Investment Portfolio
The chart depicts an efficient investment frontier based on the returns from these three price series and using the standard Markowitz portfolio construction approach.

The ideal allocation of funds across the three assets would have depended on the objective being pursued.

• If the aim was to get the highest possible returns irrespective of their volatility (Point A on the efficient frontier), the portfolio would have been composed entirely of resource sector stocks since that was the asset class with the highest absolute return.

• If the aim was to minimise the volatility of the portfolio (Point B), funds would have been split 54/46 between industrial stocks and gold bullion.

• If the aim was to maximise the risk adjusted return from the portfolio (Point C), funds would have been allocated across industrials and gold in a 78/22 split (i.e. more industrials and less gold than in the second case) without any resources exposure.

Needless to say, these conclusions are based on historical data which might not be any guide to what happens in the future. For the past two years, for example, we have been observing a radically different investment environment for banks. This alone suggests that the industrial sector returns evident in the 1990s and early 2000s will no longer be typical as investors apply a higher discount rate to bank earnings and dividends than they might have done in the past.

The volatility in resource sector returns had typically been far higher in the 1980s and 1990s than for other sectors of the Australian market and for the market as a whole. Since then, the volatility in resources sector returns has fallen and converged significantly with returns from other market sectors.

To the extent that this change holds, the market should now pay more for any given level of resource sector earnings than it did in the past relative to the same non-resources earnings stream.

In any event, the role performed by gold in this example could have been carried out by other commodities. The correlation between plywood returns, to hark back to our earlier seemingly fanciful example, and the returns from the other two assets was lower than the correlations between gold returns and those of the other investments suggesting that, from a strictly analytical perspective, plywood could do the job just as well as gold. Other commodities might also be more than adequate substitutes.

Valuing Gold Production
The equities of gold producing companies are often used by people seeking the potential benefits of investing in gold but wanting the relative ease of buying and selling equities.

The market value of a gold producing asset will depend on:

• the gold price;

• the quantity of gold mined;

• the cost of extracting the gold; and,

• the discount rate the market wants to apply to the future earnings stream from a gold mine.

Higher gold prices directly affect revenues. However, gold prices also affect revenues and costs indirectly through their effect on changes to the quantity of gold mined.

For any given price, miners will always strive to maximize production so as to minimize their unit costs. However, the higher the gold price, the greater the financial incentive to expand production more aggressively. This can be done in three ways:

• by tapping lower grade parts of an existing orebody when the higher prices are sufficient to more than compensate for higher unit production costs due to lower grade ores;

• by investing in additional processing plant to raise production capacity; and,

• by raising exploration activity as a prelude to greenfield site developments.

The first route will usually be the quickest way to expand output. The other two will take longer partly due to technical and logistical issues but also because potential producers and their financiers have to gain enough confidence in the longevity of any price rise before making the necessary financial commitments.

Australian Prices and Output
The chart shows the historical relationship between Australian dollar gold prices and the quantity of gold mined in Australia.

Australian gold output had been declining slowly for nearly 20 years before it began a very rapid ten year rise in the 1980s. New techniques permitting the industry to exploit known areas of mineralization more effectively supported higher volumes but the gold price had risen fivefold in five years and, importantly, was sustained at the higher level giving the industry greater confidence about prospective investment returns. Nonetheless, there was a lag between the rise in gold price and the subsequent increase in output of approximately five years.

Having risen, there was little net change in the Australian dollar gold price between 1980 and 2005. Output rose briefly in the later 1990s but by 2007 industry output was little different to what it had been at the beginning of the 1990s following the initial growth surge.

Only since 2006 have gold prices gone past previously experienced levels. So far, the gold industry has failed to respond with significant increases in output although the Commonwealth government’s commodity forecasting arm, the Australian Bureau of Agricultural and Resource Economics (ABARE), is forecasting the first gold production increase for the Australian industry since 2003 in 2010 signalling the possibility of a sectoral revival. Nonetheless, it would be unwise to expect very rapid industry growth in the short term.

• The mining industry as a whole has had other priorities with skilled people and capital being drawn to coal, iron ore and base metal projects in response to growing demand for these commodities from Asia.

