Gold prices: how to build a forecast without confusing scenarios with certainty

Forecasting the price of gold does not mean extending a recent curve or identifying a target as though it were an established outcome. The metal responds to several forces that sometimes move in opposite directions: interest rates, the dollar, investor demand, central bank purchases, jewelry consumption and the risk environment. A useful method therefore seeks to understand market drivers, build several scenarios and measure the uncertainty surrounding each.

This distinction matters for mining companies, investors and producing countries. A high price can improve export revenues or project value, but does not guarantee continued increases or identical profitability from one mine to another. Production costs, currencies, taxation and financing conditions change the final outcome.

Define what you are trying to forecast

Before choosing an indicator, the question must be clarified. Is the aim to estimate gold’s price in a week, assess an annual average or determine whether it will remain above a threshold relevant to a mining project? These horizons have different dominant factors. Short-term movements are often influenced by economic data, news and market positioning. Over longer periods, real interest rates, structural demand, monetary policy and central bank decisions carry more weight.

The price benchmark must also be chosen. Gold trades on several markets and is often priced in dollars per troy ounce. For an African company or state, the local-currency price may evolve differently depending on exchange rates. A forecast intended to inform public revenues or a mining company’s accounts must therefore specify the currency, the type of price observed and the costs against which it will be compared.

Track fundamental factors

The first factor is opportunity cost. Gold pays neither a coupon nor a dividend. When real bond yields rise, some investors may prefer those income-producing assets; when real yields fall, holding gold becomes relatively less costly. This relationship is useful but not mechanical: at some times, other buyers can offset the effect of interest rates.

The dollar is a second reference point. As gold’s international price is mainly quoted in dollars, a change in the US currency can alter the metal’s cost for buyers using other currencies and influence demand. Interest rates, currencies and investment flows should nevertheless be analyzed together, rather than assuming that a weaker dollar will always produce an identical rise in gold’s price.

The third factor is demand. Jewelry purchases, gold-linked financial products and net central bank purchases do not necessarily have the same motivations. Central banks may seek reserve diversification; households respond more to incomes, cultural habits and local prices; investors are sensitive to relative returns, liquidity and risk. A forecast combining these groups into a single “market” risks masking divergent behavior.

Finally, mine supply and recycling influence market balance, but slowly. A new mine requires capital, permits and several years of development. Existing production cannot always respond quickly to rising demand. In the short term, however, recycling and available stocks can ease some pressures.

Build three coherent scenarios

A simple framework establishes base, bullish and bearish scenarios. The base scenario rests on central assumptions about real interest rates, the dollar, growth, official purchases and investment flows. The bullish scenario might combine falling real yields, increased safe-haven demand and sustained central bank purchases. The bearish scenario might assume stronger growth, a firm dollar, rising real yields and outflows from gold investment funds.

The value of this approach is not a claim to know the exact path in advance. Instead, it requires explaining which assumptions would move the market in each direction. It also identifies signals that would invalidate the main scenario. Rising real interest rates accompanied by persistent buying flows, for example, would indicate that the usual drivers are insufficient to explain the price and that central banks’ contribution or other factors must be reassessed.

Statistical models can complement this assessment by linking the price to variables such as real interest rates, the dollar or investment flows. Their complexity nevertheless does not replace reliable data or consideration of regime changes: historical relationships can evolve with purchasing habits and monetary policies.

Test a forecast’s quality

A serious forecast must be evaluated using data that were not used to construct it. Comparing results with the simple assumption of an unchanged price is an initial test: if the model does not outperform this benchmark across several horizons, sophistication adds little. Average error must also be distinguished from the ability to predict large changes, because a method may follow the general trend but miss periods of stress.

Forecast intervals are more informative than a single target price. Stating that a price could fall within a range under a set of assumptions makes uncertainty visible. This range must be revised when data change, and past forecasts must remain recorded to track errors. Without this history, it is easy to remember accurate forecasts and forget failed ones.

For mining companies, the metal forecast must ultimately be translated into operating indicators: production costs, break-even thresholds, debt, investment needs, taxation and exchange rates. A high-cost project can remain vulnerable even when the price appears favorable. For producing states, budget prudence means avoiding permanent expenditure based on an exceptional price that could reverse.

A disciplined method is better than a spectacular figure

Gold market drivers extend beyond inflation or crises. The World Gold Council groups factors into four categories: economic expansion, risk and uncertainty, opportunity cost and market momentum. This framework helps examine competing forces. Growth can support jewelry and equipment demand, while uncertainty sometimes strengthens investment buying; interest rates and the dollar simultaneously change holding costs, and momentum can amplify short-term movements.

When analyzing the market, the sequence therefore matters more than the formula: define the horizon and currency, track fundamentals, establish several scenarios, test models over distinct periods and update assumptions when facts change. The result is not a price promise but a more transparent decision-making basis for comparing gold-related risks, projects and budgets.