We're reducing our long-term gold price forecast to better reflect our view on the marginal cost of mine production, which we estimate at $1,100 per ounce in real terms. We believe this focus on marginal analysis, while not without its shortcomings, is the most appropriate method for analysing gold miners and the best means for separating firms that are favourably positioned from those threatened by higher costs or weakening production.
We admit that forecasting gold prices is a tricky proposition. Unlike other metals, gold is rarely consumed--almost all of the gold ever mined remains in use as jewellery or as bars in bank vaults. Therefore, the amount of gold theoretically available for sale dwarfs mine output--at roughly 30,000 metric tons, central bank holdings alone amount to 12 years of global mine supply. Investment considerations can seem to affect gold prices more than the traditional commodity pillars of supply, demand, and marginal cost. As such, the combination of a weak US dollar, low or negative real interest rates, and macroeconomic and geopolitical uncertainty are commonly cited reasons for rising gold prices as investors flock to the yellow metal as a safe haven investment.
Despite clear deficiencies, we believe that a study of marginal costs, supply, and demand remains the best method for forecasting long-term gold prices. This methodology is especially useful for identifying low-cost miners and ones with favourable growth opportunities. While global mine supply has recently increased in response to higher gold prices, miners are still caught in a fierce battle against lower ore grades, higher capital costs and longer lead times, and a lack of meaningful exploration success. This has put pressure on both supply and costs. Meanwhile, demand for gold should be supported by exchange-traded funds (which make gold investing convenient for individuals) and the rise of the emerging market middle class as a source of jewellery demand.
Our long-term gold price assumption of $1,100 (in today's dollars) attempts to capture both the cash costs of production and the capital costs of mine development for the marginal miner. Breaking that sum down into its constituent parts, we estimate that the highest cost miners produced gold at a cash cost of $900-$1,000 per ounce in 2010. Amortisation, a decent proxy for capital costs, runs at about $100-$200 per ounce. While the industry supply curve is dynamic, we believe that, in the future, lower costs for royalties, consumables, and employee bonus schemes (all positively correlated with gold prices) will be offset by declining ore grades.
Previously, we used Comex gold futures contracts to populate the price decks in our discounted cash flow models, rather than making a call on gold prices. This resulted in frequent and sometimes material changes in our fair value estimates based on the market price for gold. Going forward, we will continue to use COMEX gold futures contract prices for the first three years of our explicit forecast horizon, which will help us anticipate near-term changes in our companies' cash flows. Beyond the first three years, we will revert to an inflation-adjusted long-term gold price forecast. In practice, this means we've cut our price forecast for 2014 to $1,200 from $1,515 and our forecast for 2015 to $1,236 from $1,579.
Our lower long-term gold price assumption results in lower fair value estimates, but the magnitude of adjustments will vary depending largely on companies' production costs and the timing and cost profile of new projects.
