known as toll refining. The refinery charges a fee for this service and theoretically at the end the miner could ask for a bar of 99.5% gold. In practice, few actually do this. Why?

Firstly, the amount of gold delivered for refining rarely equals an exact quantity of 400oz bars. If it is a prospector, then the amount may not even be 400oz. So for small lots you would end up with a bar that, while of a tradable purity, is not in a tradable size. Secondly, miners (and a fair number of prospectors) are really after cash and not physical bars. They don’t want the hassle of taking delivery from the refinery, finding a buyer and arranging shipment.

So their friendly refinery offers to do all this work for them, and therefore most usually sell their refined gold directly the refinery itself and leave them with the work of finding a buyer and shipment. This process of buying from its customers and on-selling to someone else is simple enough but it gets a bit more complex when a miner wants to settle its hedging contracts.

Loco Swaps

Miners can hedge either via COMEX futures or a deal with a bullion bank. In either case, for the miner to settle its contracts it needs deliver refined gold to a COMEX warehouse or to the account of a bullion bank in London. From the point of view of the industry as whole, however, this is not always efficient.

For example, with a huge demand in India for 99.99% kilo bars it would not make much sense for an Australian miner to ship 99.5% 400oz bars to London, then the bullion bank to ship it to a refinery, which reprocesses into 99.99% kilo bars and then ships it again to India for eventual sale. Australia’s refinery, AGR Matthey, would rather keep the gold, immediately further process the gold into 99.99% purity and directly ship to India. This is cheaper, keeps the price down which means Indians can buy more gold, which we would all agree is A Very Good Thing.

But the miner needs gold in London. So, continuing with our example, AGR Matthey offers to do a loco swap (short for location swap). In effect, the deal is “you give me title to your refined gold in Australia and I will give you title to gold in London.” The miner and refinery are swapping gold in different locations. This begs the question “why has the refinery got gold in London” and the answer is “it doesn’t have any”. Whaaat, you say? This is where our goods friends the bullion banks and their buddies the central banks come in.

As noted in previous blogs, the central banks are sitting on lots of gold which isn’t earning them anything so they ask bullion banks to try and earn a return on it for them. So AGR Matthey rings up a bullion bank and asks if it can lease 400oz gold for a month. "No problem", says the bullion bank, "I’ve got heaps sitting around to lend to you". AGR Matthey now has 400oz physical gold in London (asset) but also a 400oz lease to repay in a month (liability).

Just a side note here: AGR Matthey’s ounce assets and ounce liabilities match equally, so it is not exposed to any movement in the price. For example, if the price drops from $1000 to $800, the value of its physical gold goes down to $320,000 but the liability also drops to $320,000. Its (ounce) balance sheet always equals zero.

With gold in London, AGR Matthey can now do the swap with the miner. It transfers the London gold from its account to the miner’s account and takes control of the physical gold in Australia. The miner can then use the London gold to meet its contractual obligations, or if it is unhedged, simply sell it. From AGR Matthey’s point of view, it still has a liability (in London) for 400oz, but instead now has 400oz of physical gold in Australia rather than London.

Sale of Gold

The process of refining up to 99.99% and then finding buyers in India and shipping it to them takes time, hence that’s why AGR Matthey asked for a one month lease. During this time it owns the physical gold, offset by the lease liability. It ships it to India and when it has found a buyer, it sells the physical gold to them and using the cash from the sale, immediately buys 400oz of gold in London. It has lost title to physical gold in India but gained title to gold in London. It can now use this London gold to repay the lease to the bullion bank, who can return it to the central bank, plus the lease fee.

This process is similar in a way to what I described with miners and hedging. Once the loco swap has been performed, ultimately the lease from the central bank is “secured” or “backed” by the physical gold held in the refinery. There is no short selling involved. The central bank, by providing gold for lease, has actually facilitated the manufacturing and selling process of physical gold, enabling it to be done efficiently and therefore cheaper. As I’ve said before, leasing itself is not bad; it is who gold is leased to and what they do with the gold that matters.

Next week we’ll go into more detail about gold manufacturing and trading. Leasing will feature again and we’ll introduce loco discounts and metal accounts.

11 June 2008

An empire of imaginary metal

Rich List businessman Virendra Rastogi jailed for £350m fraud.

Only found out because a PwC auditor became suspicious when documents were all sent from the same fax machine. Maybe if auditors actually physically visited some of the "customers" they would have found out earlier, but then it is much easier to just check paper confirmations from clients and check the debits and credits add up - wouldn't want to do any real work.

