Energy Price Volatility and Forecast Uncertainty
October 6, 2009
Summary
It is often noted that energy prices are quite volatile, reflecting market participants’ adjustments to new information from physical energy markets and/or markets in energyrelated financial derivatives. Price volatility is an indication of the level of uncertainty, or risk, in the market. This paper describes how markets price risk and how the marketclearing process for risk transfer can be used to generate “price bands” around observed futures prices for crude oil, natural gas, and other commodities. These bands provide a quantitative measure of uncertainty regarding the range in which markets expect prices to trade.
The Energy Information Administration’s (EIA) monthly Short-Term Energy Outlook (STEO) publishes “base case” projections for a variety of energy prices that go out 12 to 24 months (every January the STEO forecast is extended through December of the following year). EIA has recognized that all price forecasts are highly uncertain and has described the uncertainty by identifying the market factors that may significantly move prices away from their expected paths, such as economic growth, Organization of Petroleum Exporting Countries (OPEC) behavior, geo-political events, and hurricanes. However, these descriptions do not provide a quantitative measure of the range of uncertainty regarding an expected future price. Nor do they indicate whether the uncertainty has increased or decreased since the last forecast was published.
Beginning with the October 2009 issue, the STEO will publish confidence intervals for crude oil and natural gas futures prices. A confidence interval is a range of prices between a low and a high price, i.e., the confidence limits. The range of the confidence interval is determined by the confidence level. The confidence level represents the probability that the final market price for a particular futures contract, e.g., December 2010 crude oil, will fall somewhere within the lower and upper range of prices. For example, if a confidence level of 95 percent is specified, then a range of prices can be estimated for any future month within which there is a 95-percent probability the price of the commodity in the expired contract’s delivery month will fall within that range. The higher the specified confidence level, the wider the range between the lower and upper confidence limits.