Is Gold In A Bubble?
As the gold-sceptics keep reminding us, gold pays no coupon and no dividend, it does not offer a running yield, so traditional measures of ‘fair value’ do not apply.
But gold is money, and just as the paper ticket in your wallet does not pay interest, neither does gold. Gold is a monetary asset that has functioned as a medium of exchange and a store of value for thousands of years, around the world and in almost all societies and cultures.
Many modern economists believe that gold has now been successfully replaced with state paper money, such as paper dollars, paper euros, paper yen, and so forth. Holding gold is therefore redundant. The present crisis is a stark reminder that this faith in fiat money is misplaced.
Gold is still a superior monetary asset. It is not under the control of any political institution. It cannot be printed to artificially lower interest rates and to ‘stimulate’ the economy, to create fake booms in financial assets and in real estate, to fund credit growth with printed money rather than true savings, to subsidize the banking sector and then bail it out when the banks overreached, to allow the government to run never-ending budget deficits, to make unfunded promises to voters and fund wars. Gold is hard, inelastic, apolitical and truly international money. It does not bow to anybody. Paper money is a political tool.
Now that humankind’s latest and most ambitious experiment with fiat money, launched in 1971, has once again created massive imbalances and led the world into financial crisis, gold is experiencing a renaissance. But after the spectacular rally in the gold price over recent years many observers ask if the precious metal has not moved ahead of its fundamentals – even though it is not that easy to assess these fundamentals. They ask if gold has not entered speculative bubble territory.
Even fiat money sceptics such as Bill Bonner and Marc Faber, who have long been fans of the precious metal, warn that this is not a one-way street. From its all-time high of $1,900 in September of last year, gold has already corrected by about 20 percent. The current price action in gold is certainly disappointing given the ongoing and even intensifying Euro Zone debt crisis, which should encourage additional shifts out of fiat money and into gold. This week The Wall Street Journal concluded that “Gold’s status as a safe haven is looking shaky.”
The question is, therefore, if and how we can assess gold’s fundamental value, and by fundamental value I mean its role as money and a store of value, not its role as an industrial commodity. Are there any quantitative tools to determine if gold is in a bubble?
In this essay I will try and provide some back-of-the-envelope calculations that can give us some guidance as to the fundamental drivers of the gold price and the relative weight of these drivers through time. This is not meant to be a definitive analysis of the gold price and I certainly do not intend to give investment advice (my usual disclaimer!) or make any predictions as to where the gold price is heading. What I am doing here is a simple mental exercise that may be useful as a framework for how to think about the dollar-price of gold.
Over long periods of time gold has done an extraordinary job at preserving purchasing power. Not surprisingly, it has beaten all fiat monies as a store of value.
Since the dollar came off the gold standard domestically in 1933, and in particular since it came off the international gold-standard-light (Bretton Woods) in 1971, massive quantities of new dollar currency units have been created, thus substantially reducing the dollar’s purchasing power. The question I asked myself was this: at what price would an ounce of gold have had to trade in any given year to exactly compensate its owner for the loss in the dollar’s purchasing power since 1933, the year the dollar became an irredeemable piece of paper?
In 1933, one ounce of gold was fixed at $35. By using the US Inflation Calculator we can determine, based on official US inflation statistics, how many dollars it would have taken in any subsequent year to buy the same quantity of goods and services that $35 dollars bought in 1933. For example, in 1970 it would have taken $104 to buy the same goods that $35 bought in 1933. In 1996, it would have taken $422. Therefore, if the ounce price for gold had been $104 in 1970 and $466 in 1996, gold would have exactly compensated its owner for the loss of purchasing power that the inflating paper dollar experienced.
Given gold’s historic role as money and its superiority as a store of value over longer periods of time, it is not unreasonable for the owner of gold to expect that – on trend and in the long run – he should be compensated for inflation. I call this synthetic price the purchasing-power-protection price of gold, or the PPP-price of gold, which can be thought of as some long-run lower limit of the actual gold-price.
Of course, it is unlikely that gold would trade precisely at that price. At any moment in time, the price of gold will reflect many other factors, too. There are, first and foremost, expectations as to future inflation. Then there are potential concerns about the stability of the banking system, or geopolitical considerations.
Additionally, a lot of gold has been mined since 1933, and in particular since 1950. Furthermore, we can debate whether official CPI statistics are really a good measure of the dollar’s declining purchasing power. As Mises has explained, there is no such thing as a clearly identifiable and measurable purchasing power of money. Every index that is being used is a compromise, and we know that the US government has repeatedly changed its methodology of how to calculate the debasement of its own fiat money.
it would be reasonable to assume that the market has a superior way of assessing the dollar’s loss in purchasing power and that this would then be the true driver of the gold price, rather than the government’s official measure. Then there is the question if $35 is the right staring point. Before all privately held gold was confiscated by Roosevelt’s executive order in April 1933, the dollar was fixed at $20 dollars an ounce. $35 dollars was simply a new, administratively set price.
I decided to take the $35 dollar price simply because at that stage (1933) so many paper dollars had already been printed – by the banks and with the encouragement, since 1913, from the Federal Reserve as the government’s backstop for Wall Street – that the $20 dollar price was no longer reflecting the true price of gold: Americans were ditching bank deposits and paper dollars and accumulating gold thus pulling the rug from under the banking system. This was the reason for Roosevelt’s intervention.
In any case, there are many reasons why we should take this analysis with a pinch of salt. But nevertheless, I do not think that this approach is entirely without merit as long as we understand that it is simply a rough framework and do not read too much into it.
I calculated the PPP-price of gold for the 43 years from 1970 to 2012 and compared it to the average market price for gold in every year. The time series for the average gold price was provided by my friend David Goldstone, who obtained it from http://www.measuringworth.com/datasets/gold/result.php.
It was a discussion about the gold price with David that gave me the idea for this analysis.
I also calculated a simple ratio between the two prices, the market price and the PPP price. When this ratio is below one, the gold price is below the PPP-price and gold at this level would not even compensate for the massive dollar debasement since 1933. If the ratio is above one, the gold price compensates fully for the dollar’s loss in purchasing power and also provides an additional margin on top.
As one would expect, both PPP-price and gold market price increased substantially over those forty-three years. The average ratio was 1.16, which means that gold traded on average at a 16 percent premium to its PPP price. This makes sense. We would expect gold to compensate – most of the time – not just for present inflation but also for the extra risk of accelerating inflation in the future, or to also compensate for other risks. It should neither be surprising that the volatility of the market price around PPP was substantial. Only in about half the years in our study was gold trading even within 25 percent on either side of its PPP price.
Using the PPP-measure, gold was most undervalued in the early 70s when it traded at a more than 60 percent discount to PPP, and between 1999 and 2001, when it traded at a 40 percent discount. (I will discuss the reasons for this in the second part of this essay).
Gold seemed most overvalued in 1980, when it traded at more than twice its PPP price, and from 2010 until today, when its price is again more than twice the level that would be justified on up-to-date paper dollar inflation alone. It seems obvious that in both these instances the gold price reflected concerns about an imminent further acceleration in inflation or even, I would argue, imminent monetary regime change. Before we look at these instances a bit closer let’s try and develop a narrative for our time series.