Accurate forecasts for medium-term commodity prices are essential for resource companies committing to large capital expenditures. The inaccuracy of conventional forecasting methods is well known because they tend to be extrapolations of the current price trend. The inevitable reversal catches many by surprise.
This paper demonstrates that medium-term (2–5 years) commodity prices are not strongly linked to economic health and commodity demand-supply, but are instead inversely controlled by supply-demand for the United States dollar (USD) and consequent valuation. P90, P50 and P10 projection bounds for future valuation of the USD are presented based on the successful probabilistic techniques of the petroleum exploration industry.
This allows probabilistic projections for the oil price, which is inversely related to the USD valuation. I show that the USD is significantly undervalued at present. Probabilistic projection of the USD valuation indicates that likely appreciation will put downward pressure on commodity prices for the next 2–5 years. If the USD premise is correct, likely appreciation of the dollar during the next 2–5 years will hold stable, or even decrease, oil price to around USD $50 BBL. This is a contrary expectation to most forecasts—one which, if it eventuates, should give cause for reflection before committing to large capital expenditures.
Further investigation could examine the extent to which the USD valuation can be modelled as a fractal phenomenon. If so, it would mean the USD valuation is not driven by conventional economic fundamentals; instead, it is a semi-random number series with serial correlation. If true, probabilistic forecasts of the USD can be significantly improved, hence that of medium-term commodity prices.