Gold prices likely haven't hit a bottom, but we can already see the path to recovery. We expect gold prices to slip below $1,000 per ounce in 2016. Higher U.S. interest rates will weigh heavily on investor demand, while continuing deflation pulls costs lower.
We expect jewellery buyers to partly fill the vacuum left by shrinking investor demand in 2017. Prices won't recover until 2018 and 2019, when rising Chinese and Indian jewellery purchases fully close the gap left by investors. We forecast a nominal gold price of $1,300 per ounce by 2020, or roughly $1,160 per ounce in constant 2015 dollars. This compares with roughly $1,050 per ounce today.
How We Forecast Gold
The market tends to focus on interest-rate and inflation expectations to forecast gold prices, which is sensible at face value. For decades, investor demand has largely defined gold's ups and downs. Surging investor demand led to the most recent bull run for gold. When rate and inflation expectations shifted and investor demand declined, the gold price fell.
While we recognise the key role of rate and inflation expectations in guiding investment demand, we think the focus risks overlooking other sources of gold demand that will play an increasingly powerful role globally. That's why we base our long-term outlook largely on more traditional supply and demand considerations, with a focus on marginal cost.
Our approach relies on iteration, as some categories such as recycling supply are influenced by the prevailing gold price. Rate and inflation expectations are subsumed within our methodology. They do not define it. In the long run, gold prices and cash costs have largely moved in tandem, a strong historical relationship that supports our argument for using marginal cash costs to forecast gold prices.
To forecast demand, we use a sector-by-sector approach, since the demand drivers differ greatly among categories. For each sector, we construct a detailed build-up, identifying and forecasting specific factors that will drive future gold volumes. To forecast supply, we first estimate supply from scrap based on the historical relationship between recycling rates and the prevailing gold price, which requires an understanding of above-ground gold stocks.
We then take our total demand forecast and subtract our recycling forecast to determine mine production's contribution to supply. Based on our expectations for mine production, we reference the industry cost curve, which we shift according to currency and input cost expectations, to arrive at the price where supply and demand intersect and the market clears. If faced with an implied supply shortfall, we evaluate the economics of prospective mine projects to arrive at an incentive price.
