Dust flux, Vostok ice core

Dust flux, Vostok ice core
Two dimensional phase space reconstruction of dust flux from the Vostok core over the period 186-4 ka using the time derivative method. Dust flux on the x-axis, rate of change is on the y-axis. From Gipp (2001).

Friday, March 29, 2019

How hard is it really to discover gold?

The question of peak gold has been on my mind lately.  I have looked at this problem in the past and  don't have a complete answer, although I am biased towards us being well short of peak gold..

Data from S&P Global Markets
Inflation adjustments calculated from here.

I would like to discuss this diagram. It seems reasonable to assume that an increase in exploration budgets should translate into gold discoveries. A glance at the above graph shows peaks in discoveries associated with the first two peaks in exploration expenses. The last peak in exploration expenses, around 2012, appears to be correlated with a spectacular lack of success on the exploration front, leading to the argument that we have reached (or passed) the time of peak gold.

The devil, however, may be in the details.

There are two confounding details that could change the interpretation of this diagram. The notes on the figure in the original diagram suggest that any new gold discovered on a project is attributed to the year in which the project was discovered. For instance, a junior miner might have discovered a 2 Moz deposit in 1994; and the project is developed over a number of years, probably changing hands in the process. In 2012, the major that now controls it embarks on an exploration program to expand the resource, and discovers a further 3 Moz. That 3 Moz discovery is attributed to 1994, as that is the year the project was discovered (see below).

This is the original version of the figure at the top of this post. The dark blue bars represent gold discovered in 2017, much of which is being attributed earlier years, as those years represent the time of the initial discovery of the projects on which this new gold is found. The small amount of gold attributed to the year 2017 represents new discoveries. If all gold found in 2017 had been attributed to 2017, it would look like a lot more gold was discovered that year--almost 150 Moz, as opposed to the 20-something Moz that appears on the graph instead. Now remember, that this is also true for 2016, and 2015, and 2014, and . . .

This is not an effect that can be easily removed. If it were, we would probably find that a lot of gold was indeed found in the past few years, and the discrepancy between the amount of gold discovered and the money spent on exploration might disappear.

So this graph cannot be used to make the argument that gold is much harder to find than it used to be. It may be used to make the argument that it is harder to find new gold deposits than it used to be, although unless you subtract out the money being spent by majors boosting the resources of existing projects, the graph is still misleading.

And there is another confounding detail that makes even this conclusion difficult to support.

The world of exploration has changed over the last two decades, particularly when it comes to resource definition. It was a lot easier to drill a few holes in 1990 and announce you had a 1 Moz deposit than it is now. It is also a lot easier to expand an existing resource than to define an entirely new one. For this reason, it seems to me that the economic incentives favour spending money on increasing existing resources rather than discovering new ones. At some point in the future this will probably no longer be the case, but it is difficult to predict when that will be; but at that point, there will be a lot more appetite for greenfields exploration again.

In conclusion, I don't consider this graph convincing evidence that gold is in increasingly short supply. It may be running low at a given price, but higher prices will liberate more. I still favour my earlier conclusion of up to 400 deposits in the 1-3 Moz category still to be discovered--and that doesn't include areas like the unexplored parts of Greenland, Antarctica, and the deep ocean.