12 July 2008

The Gold Value Chain Part V – Trading & Loco

Loco is short for Location. Location is relevant for gold because it is a physical commodity and it costs money to move it between locations. People new to gold, who have been used to trading shares and bonds and other “virtual” products, can sometimes not consider the implications and risks of dealing in something that is physical.

I am sure most people would appreciate that buying physical gold involves freight costs – even if you are personally going to pick up some coins or bars from your local bullion dealer, you would still appreciate that there was some cost paid by the dealer in getting the gold shipped to them in the first place. This is one of the many costs associated with dealing in physical gold (another is insurance) and they are usually embedded in the margin the dealer charges, either added to the fabrication price or the spot price. I talked about spreads last week and the reason for focussing on that was that dealers can play around with how their prices are perceived by either adding parts of their margin to the spot price or fabrication charge. The spread is the only way to get to the bottom of this.

What I will focus on here is the spot price. A lot of times people think of the spot price like an exchange rate, a “base” price on top of which fabrication, freight and profit are added. Exchange rates are for fiat currencies, something as virtual as you can get. The spot price, by contrast, is for something that is ultimately physical. As a result, the spot price has a location cost embedded into it (or should have). One of the biggest misrepresentations by dealers is not really making this clear to investors and the risks associated with it.

In the industry “spot price” is really shorthand for “the price of gold located in London”. Why London? Because that is where, historically, gold was traded. It is where the major bullion banks have head offices and where some pretty big vaults are located and a fair amount of physical gold is located. Thus the price quoted on Reuters is for gold located in London, or in industry jargon, “loco London”. If you are dealing with a dealer who has trading accounts with bullion banks, the spot price they are quoting is effectively a loco London price. This is effectively gold’s “base” price.

If you deal in gold in other locations, e.g. loco Perth, the price has to be different to gold’s price loco London. Sure, you say, it costs money to freight gold between locations so I can understand that the spot price for the basic wholesale form (400oz bars) is different between locations. But the question is, who pays this difference? Answer is, supply and demand.

I’m sure everyone understands supply, demand and price. More supply than demand and price drops, other way around and price goes up. Same applies to gold in each location. For example, Australia produces way more gold than it needs domestically, so loco Perth supply is higher than demand. Thus miners can’t sell it all to Perth buyers and will therefore have to freight it to London first to get the loco London spot price. There is a cost associated with this, let’s say $1.00 for simplicity. In practise the miners don’t ship it and instead AGR Matthey does a loco swap (see Value Chain Part II blog below for details). However AGR says to the miner “since it would cost you $1.00 to ship the gold to London, I will only give you the loco London spot price less $1.00”. The miner in retort may say “yes, but you want the gold to make into bars to sell to Indians, so it would cost you $1.00 to ship gold from London to Perth”. In the end there is a bit of horse trading and they settle on a price, the most fair one being the mid-point, say $0.50. As a result, in this example gold loco Perth would trade at a discount to loco London gold, or in industry jargon, the loco discount for Perth gold would be $0.50.

This little game is played all over the world and each location trades at a premium or discount to London depending upon local supply and demand at that time and the relative bargaining power of the players. As a result, loco discounts/premiums are not fixed and change over time as supply/demand changes. Normally the supply/demand situation is stable, which is another way of saying that the physical flows around the globe are stable. However, this can change. For example, one major flow of physical gold is from Australia to India. But what happens when the gold price rises and Indian demand dries up. Then the gold flow is from Australia to London, or wherever the demand is. This changes the loco discount of Perth.

When a dealer sells to you, they quote the loco London price because they are backing this deal by buying gold loco London. They then want to swap this for, say, loco Perth gold. In this case, because loco London trades at a premium to loco Perth, AGR Matthey will pay the dealer $0.50 for the London gold in exchange for Perth gold. When a dealer buys back from you, the process is reversed. The dealer sells gold in London, but has your physical in Perth. They need to swap this for London gold so they can settle their sale trade in London. AGR will do a swap but the dealer is now in a similar situation to the miner, so they have to pay $0.50 to AGR. In the end the dealer gets $0.50 on the sale to you and then pays $0.50 when buying back, so it netts out. In normal markets (where the loco discount is constant) a dealer can therefore make their loco price the same as the loco London price.

This can therefore give investors the impression that there is one global spot price for gold. I think this is misleading because when markets change and there is sustained buying or selling imbalances in a location, the discount can start to become quite large and netting not possible, resulting in investors getting a price much lower than they expect. That’s why I am writing this, so you know the risks involved. You may not be able to do anything about it at the time, but at least you are aware. You also want to consider where you are storing your gold, because if it is in a location with low volumes/low liquidity or far away from London, the discount could become quite large. To understand these risks you need to understand how loco discounts are priced.

