May 082015

Last week I did a post on the London gold Auctions, noting at the end “the number of days where the volume offered for sale does not change, or changes very little, through each round”. I’d like to explore that in more detail in this post. As and example of what I was talking about, see the chart below.

London Auction 30th March

The red and green lines show the volume of ounces offered to buy (bid) or sell (ask) as each round progressed (note: the auction finishes if the volume bid and ask is within 20,000oz, hence the red and green lines don’t come together at the final successful round). The purple line shows the price set by the Chair of the auction for each round.

On this day you can see that amount offered for sale pretty much remained the same as the price dropped. This is counter to theoretical laws of supply and demand – as the price drops, you would expect quantity supplied to reduce. In the case of the buyers, we see what is expected – quantity demanded increasing.

To date we have had 63 London Auctions, so what has been the average behaviour over this time? First, each auction has had different quantities offered, so we convert the chart above into an index, setting the mid-point of the final bid and ask volumes to 100. That way all volumes can be compared with each other. This results in the chart below.

30th March Index

Then we can graph all the bid volumes and all the ask volumes, eliminating Auctions where there was only 1 or 2 rounds as that doesn’t show much information on how buyers and seller reacted to price. The chart below graphs all the ask/selling volumes with the big fat red line being my “average” behaviour (excuse the x-axis, 12 is the last round, 11 the 2nd last etc).

Auction Ask

You can see with the clustering of the lines that most of the 45 auctions showed very little reduction in the volume offered for sale, say a reduction of only 10-20% in general. Contrast that to the buyers, below.

Auction Bid

Here we see a lot more movement, with the buyers on average starting off a lot lower in the bid volumes (between 30-50% below) compared to what the sellers do.

Of the 45 London Gold Auctions of 3 or more rounds, 38 have seen reductions in the 1st round price to the final round price. Of those 38, the average starting volume offered by sellers as an index to the final mid-point volume was 114.2, a move down of 14.2. For the buyers, the average was 52.5 a move up of 47.5.

So it seems that on the majority of days the sellers are generally price insensitive, putting in “sell at market” type orders. My guess is that it is mostly mining companies doing this selling, as for many of their managers it is safer to just sell on a public benchmark than try and “trade” your production in the spot market – if you do better than the London Auction you probably don’t get as much praise to offset the angst you get from your boss when you achieved a price below the London Auction.

You’ll note the big jump in volume from the 2nd last round (average of about 30% under the final mid-point volume) to the last successful round (where it is about 10% off the final volume). It is as if the canny buyers have observed this “sell at market” behaviour and are holding out as long as possible with the volume they offer to try and get the price down (or maybe I’m just reading too much in here).

If any of you are investors in gold mining companies (commiserations) you may want to ask them if they sell at market on the London Auctions. If the answer is yes, you may then want to ask them if they know who the sucker is at the London Auction.