09 June 2008

The Gold Value Chain Part I – Mining


So let’s say you have been lucky enough to find some land with gold in it. You estimate that there is 1 million ounces in it but to get it out you need to build a mine and employ people and to do that you need (paper) money. You can get that either by selling some shares in your new company Goldmine (equity) or by borrowing some cash from your friendly banker (debt). Ideally, the source of funding you would choose would be the cheapest, but you will probably find that there are limits to how much you can get of the cheapest source of funding. For example, a bank is unlikely to lend you the entire amount you need without yourself or other investors putting some cash in. This does seem a bit unfair because they have been stupid enough to lend people 100% of their house price (what happened to the days of 10% deposit) but I think they are learning their lesson (yet again).

As a result Goldmine is probably going to need a bit of debt and equity. Now the equity people accept a bit of risk and understand that they don’t have any guaranteed return on their investment, but they think you are trustworthy and are happy with your projected profitability so are prepared to give you a go. Generally for that risk they are expecting, however, to earn more than the rate they could earn by lending their cash out. Let’s say for simplicity you have calculated that they will earn 10%. The bank however, is not into that sort of risk so they want a charge on the assets of Goldmine or mortgage on your assets but in return they just expect a boring cash return of 5%.

The process I’ve described above is pretty much common to any business, not just mining. Debt is usually cheaper than equity, but it is limited and requires guarantees of some sort. But things are a bit different with gold, however, because Goldmine also has the ability to borrow gold itself, which makes a bit of sense because it is gold that you want to get out of the ground. Why would Goldmine want to borrow gold? Well let’s enter the world of central banks, bullion banks and leasing.


One of the first things that struck me when I started in this industry was the fact that you can borrow gold (however in the gold industry they don’t use the word borrowing, but call it leasing). I found it intriguing because you can’t borrow copper or tin or other commodities and also because borrowing is something usually associated with money.

One of the funniest things I think is those commentators who rant about the evils of leasing. The fact that you can borrow gold, and that gold has an interest rate (called the lease rate), is one of the strongest argument that gold is a legitimate form of money. As a result, all people who do not believe in paper money should be supportive of the gold leasing market because it is proof gold is money and in a transition from paper money to hard money you will need a borrowing and lending market to exist for your hard (gold) money. It is pretty hard to set up the infrastructure and systems for borrowing/lending market from scratch, so supporting the existing leasing market means you are ensuring hard money will be ready to “rock and roll” when money = gold?

If you are new to this whole gold thing, you might be asking how did it come to pass that gold (and silver, and to a lesser extent, platinum and palladium) can be leased? My simplistic explanation is that in the good old days of the gold standard, money was gold (or convertible into it). What is interesting is that when countries went off the gold standard and money became just (fiat) paper, gold continued to be borrowed and lent. I suppose that is because central banks were used to treating gold as money and had arrangements and systems that treated gold as money, so just because money wasn’t gold anymore didn’t mean they couldn’t continue on as they had. They still held reserves of it and still wanted to earn a return on their assets, so why stop lending it out if other countries or legitimate businesses wanted to borrow it? Interestingly, it was Australian miners who rejuvenated the leasing market by realising its potential in supporting gold mining.

Before I get back on topic I want to talk about why the industry uses the word leasing instead of borrow/lend? My theory is that leasing implies ownership by the person lending it to you. The thing about paper money is that it is virtual, there is no real thing you have a claim on. The words borrow/lend refer to this virtual liability. In contrast, you can’t borrow a house; you rent/lease a house. You don’t borrow a car, you lease it. In these situations, the word lease is saying that I still own the house or car, but am letting you use it for a period of time for which you pay me a rental or lease fee. After you finish using it, you give the asset back and any gain or loss in the value of the house or car is the owner’s problem, not yours.

Given that gold is a hard asset, not virtual, I don’t think it is just chance that the industry uses the word leasing. It is a way of saying or reinforcing that the physical gold I give you is still mine, I benefit from any increase in its price – you are just “renting” it and can use it but it is mine and I want (the) gold back at the end of the lease period. Of course, in reality it is not possible to get the exact physical atoms back that were leased, so you only need return gold of the same weight and purity (just the same as if you lend your friend $20, you don’t expect the exact same note back with the same serial number on it).