• Gold mines are frequently based around deposits containing as little as two or three parts per million of gold metal. The most easily mined gold has already been taken with progressively less accessible and more costly sources of ore now having to be tapped.

• Given the likely sources of new metal, a response lag at least as long as the lag between price rises and output growth in the 1980s would be consistent with normal industry decision making.

Equity Market Risk
Buying gold assets or listed gold companies incurs an equity market risk which is not present in a purchase of gold bullion.

In valuing a future income stream, the equity market will implicitly apply a discount rate. Currently, the market is more risk averse than it has been for many years and is applying a far higher discount rate, paying correspondingly less for earnings across all market sectors.

In Australia, the stock market fell by 48.9% between the market peak in 2007 and the end of January 2009. Despite the US dollar gold price having risen by over 20% during this same time frame and the Australian dollar falling by 30%, the S&P/ASX All Ordinaries gold mining index fell 26.7%.

The equity market effects, from which investors have been attempting to flee, have overwhelmed the impact on equity values of the positive gold price effects.

Not only do gold equities incur general equity market risks but they also have to contend with the risks of individual mines falling short of their potential for operational reasons unconnected with their commercial environments.

This throws some doubt on the wisdom of using gold equities as a bullion substitute and is consistent with the research conclusion of Gary Gorton and K. Geert Rouwenhorst (G&R) in “Facts and Fantasies about Commodity Futures.

G&R found that the share price performance of commodity producers is more likely to adopt the characteristics of the relevant equity market than the qualities of the markets of the underlying commodities which they produce. They concluded that “an investment in commodity company stocks has not been a close substitute for an investment in commodity futures”.

While gold equities have been dragged down by general equity market conditions, they have fared better than equities generally. Between the end of 2006 and January 2009, the gold price increased 45.4% while gold equities declined by 2.2%, a 47.6 percentage point difference in favour of gold bullion. Over a shorter period in the current cycle, since the peak in equity markets in October 2007, the gold price has risen 16.5% while gold related equities have fallen 26.7%, a similar performance differential.

However, between the market peak in 2007 and the end of January 2009, industrial company stock prices dropped in value by 49.8%. Quitting industrial equities at their October 2007 values and switching to gold equities would have delivered a 46.0% better outcome than holding industrial equities through the cycle.

Relative Returns
To the extent an investor is concerned with relative returns, using gold equities as a hedge against general equity market weakness would have been a beneficial alternative in the current cycle.

Over a more lengthy period of time covering less volatile market conditions, the data support the idea that gold equities have offered a leveraged exposure to movements in gold prices (i.e. gold related equities have risen or fallen more than gold bullion prices).

The chart shows monthly movements in gold prices on the horizontal axis and relative price movements between Australian gold equities and Australian industrial equities on the vertical axis. The chart is based on data from January 1992 to December 2008. An analysis of this data suggests that, on average, a 10% rise (or fall) in the price of gold bullion has been associated with a 16.4% gain (or loss) by gold equities over and above any movement in industrial equities.
Of course, these are generalizations which might not hold for the equities of individual gold producers or different combinations of gold producers.

The averages, themselves, can also conceal some variability with the potential to impact investment decisions.

• Over the period covered by the data, the direction of monthly gold bullion returns has coincided with the direction of gold equity returns more often than not, as one would expect. However, 31% of the time the direction of the two returns has not coincided.

• In those instances in which the direction of returns has coincided, equity returns were less than the bullion returns 44% of the time.

Conclusions

• The analytical basis for using gold in a world of expanding access to a fast multiplying range of investment products is less compelling than it once was.

• Gold prices have no more fundamental analytical base than any other currency. We are relying on a continuing act of faith rather than anything more substantial in expecting gold prices to rise.

• The low correlation between gold returns and those of other investments can assist portfolio construction.

• Using gold equities as a substitute for gold bullion, on average, provides some additional leverage to gold price movements.

• The probability of the expected connection between bullion prices and equity prices failing to hold represents a large ongoing risk to the use of gold equities as a bullion surrogate.