This article is a short version of the paper at the bottom of the page!
Most gold market research is based on the premise that the supply side of the market can be characterized by the quantity supplied and demand side by the quantity demanded. The specific cause and effect relationship between these two variables and price is often unstated; and perhaps rightfully so: is it not obvious that a greater quantity demanded is the cause of a higher price, and that a greater quantity supplied is responsible for a lower price? No!
This article will show that market forecasts based on quantities of gold are meaningless. Widespread statements like “Gold demand was up by 15% in 2012” are true but only if they are understood in a misleading sense. The supply and demand sides of the market consist of supply and demand schedules, not quantities. A price forecast based on quantities is a non sequitur because there is no causal connection from the quantities to the price. This error has side-tracked the majority of analysts into an obsessive focus on quantities while ignoring the actual drivers of the price.
This article does not entirely stand alone; it builds upon other articles that I have written about the gold market, and on the marginal price theory of the Austrian School. Some parts of this article will not make sense unless you are familiar with some of these concepts. I chose to do this partly to avoid repeating ideas that I have already published, and partly to control the length. I have linked to backup material that I believe is relevant. This article contains the key insights and principles of the paper at the bottom of the page. The paper elaborates each section in some more detail and has a detailed analysis on the CPM Group’s Model.
The Usual Explanation
First, let’s look at the examples. Most published analysis of the gold market is concerned with supply and demand numbers.
As the gold price increases, demand for gold and other precious metals has continued to grow. Demand for gold has continued to grow in 2012 and is predicted to increase further next year. Research by Source, a provider of exchange traded products, shows that inflows into European gold ETPs have reached $6.8bn this year to date, constituting a staggering 15.4pc growth
The World Gold Council’s web site contains the following:
Since 2003, investment has represented the strongest source of growth in demand. The last five years to the end of 2011 saw an increase in value terms of around 534%. In 2011 alone, investment attracted net inflows of approximately US$82.9bn.
My third example cites CPM Group’s 2012 Gold Yearbook Press Release:
Investment demand, the key driver for gold prices, remained at historically high levels last year. Net additions to private investor gold holdings declined to 34.3 million ounces in 2011, down 5.8% from 2010 levels. Even though net additions to private investor holdings slipped lower in 2011, a year in which prices touched a record high, the decline had followed two years of double-digit growth from already high levels of net additions to investor holdings. (p2)
Gold fabrication demand rose 0.6% to 72.9 million ounces in 2011, slower than the 2.3% growth in 2010 due to higher gold prices. Despite higher prices, many consumers sought to purchase more gold jewelry, specifically in developing countries, as a hedge against inflation and form of savings. Developing countries’ demand for gold in the form of jewelry rose to 50.2 million ounces, up from 49.6 million ounces. (p3)
Market Sectors and Flows
Most gold analysts divide the market into sectors. This section will discuss how this is done and what the quantities mean in relation to the sectors. A typical sector breakdown is: mines, industry, jewelry, investors, and the official sector (central banks). Some writers break the investment sector down into bars, coins, and ETFs. The choice of sectors is not critical to the points that follow; none of the conclusions would change if, instead of these sectors, flows between countries were used instead. Some reports combine the two approaches, dividing the developed world market into sectors and treating the rest of the world on a country or regional basis. Any of these breakdowns would serve equally well.
Below is a list of the sectors, their buying, and their selling:
Inter-sector Flows and Quantity Balance
The quantity balance between sectors is at the core of most market analysis. The quantity balance is an equation relating all of the flows in the market to each other. (A flow is the quantity bought and sold, while a stock is a quantity held by someone over time). Quantity balance is the requirement that every movement of gold must be accounted for on the buy side and the sell side. It is similar to the way that double-entry bookkeeping works. This section will derive the quantity balance equation. The following section will discuss its significance.
Over a one-year period, every trade that takes place between a buyer and a seller is counted in the following way: the quantity of gold bought (and sold) is added the buying sector’s gross quantity bought and to the selling sector’s gross quantity sold.
At the end of the year, net flows for each sector are calculated. The definition of the net flow for a single sector is:
sector net flow = sector total buying – sector total selling
Sector net flow can be a positive number, meaning that the members of the sector bought more than it sold; or a negative number, indicating that the members of that sector sold, in aggregate, a greater quantity of gold than they bought.