The Lease Rate

So why would your company Goldmine be interested in borrowing gold instead of dollars. Well, because it is cheap, really cheap. Currently you can lease gold at 0.2% per annum, compared to USD cash rates of 2%, or AUD of 8%. Needless to say, as the manager of the company you would be crazy to turn down such a cheap form of funding, because it is going to increase the profit Goldmine makes, and therefore the dividend you can pay your shareholders.

Why is the gold lease (i.e. interest) rate so cheap? Well, basic supply/demand. An interest rate is the “price” of borrowing/lending something. If you have a lot of demand for borrowing and less lending supply, then the price (interest rate) goes up. So a low gold lease rate means there is a lot of lending supply, but not as many borrowers.

So who is on the supply side? Well those who hold gold and want to earn a return on it. As I noted in my last blog, central banks hold approximately 1 billion ounces and they are really the main supply of lent gold. In theory private investors could also lend their gold, but there aren’t any mechanisms in place in the market to deal with small players, so it really is a wholesale market and only accessible by central banks or large institutional holders.

Who is on the demand side, well three groups really – miners, fabricators and short sellers. The fact is there are not a lot in the first two groups relative to supply. In respect of the short sellers, banks (should) be reluctant to lend too much because of the risks involved (that is the topic of a whole other blog). Therefore there is a limited market demand to lease gold, hence the rate is low, and has been low, or lower than cash rates, for a very long time.


In a world where money is gold, your Goldmine company would simply borrow gold, use that gold to pay for equipment and wages, get the gold out, have it refined and then use that gold to repay the loan, with hopefully some gold profit left over to distribute to shareholders. However we live in a world where money is paper, so the people selling Goldmine the equipment and the workers want to receive paper dollars instead (crazy some would say). But even though we live with paper dollars, Goldmine could still do the first and last part – borrow and repay gold.

So in practice Goldmine leases just enough gold to cover setup costs and extraction costs. With lease rates at 0.2%, the “interest” bill is much lower than if it borrowed dollars. However it needs dollars, so it immediately sells the gold for dollars and uses those dollars to build the mine and pay workers.

Now the alarmists out there may say “that is not good, Goldmine has short sold and is exposed”. Not so, because the definition of short selling is that you have sold something you don’t own and will have to buy it back in the future. But Goldmine does have gold in the ground, so it has actually just sold gold it already has in advance, not sold short.

Let’s assume for simplicity that all up production costs (equipment and wages) for Goldmine are $300 million, the current gold price is $600 and that the whole 1 million ounces will be mined in one year. To get the $300 million to mine the gold, Goldmine leases 500,000 ounces at 0.2% (interest cost at the end of the year is therefore 1,000 ounces) and sells this at $600 per ounce to generate the $300 million.

Everything goes well and at the end of the year it has mined 1 million ounces. It repays the 500,000 ounce gold lease plus 1,000 ounce interest, leaving 499,000 ounces. This is sold at $600 per oz = $299.4m. Yippe, happy shareholders!

So what about the alternative scenario, where Goldmine just borrows $300m cash? Well at the end of the year it would have 1 million ounces which it sells for $600m. It repays the $300m loan plus $6m interest (at 2%). This leaves $294m cash, $5.4m less than if it leased gold.

What happens if the gold price goes to $900 per ounce at the end of the year? In the leasing example, Goldmine would still have leased 500,000 ounces at the start of the year because the price was $600 at that time. So at the end of year it sells the 499,000 ounces @ end of year price of $900 = $449.1m. Yippe!

Under the alternative method using cash borrowings, Goldmine would sell the entire 1 million ounces for $900m. It repays the $300m loan plus $6m interest (at 2%). This leaves $594m cash, $144.9m more than if it leased gold. 32% extra yippe!

Now you may say that leasing doesn’t look too good anymore. In a stable price market you are slightly better off but with a rising price you are a lot worse off. Correct, but what happens if the price drops below production cost, to say $200 per ounce?

In the leasing example, Goldmine would still have leased 500,000 ounces at the start of the year. So at the end of year it sells the remaining 499,000 ounces @ $200 = $99.8m. Interesting, even though the gold price is below production cost Goldmine still made money. Yippe!

Under the cash borrowings method, Goldmine sells the 1 million ounces for $200m. But it can’t repay the $300m loan plus $6m interest. It is $106m short and your shares are worthless. No yippes!

