When it Comes to Iran, Energy Market Shows the World Believes U.S. is Bluffing
By David P. Goldman
By David P. Goldman
The Obama administration has utterly failed to convince the world that it is serious when it says the U.S. would attack Iran if it does not halt its nuclear weapons program. Analysis of energy markets reveals that crude oil is trading today with no greater risk than stocks and currency. Even China, notoriously conservative when it comes to the energy imports upon which its economy depends, demonstrates little concern over the prospect that Gulf oil flows might be interrupted by American military action against Iran.
Do oil prices reflect the risk of a military strike against Iran’s nuclear weapons program? This is not an academic question. If oil buyers ignore President Obama’s claim that he is not bluffing about Iran, it is likely that the Iranian government will ignore it as well. Measurement of a risk premium in the oil price indirectly gauges the credibility of the Obama administration’s stance towards Iran. The oil price has fallen by 16 percent during the past 12 months (from $107 to $90 for the benchmark West Texas Intermediate) as the Iranian nuclear threat has escalated. That suggests that the oil market does not assign a high probability that military force will be used against Iran.
Markets are not necessarily good forecasters of strategic events. Nonetheless, corporations and investors devote vast resources to hedging prospective oil price movements, and the market consensus is an important data point.
Last March, President Obama claimed that “uncertainty about what’s going on in Iran and the Middle East” was adding a $20 or $30 premium to oil prices.” Two weeks ago the Organization for Economic Cooperation and Development (OECD) claimed to measure a big risk premium in oil prices. But a strong case can be made that oil is trading on economic factors alone, with no strategic risk premium at all. My own firm, Macrostrategy LLC, has built an econometric model of oil prices that explains nearly 90 percent of weekly oil price movements during the past four years.
Major oil buyers continue to display a high degree of complacency about supplies. For example, China’s oil stockpiles have shrunk to the lowest level since March 2012, Bloomberg News reported Feb. 20: “Crude supplies, excluding emergency reserves, dropped by one percent from a month earlier, a report from Xinhua News Agency’s China Oil, Gas & Petrochemicals newsletter showed today. Inventories dropped to 28.86 million tons, which was the lowest since March, according to calculations by Bloomberg based on the Xinhua data.” Although Chinese strategic oil reserve data are secret, the International Energy Agency believes that China stopped adding to reserves last November. China’s strategic petroleum reserve aims to reach 500 million barrels, or 90 days’ of imports, but presently holds perhaps a quarter of that amount.
Economists at the Organization for Economic Cooperation and Development claimed in a working paper titled “The Price of Oil – Will it Start Rising Again?,” to have measured a large risk premium in the present price of oil equal to about a fifth of the oil price.(1) The OECD makes the large assumption that whatever their model fails to explain can be interpreted as a risk premium.(2) Using the OECD’s flawed method, it is possible to attribute virtually all the movement in the oil price during the past several years to economic expectations.
Macrostrategy LLC, conducted a different modeling exercise utilizing weekly data for the past four years that simulated the change in the West Texas Intermediate oil price based on three variables: stock prices (a proxy for growth expectations), the trade-weighted value of the dollar (a proxy for the cost of oil to other major oil importing countries), and the level of crude oil inventories in the United States (a proxy for supply and demand).
The Macrostrategy model estimates weekly changes in the oil price, and adds up the weekly changes to simulate the level of the oil price over the four-year interval. The factors are intuitive and their relationship to the oil price is evident. Each of these variables clearly tracked the oil price.(3) When combined in a rigorous econometric model, they produce an exceptionally good fit.(4)
Volatility is another gauge of the degree of perceived risk in the oil market. The traded oil price is now around 30-year lows, whether volatility is calculated using a rolling standard deviation of returns or with more sophisticated methods.(5) Oil price volatility is also very low because the volatility of economic variables in general is very low. The implied volatilities of the S&P 500 index, gold, oil, and the Euro exchange rate with the U.S. dollar have moved in lockstep during the past four years. The fact that oil price risk as gauged by the options corresponds quite closely with economic risk further confirms the conclusion of the Macrostrategy econometric model.
We conclude that oil is trading on economic not strategic risk. Recent variations in the oil price can be explained almost entirely by variables reflecting economic expectations with a rigorous econometric model. In light of this result, there is no reason to believe that there is a strategic risk premium in the oil market. Evidently, the world remains convinced that when it comes to the use of military force against Iran’s nuclear weapons development program, the United States is bluffing.
David P. Goldman, JINSA Fellow, writes the “Spengler” column for Asia Times Online and the “Spengler” blog at PJ Media. He is also a columnist at Tablet, and contributes frequently to numerous other publications. For more information on the JINSA Fellowship program, click here. For more information on the JINSA Fellowship program, click here.
(1) The OECD writes:
Starting from the earliest possible quarter given data availability constraints, namely the second quarter of 1992, the model is solved for each quarter of the simulation period using a linearization which assumes that the variation in oil demand/supply is small relative to its level. The dynamic simulation underestimates oil prices in the second quarter of 2012.It is assumed that half of this gap reflected a risk premium due to fears of more severe supply shocks in the future.
(2) In my view, the model is flawed. Like most conventional macroeconomic models, the OECD estimation uses backward-looking data that do not take into account the expectations of oil market participants. It takes quarterly data for oil production, which is known, and attempts to estimate demand for oil based on three factors: “the real price of oil, the level of national disposable income, and the preferences of the consumer.” In other words, oil demand falls when the oil price is relatively high, when national income falls, and when consumers feel like buying other things than oil. Consumer preferences are notoriously hard to gauge, and the OECD model piles assumption atop assumption in its estimation. That is not the model’s worst failing, though. The overriding error in this approach is that it does not take into account expectations about economic growth and inflation.
(3) The Granger Causality Test confirms that the predictor variables anticipate changes in the oil price rather than vice versa.
(4) To maintain econometric rigor, Macrostrategy employed a two-stage modeling process. In the first stage, weekly changes in the oil price were regressed against weekly percentage changes (returns) to the trade-weighted dollar index, the S&P 500 equity index, and the level of oil stocks as reported by the Energy Information Agency.
The model explains 31% of weekly changes in the oil price. The three predictors all are significant above the 99 percent confidence level.
In the second stage, the weekly changes are added up into a forecast of the level of the oil price during the past four years. The forecast of levels has an 87 percent fit with the actual oil price. As of March 1, the oil price was trading marginally lower than the model forecast-quite the opposite of a risk premium.
(5) Generalized Auto-Regressive Conditional Heteroskedasticity is an econometric technique used to calculate volatility by giving extra weight to extreme movements.