All precious metals are down, but only one is obviously cheap

21 Jun 2013 by Jim Fickett.

Taking into account both history and cost of production, silver is a little expensive. Gold is expensive on an historical view, but perhaps cheap if you think current monetary expansion will lead to high inflation. The palladium market is in the midst of a major adjustment, as Russian state sales decline. It is hard to tell whether the current price has already accounted for the necessary adjustments or not. Platinum is priced well below its cost of production, and is a bargain.

The recent fall in the prices of precious metals has been substantial, as can be seen from the price history for the last few years for each metal (from Kitco):

Despite the substantial downward adjustment, a longer view suggests prices could possibly have further to fall.

We can compare the current spot price, inflation adjusted, to the history of the same over the last four decades. Even after the recent drop, all four metals are well above their long-term average inflation-adjusted prices – silver by 25%, gold by 74%, palladium by 91%, and platinum by 40%:

20 Jun 2013 nominal 20 Jun 2013 real mean 1969-2013 real 20 Jun 2013/mean
Silver 19.60 8.41 6.72 1.25
Gold 1277.80 548.41 315.61 1.74
Palladium 662.00 284.12 148.73 1.91
Platinum 1357.00 582.40 414.85 1.40

Of course the world is not static, and the historical average price is a useful guide, but will not necessarily be the average price going forward. It is useful to look at the particulars of the market for each metal individually.


Silver experienced a genuine bubble over the last few years (Silver bubble). It is primarily an industrial metal, and is not in short supply, so it could be rationally priced if we knew the actual cost of production. Unfortunately about three quarters of all the silver produced comes as a byproduct of mining something else, so it is difficult to tease out the true production cost. The Silver Institute estimates the marginal cash cost of production was about $9/oz in 2013. The true total cost of production would add to this general and administrative expenses, royalties, tax, and capital expenditures, and would certainly be higher. Since it is hard to know how much higher, perhaps it is best to be guided by history, which shows that the current price is 25% above its long-term average. In other words, silver is back down to only being somewhat expensive.


Although there are important industrial uses for gold, it is primarily valued as money. That is, most transactions in the gold market are between investors, and both mining supply and industrial consumption are small in comparison. In my view, then, there are only two way to try to put a rational price on gold. First, you can look at the historical average. Even after the impressive drop we have just experienced, gold remains 74% above its historical average price. This suggests very considerable caution. Second, taking a purely monetary view and looking only at the world's dominant currency, you can compare the total number of dollars in existence with the total number of ounces of gold, and try to deduce an exchange rate (The gold-dollar exchange rate suggests gold is fairly valued). This would suggest that the current price is somewhat (though not drastically) low.

Both of these points of view have their strengths, but neither can be viewed as anything like precise. Overall, I would say that if the long-term unfunded liabilities of the US, together with the risks the Fed is taking, lead to high inflation eventually (which is quite likely but not guaranteed), the current price of gold will be seen in retrospect as attractive. But the price remains high enough that there is still no margin of safety in buying, even after the recent fall.

My bottom line views on investing in gold are summarized in Everything you need to know about gold in eight bullet points.


The biggest driver of the palladium price in the last few years has been supply from Russian government sales. For many years, supply from mining and recycling has been less than total demand, and the difference has been largely made up by sales from a large Russian stockpile built up during Soviet times. Recently Russian government sales have declined precipitously. From about a million ounces in each of 2008, 2009, and 2010, sales dropped to 775,000 in 2011 and 250,000 in 2012. The consensus in the industry is that Russian stocks are nearly depleted, and the market is about to be in chronic shortage. Hence, even with the recent drop, the price remains nearly double its long-term average.

There are several difficulties in trying to come up with a rational price for palladium. First, the state of the Russian stockpile is a secret. It is likely true that lower state sales indicate depletion of the stockpile, but no one outside of Putin's inner circle really knows if that is the case. Second, current investor holdings total around nine million ounces – enough to replace the old Russian state sales for a decade. So the fickleness of the market will be self-reinforcing, as is so often the case. If investors buy out of fear of a shortage, they will create a shortage and the price will rise. If the current fall leads to sales, there may be no shortage after all, leading to further falls. Third, no one really knows how the market will respond to lower supply from Russia. Higher palladium prices may lead to some switching from palladium to platinum in automobile catalytic converters, or jewellery demand (about half a million ounces annually) may drop. And finally, as with silver, most palladium is produced as a by-product of mining other metals, so it is difficult to estimate its cost of production.

All told, I think some premium to the long-term price is very likely justified, but it is hard to be quantitative about what the market price should be. The current 91% premium to the long-term price is substantial, and one should really take time to gain serious expertise before jumping in.


The platinum market is mostly industrial rather than speculative (unlike gold) and is not in the midst of any major adjustments (unlike palladium), so one can arrive at a rational price by looking at mining costs. The typical cost of production is at least $1500, and probably more like $1700 (Platinum mining costs are likely to drive the long-term price moderately higher). Further, costs are more likely to continue rising than to fall. So the current price, at $1357, is not just on the low side, it is low enough to have some margin of safety built in.

Platinum is currently a bargain.