The main issue for investors is what price are you going to get when you sell. I am assuming that you are accumulating gold at these low low prices on the expectation that gold is going to go much higher. Hopefully you will time it right and sell before the peak, but what happens when the price starts dropping and all the people who rushed in at the end try and sell at the same time? In that situation you will have a glut of physical gold in that location for sale but few buyers. Technically the price could diverge from the loco London price by quite a bit as no one is buying. But then arbitragers will step in to buy loco Perth, ship it to London, and sell loco London.

Note that ultimately anyone can get the loco London price; they just have to ship their gold to London and sell it to a dealer there. Of course in reality small investors can’t do this. They won’t be able to arrange insured freight for small amounts, and have the risk that when the dealer gets the gold they just don’t pay on the basis that you aren’t going to spend thousands to fly over there. So this does present a great business opportunity to buy gold from all the idiots lemmings at a discount, aggregate it, ship it, and sell it in London at a higher price. So what is involved in arbitraging physical gold?

Firstly, your idiots are going to be so desperate with the gold price tanking that they will want their cash straight away. They aren’t going to want to wait for you to ship it to London. So what you have to do is lease gold loco London, sell that gold at the London spot price and use the cash to pay your idiots. Fine, but you now have to get that gold to London to pay back the lease. The longer you wait, the longer the lease runs for and the more your lease cost is. But you can’t ship the gold straight away, you have to wait for some more idiots to come along and buy gold from them until you eventually have a large enough pile of gold to get a good bulk freight deal. This means the all up cost to you to buy loco Perth is composed of:

1. Freight cost, which is dependent upon the size of the shipment
2. Lease cost, which is dependent upon the lease rate and the time it takes to accumulate your shipment size plus the time to ship to London

There is a bit of a balancing act here, because the longer you wait the bigger your shipment size and therefore the lower your per ounce freight cost, but the greater your lease cost. Depending upon the variables there will be an optimal point at which you should ship. Let’s do some numbers.

Let’s assume lease rates are 0.2%, price $1000 per ounce, freight time is about 3 days and let’s say it takes you about 12 days to accumulate your shipment. This means your per ounce lease cost is $1000 x 0.2% x 15 days / 365 days = $0.08 per ounce. Not a lot. What is the freight cost? Hard to say given it is highly volume related, but you can consider it to be around the tens of cents per ounce for shipments in the tonnes.

There are two important risks to be aware of here. Firstly, the lease rate. If you look at lease rates over the last two decades, it has averaged 1%, spiked above 3% for durations of months on 5 occasions and has got as high as 6-7%. If we plugged 3% into our model, the lease cost becomes $1.23, still lowish considering a spot price of $1000, but a big jump from $0.08.

Secondly, we are assuming we can actually freight it out when we want. There is only limited air freight capacity out of any location, and combined with the fact that insurers are probably not going to be happy covering a whole plane load of gold (lest the pilot decides to divert the flight), you could find yourself having to sit on gold loco Perth for a while until freight capacity becomes available. Alternatively, you could bid the freight price up to get earlier flights. Either way your costs are going to go up. With the price falling, you could have a fair amount of selling going on so could build up your loco Perth stock pretty quickly. Let say the lease rate was still 3% but you had to wait 2 months before you could get flights out of Perth. That is now a lease cost of $5 per ounce.

Now you might say what is the chance of that happening? Well consider that if the price is tanking then miners are going to start to want to hedge (or alternatively, miners decide it is wise to start hedging at these high prices before anyone else does, and then the price starts to tank). This means they need to borrow gold, which will push up the lease rate. Oil isn’t getting any cheaper, so freight costs will also be up.

Where it gets really scary is with silver. This is because it is bulky. At a ratio of 50:1, you simply cannot fly silver to London, it has to be shipped. How long does that take? Well, 2-3 months. Currently lease rates for that time period are say 0-0.1%. At a spot price of $18, this equals $0.005 per ounce. However, there was a time in early 2002 when the 2 month silver lease rate exceeded 15%. This would result in a lease cost of $0.68. Hopefully there is more capacity to take big tonnages and frequent shipments when you are dealing with ships as opposed to planes, so you may not have to hold on to your silver buybacks for too long.

Therefore, the smaller the local market for gold and silver, the fewer flights/ships, the further away from London you are the more exposed you are to a big discount in the selling price occurring during abnormal market situations. This explains why the big ETFs hold their metal in London. It lowers the risk as that is the biggest physical spot market. I trust this also gives an insight into the sort of issues and factors that a dealer has to work with when setting their trading prices.


  1. Great way to explain physical gold trading. I used to do it. And yes, the costs of moving gold from one place to another is high. Whenever I see an armored truck I feel I paid for it. No kidding. If you are in this business for a while and trade metals on a daily basis, you end up paying for a fleet of trucks in no time. But still make money

  2. excellent articles here, very informative! Thank you so much for your work