These examples demonstrate that leasing gold and selling it in advance (or forward selling) is a way of hedging (or protecting) against a fall in the gold price. In the current market where the gold price is well above the cost of mining gold, and is expected to go higher, there is little need to hedge. However if the gold price drops towards the cost of mining, then Goldmine would be crazy not to protect itself.

Hedging (and as a result, leasing, because it is what allows hedging in the first place) is not of itself bad in my opinion. Whether it is reasonable is entirely dependent upon one’s risk profile and assessment of the risk of the gold price declining. Those who say that no miners should be hedging and leasing should not be allowed as just dictators, forcing others to accept the risk/return tradeoff they consider acceptable. I don’t think that there is anything unreasonable about someone saying I’d rather Goldmine hedged so it will not go bankrupt if the price drops below $300 and I’m prepared to trade that off against higher profits if the price goes up.

However, the way gold mining is done is that the investor doesn’t really have control over the hedging decision. You may have a good company with solid management that you like, but a hedging program (or lack thereof) that you don’t like. Or vice versa. Too bad, they both come packaged together. It would be interesting to see a market where miners did capital calls on shareholders to pay cash costs as they occurred and disbursed all gold mined as dividends. Individual investors could then go into a separate retail market where they could enter into arrangements with a bank to forward sell their gold dividends if they thought the price was going to decline or simply sell at spot as they received the gold if they thought the price was going to rise. That suggestion is probably impractical (the company would have to go after those shareholders who got it wrong and couldn’t pay the future capital calls), and inefficient (bulk hedging by a miner with a bullion bank is cheaper than many small shareholders doing it), but the “I’m in control” part I like.

The examples I have went through above about Goldmine leasing gold directly and selling it in reality doesn’t occur. If a miner wants to hedge, they simply go to a bullion bank who offers them some sort of contract, be it a forward sale or option, customised around their unique circumstances and projected production. However, whatever financial instrument is constructed, behind the scenes the leasing of gold and its sale is ultimately involved.

Where’s the gold?

One final thing I would like to cover is the global “balance sheet” for gold in the example I have put forward, because there is a lot of talk about central banks and whether they have the gold or not and whether it is leased out and whether it is doubled counted. Let’s go back to our scenario of Goldmine leasing 500,000 ounces to hedge its production costs. At the beginning, the global gold stocks look like this:

Central bank asset - physical gold in vaults – 1,000 million ounces
Goldmine shareholders asset - physical gold in ground – 1 million ounces
Private investors’ asset - physical gold in vaults – 1,000 million ounces
Total – 2,001 million ounces

After the central bank lends gold to bullion bank who lends to Goldmine who sells it, the situation is:

Central bank asset – physical gold in vaults – 999.5 million ounces
Central bank asset – claim on bullion bank for leased gold – 0.5 million ounces
Bullion bank asset – lease to Goldmine – 0.5 million ounces
Bullion bank liability – lease to central bank – 0.5 million ounces
Goldmine shareholders asset – physical gold in ground – 1 million ounces
Goldmine shareholders liability – lease to bullion bank – 0.5 million ounces
Private investors’ asset – physical gold in vaults – 1,000.5 million ounces
Total – 2,001 million ounces

In this interim stage what has happened is that the central bank has traded physical gold for a claim to future physical gold, plus interest. Via the bullion bank through to the miner, on a global scale the central bank effectively “owns” part of Goldmine’s gold in the ground. However, it is a claim only – the central bank has a counterparty exposure to the bullion bank, which has an exposure to Goldmine. After the gold is mined and all leases repaid and Goldmine’s remaining gold sold the situation looks like this:

Central bank asset - physical gold in vaults – 1,000.001 million ounces
Private investors’ asset - physical gold in vaults – 1,000.999 million ounces
Total – 2,001 million ounces

In the end, the gold in the ground ends up with investors, with the central bank back with its physical metal plus a bonus 1,000 ounces. There is nothing inherently bad about this in my opinion, with the central bank facilitating the mining of gold, a good thing I think we would all agree.

The issue in the real world is does the central bank really know what the bullion bank is doing with the leased gold, does it know if it has been prudently lent out to reputable miners who are not excessively hedging or speculating (e.g. Sons of Gwalia)? The fact that a central bank has leased out gold is not a problem for me, the question is to whom and what have they done with it, and will they be able to repay the loan.