Assuming that mines sell all of their production, which is nearly always true, mine sector net flow is always a negative number.
mine net flow = mine buying – mine selling = 0 – mine selling = – quantity mined
For every trade, the quantity bought is equal to the quantity sold. This means that the sum of all sector net flows is zero. By the rules of algebra, this arithmetic identity can be rearranged in several ways:
(1) quantity mined + net industry + net jewelry + net investor + net official = 0
(2) net industry + net jewelry + net investor + net official = quantity mined
The CPM Group uses a different sector breakdown than I have used here, so their quantity balance is a little bit different. They use the following market sectors: total supply (mine plus scrap), fabrication demand (industry plus jewelry), official sector and investment. Their quantity balance in their partitioning is summarized in equation (3), below.
(3) quantity mined + industry sold + jewelry sold
= industry bought + jewelry bought + net official + net investor
In this breakdown, official and investor sectors have a net flow on the right side of the equation but the jewelry and industry sectors have gross purchases on the left of the equation and gross sales on the right.
The preceding equations are all saying the same thing: all the gold that comes out of mines ends up as net inflow into one or more market sectors. These identities all follow directly from the laws of arithmetic. They contain no new information. They are only a restatement of the original assumptions, namely, that miners sell all of their production, and that no gold is destroyed during a trade. The mine sector net flow is always negative but the other sector net flows could be positive, negative or zero.
The False Logic of Quantities
I believe that the error of attributing gold price moves to quantities is based on the following invalid thought process on the demand side (with similar thoughts on the supply side not shown here):
- The gold price is driven by supply and demand
- Supply and demand are quantities
- Looking at the demand side, more demand implies a higher price, less demand a lower price.
- More supply means a lower price, less supply a higher price.
- The key to forecasting the gold price is therefore measurement of gold supply and demand.
The main problem with this view, as I will show in the next section, is that there is no cause and effect relationship between the quantities and price.
Flows not the Cause of Price
The financial media commonly reports that buying is the cause of the price going up. Stories in the financial media usually report only one side or the other side of the market. For example, an increasing number of small investors buying coins is often cited as the cause of gold price strength. However, the same story could equally well have been written as a bearish report about the increasing number of investors willing to sell their coins. Either story would be true, at least from a quantitative standpoint and both would be wrong in attributing the movement in the gold price to one side of the market only.
If the reporter accurately described a large volume of coin buying and an equal volume of coin selling, then what conclusion about the price should the reporter draw? Exactly none. Buying as such is not the cause of the higher gold price, nor is selling the cause of price declines. If buying could take place without selling or selling without buying, then one or the other could be an independent cause of price moves. But neither can occur without the other. Buying and selling occur always in equal quantities, and, at the same time. For every purchase of gold by a buyer, an equal quantity is sold by the seller. The quantity of buying, which is always the same as the quantity of selling, is not the cause of the gold price.
While everyone agrees that the gold price is driven by supply and demand, not everyone who voices agreement means the same thing. The correct version is: the gold price is driven by supply schedules and demand schedules. Most analysis of the gold market is based on an incorrect interpretation of the statement, namely, the gold price is driven by the quantity supplied and the quantity demanded. An increase in gold demand is the cause of a higher price if an increase in demand, means a change in the preference rankings of coin buyers for more gold/less cash. In that case, all other things equal, transactions would occur at a higher price.
The quantity balance equations are logically valid at all times, but they are accounting identities, not statements of cause and effect. The quantity bought and sold is not an explanation of why the price moved. All inter-sector flows must balance, but flow is not the cause of the price; it is a summary quantity of gold traded, at whatever price. Any combination of positive, negative, or net inflows or outflows into any one or more sectors could occur during a year where the gold price was higher, lower, or unchanged.
Suppose during the last year that net investor inflow is a positive number and net official inflow is negative. This indicates that over one year, investors purchased gold from central banks. But this fact is an arithmetic identity, not a cause of the gold price movements during this year. If, the following year, central banks on net purchased gold from investors, we are still no closer to knowing at what price the gold was purchased, and whether that price is higher or lower than the current price.
