more articles by

Doug Dillon

Click to enlarge Click to enlarge


There's Just Not Enough Gold; Modeling A Dollar Flight To Gold

By Doug Dillon      Printer Friendly Version
Jan 17 2008 2:47PM

A significant rise in inflation has been seen (and can be expected to continue) given the double-digit growth of the money supply now occurring in the US, the EU and other G-20 economies


A flight from the dollar to gold is now expected by many observers and is even being covered by mainstream newspapers


This article models, in a simple fashion, a dollar flight to gold’s impact on the price of gold.

This growth in the money supply has actually been taking place since the mid-1990s. The excess money has to go somewhere and since the mid-1990s it has found its way into the Internet stocks producing a bubble and, following that, into the recent housing market causing the housing / subprime-mortgage bubble. This article concludes with some thoughts about how gold differs from the Internet stock and Housing bubbles. 

Table A provides a basic perspective on how the gold mining industry compares to the global economy. As you can see, even at recent elevated gold prices, the value of gold being mined is only .2% of the world GDP. Gold mining is thus an extremely small component of the world-economy. Total existing gold demand is balanced with supply at current prices and is roughly 98 billion dollars annually. Most of gold mine production (74%) is used for fabrication (primarily jewelry). The remaining demand is classified as investing demand. Importantly, gold investment demand constitutes just 0.05% of global GDP. All of this points to the fact that even a very minor shift of the world’s investment into gold will have a huge impact on gold demand.

This article explores this impact by means of a model. The basic outline of the model is as follows:

  • In a flight to gold scenario, supply is assumed to be inelastic. That is, existing holders of gold will not want to part with their gold (because its price is rising and the dollar is unattractive). The other supply of gold is what comes from gold mines. The lack of production increases in the face of the rapidly rising price of gold over the last few years demonstrates that supply of gold is also inelastic. So, the model has the supply of gold being perfectly inelastic at the current rate of mine production. This is a bit of a simplification, but it still allows real insight into the effect of a flight from the dollar to gold.

  • Existing demand is considered separate from any dollar-flight demand and is assumed to remain constant in dollars. Thus the amount of oz of gold going to this existing fabrication and investment demand scales inversely with the price of gold.

  • Additional annualized dollar-flight demand is assumed (based on illustrative guesses by the author of what seems reasonably likely).

  • The modeled price of gold is then the total annual demand divided by the annual mine production.

Table B models the effect of a flight of petrodollars to gold. The oil producing countries listed in the table have already expressed, to a great or lesser degree, dissatisfaction with the use of the dollar as the reserve currency, either for political reasons or because of inflation resulting from the dollar’s recent devaluation. Table B models the effect of those countries investing a small fraction, 5%, of their oil revenue in gold rather than dollar denominated financial instruments. This results in an additional $54 billion of gold demand which pushes the modeled price of gold to $1401 / oz. This change of investment policy could continue for a number of years pushing the price of gold up indefinitely.

China has a huge trade surplus to the United States and is holding 1.2 trillion dollars of reserves. Table C models the effect of China, over the course of a year, shifting 5% of those reserves into gold. This increases gold demand by $60 billion pushing the modeled price of gold up to $1454 / oz. This change in investment policy might continue beyond a single year as the trade imbalance persists and as China continues to move additional fractions of its reserve from the dollar into gold.

Table D models the effect of a very small fraction (1%) of US household financial assets being moved from where they currently reside (equities, mutual funds, bonds, pension funds holding the same) into gold. US household financial assets are so large that even a 1% shift increases gold investment by $422 billion pushing our modeled price of gold up to $4769 / oz! This illustrates how sensitive the price of gold is to a change of investor sentiment.

Table E models the effect of a minor panic flight from the dollar. In this scenario all of the previous shifts take place at once and the amount shifted is doubled. Having multiple different simultaneous dollar flights to gold is not unreasonable. If the dollar becomes severely unattractive you would expect many holders to head for the exit door at the same time. The doubling of the previous four modeled flights seems to the author to underestimate the effect of even a minor panic. This scenario, as modeled, yields a modeled price of gold is then $10,771.

The graph below summarizes the modeled results.

The author does not consider any of the above modeled-gold prices to be forecasts. He does consider them to be very illustrative of how sensitive the price of gold is to even minor shifts of investment from the dollar into the gold.

There have recently been a spate of forecasts of upcoming gold price rises based on adjusting for inflation the 1980 peak price of gold. The author submits that these forecasts could be way too low if a serious flight from the dollar to gold takes place.

As a postscript, there remains one important thing to say about how a dollar flight to gold differs from the Internet stock bubble and the Housing bubble. The supply of gold is inelastic. Even though the price of gold has more than tripled in the last few years (from $250 / oz to around $900 / oz), gold mine output has stagnated.

The supply of worthless Internet startups was, as we found out, anything but inelastic. After the mania got going, Wall Street produced out of thing air a limitless supply of worthless Internet startups. Soon the bubble popped and the price of those worthless startups price reverted to their actual value, zero.

Similarly, homebuilders have found that the supply of new houses, while initially inelastic, was quickly made elastic by the flood of houses brought to market. Home prices are now returning to their actual value.

In the first really classic bubble, the Dutch discovered that the supply of tulip bulbs, while initially inelastic, after a couple of growing seasons was completely elastic. The bubble burst and the rest is history.

Because the supply of gold is inelastic, the rise of the price of gold does not constitute a bubble, at least not the same as the Tech bubble, Housing bubble and Tulip Mania. It is not susceptible to the same kind of wall of supply induced collapse.

The dollar is completely different from gold. Its supply, like the other fiat-currencies, is like worthless Internet startups: perfectly elastic. Just as venture capitalists could produce a limitless supply of Internet startups, central banks can produce a limitless supply of fiat currency. It is this very difference in elasticity between the dollar and gold that could trigger a flight (or panic) from the dollar to gold.

Doug Dillon



You can follow Doug Dillon's learning experience on his blog, World of WallStreet (, on which he frequently posts.