I’m a firm believer in free markets, so I don’t believe in regulating gold in any way and that includes leasing, or hedging or financial derivatives. No one had the right to tell someone else whether they should or shouldn’t lend their gold (or borrow it). But one can’t be half free, so I’m also a believer in free information. So the problem is not so much leasing itself, but that there is no transparency in the gold market in regards to the global gold “balance sheet”, telling us where the real physical is and where the paper claims are.

Next week, how refiners go about converting raw mined gold.

04 June 2008

How much gold is there?

The World Gold Council states that the total amount of gold ever mined is 158,000 tonnes, which we can round out to 5,080,000,000 ounces. I haven't been able to find out the methodology of how this number is calculated, would love to know. According to them this is composed of:

2,630,000,000 oz in jewellery
600,000,000 oz in industrial applications
830,000,000 oz held by private investors
915,000,000 oz held by central banks
105,000,000 oz unaccounted for

So we can clearly say that investment holdings are the 830 + 915 = 1745 million ounces (and for the point I'm making it is irrelevant if you believe central banks actually have it or not, if they don't have it, then someone else must). Most of the industrial stuff would be in non easily recoverable form (electronics etc), so that we can consider as not likely to come onto the market. Some of the "jewellery" could be considered investment (especially in India) with the rest of the stuff around Western women's necks only likely to come on to the market if the price was really really high, so at the moment can be considered non-investment (or long-term core deep investment holdings, depending on your viewpoint). Lets assume 9.7% of the jewellery is really investment, for no other reason than 1745 + 255 = 2 billion ounces, which is a nice round number.

So there is 2 billion ounces of gold out there as investment. As I mentioned in my last post on short-term trading, there isn't much transparency in the gold market. So what can we see?

Below is a chart of all the gold products/systems/ETFs, which I have been tracking weekly since 2003. Who is in this list? Well the big ETFs (World Gold Council's Gold Bullion Securities Australia, South Africa, UK, USA & Barclays USA), Central Fund of Canada, GoldMoney, e-gold, BullionVault and Milenium Bullion Fund.

The total ounces of these products comes out at 28 million ounces. COMEX warehouse stocks in March 2008 were 7.5 million. TOCOM stocks April 2008 were 270,000 oz. That means the total "transparent gold" the trader can analyse the movements of is approximately 36 million ounces.

That is about 1.8% of the 2 billion we estimate is out there. Looks like we have a little bit of way to go before we are fully informed.

02 June 2008

The gold value chain

Next week I am going to start a series on the gold value chain. We'll follow physical gold from miner to refiner to manufacturer (jeweller or mint) to distributor to end buyer (you).

I have seen a lot of misunderstandings out there about how things work in the gold industry and hopefully this will provide much needed background for people new to gold and maybe fill in the gaps for others who have picked up bit and pieces of knowledge as they have trawled the net.

01 June 2008

Investment timeframes - Part II

Short Term

Following on from last week, why do I say that the nature of the gold market itself makes profitable day trading difficult? It is all about information, or lack thereof. Apart from pockets of relative transparency like COMEX or ETFs, the vast majority of the market is opaque. The “retail” or “average Joe” trader simply does not have access to the same amount of information about the status of the market and its flows that a “wholesale” trader does (and even they can’t see the entire market). Without understanding what is really driving short term changes in the price, I doubt it is possible even for the most astute and disciplined trader to make consistent profits.

Let me contrast it to stock markets. There are a few key features that help the day trader. Firstly, you know exactly how many shares have been issued and if there are different classes, how many of each and what the differences/rights each has. Secondly, all trading is done through (usually) one regulated market. Thirdly, you can see daily trading volumes. Fourth, at any point in time you can see the depth of the market – how many shares are being offered to buy or sell at each price level. This mass of data, combined with analysis of price charts, gives the trader room to apply skill and a bit of gut instinct to the task of making a profit.

How does the gold market stack up on these features? Firstly, no one knows for sure at any point in time how much gold there is out there to be traded, nor in what form or in what locations. The wholesale market may have a bit of an idea, but no such information is published to retail traders on a daily basis. Even if one did know the size of the gold out there at that point in time, due to its refinability, scrap gold can flow back into the market quickly (a factor if you are planning on holding a position for a few weeks), so the total “shares on issue” in not fixed and changes in response to the price.

On the third and fourth points, there is no published information on daily trading volumes or market depth; indeed, no volume data is published at all. I’m talking here about the whole market. Sure, some gold is traded on regulated markets but that information is only part of the picture and certainly little of it is live. If you only have part of the story, you don’t have the story at all in my opinion.