The True Cause of the Gold Price: Marginal Preferences
Each investor strives to maintain their desired holdings of all potential assets, including cash (i.e. one or more national currencies such as the US dollar or euro). As their preferences change, and as market prices change, investors adjust their portfolio holdings, at the margin, to bring them in line with their preferences. The price of an asset emerges as investors balance, bid for assets they wish to hold more off and offer assets they prefer to hold less of.
Supply and demand as they contribute to the price must be understood not as quantities but as schedules. Market prices balance the aggregated supply and demand schedules of the entire market. These aggregated schedules are also known as the more widely used supply and demand curves. In the standard micro-economic presentation, the supply and demand curves intersect at a point, marking the price and the quantity.
I have written about the application of supply and demand schedules to the gold market in Does Gold Mining Matter? There I explain that the supply schedule for gold (in dollar terms) is dominated by the owners of the world’s existing stockpile of gold, and that mined gold during any one year period has a relatively small impact on the supply schedule. The price is set primarily by the reservation demand schedules of the owners of the existing gold. In the same piece,I show that the quantity mined, which many analysts incorrectly believe is “the supply”, has little influence on the gold price.
The quantity balance constraint cannot be a cause of the gold price because balance equations contain only quantities. The gold price is the quantity of money exchanged for the quantity of gold. Any explanation of the gold price must contain some reference to the quantity of money involved. Equilibrium price theory provides a complete theory of the cause of the gold price, taking into account the gold and money sides of the market.
If the gold price is higher now than it was at some point in the past, that can only be due to a shift in preference schedules. One of the following must be true: 1) either buyers valued the gold more highly and thus were willing to pay higher price, or 2) sellers valued their gold more highly and were only willing to part with it at a higher price. Historical net flows provide a summary of where in the market were the buyers who valued gold the most highly, and the sellers who valued it the least.
Up to this point I have argued that the quantities supplied and demanded are not the cause of the gold price. The true causal relationship between price and quantity is nearly in the opposite direction. Transactions occur in the market because there are some investors whose mix of cash and gold holdings is not consistent with their preferences. Trading will occur until everyone has adjusted their portfolios, at the margin, to their preferred holdings. If no one changed their preferences after this moment, and no new gold were mined, then no more trading would occur.
Trading continues because people are always changing their minds about what they want to own. Individuals who did not previously consider themselves gold investors enter the market; others no longer consider gold a good investment and sell out. The more individual investors that have changed their preference rankings since the last market price, the greater the disequilibrium in the market, and the more change in the ownership of gold and cash is necessary in order for investors to reach their desired holdings. The volume of trading reflects the extent that holdings of some individuals no longer reflect their preferences.
Demand Schedules Not Measurable
So far I have argued that the gold price is an outcome of the preference schedules of investors. A preference schedule is not a number. It is a spiky curve representing a range of quantities and prices. Schedules are not directly measurable in the way that quantities are, because they include hypothetical quantities that would be supplied and demanded at prices above and below the market. In order to have the complete supply and demand schedules, the analyst would have to know how much gold would be sold and purchased at every price. When gold trades, we know only that the quantity supplied and demanded at one price.
While I have spent most of this article discussing demand, the supply side of the market works the same way. Supply schedules and demand schedules together drive the gold price. Supply schedules are immeasurable as are demand schedules.
The main point that I have tried to show is that the demand numbers used in most gold market reports do not measure the demand side of the price formation process. The same could be said about the supply number. These two numbers are connected through the quantity balance constraint but they are not the cause of the gold price.
Gold market analysts have a tougher job than other financial analysts. In Value Investors Hate Gold, I argue that it is more difficult to analyze the yellow metal than equities because quantitative measures such as yield, cash flows, balance sheet leverage, and growth rates provide a fundamental basis for analysis. Gold has none of those things.
The fundamentals of gold are the current purchasing power of money; expectations about the future purchasing power of money; the growth rates of various national money supplies; the volume of bad debts in the system; expected growth rates of bad debts; the attractiveness of other available investments; and the investor’s preference for consumption rather than investment. These factors do not act directly on the gold price. Instead, they are focused through the prism of investor preferences, which are not measurable. The price is the ultimate measurement of how investors view these factors. The paradox of gold is that which drives the price cannot be measured, that which can be measured does not drive the price.