Network Nature of the Market

The key killer for me is the second point – gold is simply not a publicly, regulated traded thing. And COMEX and ETFs and the like don’t go anyway towards solving that because they are not closed systems. It is easy to run a position in those markets that are offset or hedged with an opposite position in the spot/forward market. Detailing how that is done is for another blog. The gold market operates much like the internet – it is a network of wholesale dealers, independently trading with each other, and it is the sum of those individual trades that makes up the “spot market”.

It was always amusing to me when clients would ring up to buy and we would quote a price and then, naturally, they would say “Well, where can I get what the spot price is?” so they could work out if our price was “fair”. The answer was, “It doesn’t exist. You could spend a few thousand getting a live Reuters data feed, but even that is just indicative.” Being used to the comforts of a stock market, many didn’t like that answer and thought we were pulling a shifty on them. In the end, all we could say was that they had to do what we did, which was ring around to see who was offering the best price at that time. It made some uncomfortable, but as I pointed out in my first blog, this is the “pointy end” of investing, it’s real trading, it is about bargaining, haggling, being in the know.

The funny thing is that this network nature also gives the market strength. Transparency is nice, but not at the expense of robustness. Just like the internet, where if a part goes down then data can be rerouted, if London was nuked for example, then trading in gold could still continue. Sure, liquidity would be reduced, but as deals in the end are done over the phone, it is just a case of dealing with other counterparties in other countries. Because it is not locked in to one “exchange”, gold can be resilient in the face of a failure in part of the network. And this is how the medium and long term investors want gold to trade if it is to be the asset of last resort. The market needs that flexibility to if it is to continue trading.

But the network nature of the market and the corresponding lack of transparency is a problem for the day trader. To understand what the retail trader is up against, it may be better to explain what happens when they call up to buy some gold from a dealer. There are many small variations to how this can work; this is but one way, probably the simplest where a dealer just lays off a trade with someone else immediately instead of holding a position.

Spot Trading

Any trading desk needs an indicator of where the market is, and most use Reuters. However the price displayed on Reuters under code XAU is just an indicator. It is updated by the bullion desks of the big banks and is in effect, a bulletin board or forum where banks can publish their prices in the hope other dealers will call them up to do a trade. Sort of like an advertisement. Unlike a stock market, it is not a commitment to deal at those prices, but most times you can. However there are many times, especially when the market is moving quickly, when the dealers don’t have time to update their quotes on Reuters and so when you ring them up, they say “Sorry, Reuters off the market, my current price is $5 below the screen”.

As a result, when you call a dealer for a price, they themselves cannot really know exactly where the market is. They see $900 on the Reuters screen, but this is what they will be charged, so they have to add something on to it as they have to make a profit (you expect something for nothing, it costs to run a trading desk, to talk to you, to do the other side of the trade with the bank, to settle the funds, bank fees etc). The dealer also has to consider that by the time they get off the phone with you and then call another wholesale dealer the market may have moved, so they might need to add a buffer on top of their margin. Sometimes if your deal is big enough and the market volatile, they’ll get another trader on their desk to call a bank and get a firm price before they quote to you (and they’ll want an answer quick because the bank ain’t gonna want to sit on his quote for too long because he/she has got to trade it as well).

Dealers have a network of other dealers they trade with. Each dealer has a different bit of the gold market pie, they can see what is driving their deals (be they a refiner selling a miner’s gold, a bullion dealer selling coins and so on), but not necessarily what is driving other flows. They are in constant contact with each other, doing deals, talking and exchanging information on what they are seeing in the market, watching the Reuter’s price movements. They use all this information to set their prices and to ensure they don’t lose money. Over time they build up a gut instinct, a feel for market movements and where it might go that day.

If you call up the Perth Mint to trade, you are likely to speak to Deniece, the Mint's senior bullion dealer. She has what I would consider probably the best background training for a bullion dealer - croupier at Sun City. When the market is moving you can do any number of deals before you have time to enter them into the system to confirm your position and profit, so you have to be able to run them in your head, to know where all your 'gold chips' are on the 'table'. She has been there since 1994 and every working day she has been sitting in front of a computer screen watching the Reuter’s gold price tick up and tick down, talking to people like you wanting to buy or sell gold – that’s coming up to 4,000 days of trading. And you think you can day trade against dealers like her, with zero information about what’s going on in the market? Get real.