Volatility has always been the defining characteristic of oil and other commodity markets, defying repeated attempts to stabilise prices.
Volatility is present at all timescales from seconds, minutes and hours to days, weeks, months and years ("Oil makes only one promise and that's volatility", Reuters, December 2015).
At the macro-scale, the oil market has been characterised since its inception in the 1860s by a series of booms and busts lasting for years at a time.
Efforts to tame the boom-bust cycle through the control of prices and production have repeatedly broken down ("Crude volatility: the history and the future of boom-bust oil prices", McNally, due to be published in 2017).
Volatility appears to be an intrinsic quality of oil markets rather than an incidental problem to be solved through improved forecasting, management and coordination.
Traditional explanations for volatility focus on the low price-elasticity of supply and demand as well as long delays in investment and the backward-looking nature of price expectations.
But the concept of feedback loops borrowed from control engineering as well as the theory of complex systems can also help explain some of the endemic instability in the industry.
In Constant Crisis
"There is always too much or too little oil," economist Paul Frankel complained seven decades ago ("Essentials of Petroleum", Frankel, 1946).
The industry has "an inherent tendency to extreme crises" because production and consumption are not "self-adjusting".
Frankel blamed the lack of smooth adjustment on the limited responsiveness of supply and demand to moderate changes in prices, at least in the short run.
The result is "a price structure that allows for ups and downs which fail to bring relief from dearth or glut", Frankel wrote.
The industry is subject to "continuous crises" in which "hectic prosperity is followed all too swiftly by complete collapse".
Frankel's words, published just after the end of World War Two, are a perfect description of the subsequent boom and bust in oil during the 1970s and early 1980s, and again between 2004 and 2016.
Frankel argued the only hope for stability lay in the role of the major international oil companies, singly or in combination, as well as governments, acting as "eveners", "organisers" and "adjusters".
Frankel was writing before the creation in 1960 of the Organization of the Petroleum Exporting Countries, which was explicitly committed to stabilising the market.
OPEC's founding statute committed the organisation to stabilising international oil prices with a view to "eliminating harmful and unnecessary fluctuations".
But OPEC has been no more successful at ending the boom-bust cycle than previous stabilisation arrangements operated by the Texas Railroad Commission, the Seven Sisters, the Achnacarry Agreement and Standard Oil.
Sources of Instability
Frankel blamed the boom-bust cycle on the low price-elasticity of supply and demand, which was in turn due to the fundamental characteristics of production and consumption.
The riskiness and uncertainty of
oil exploration; high proportion of fixed costs in refining, transport and marketing; and lack of readily available substitutes for petroleum fuels and lubricants, all contributed to the unresponsiveness of production and consumption to moderate price changes.
Economists have also developed "cobweb theorems" which blame unstable prices on backward-looking price expectations coupled with delays in adjusting production ("Why oil prices are so unstable", Reuters, 2014).
Price expectations indeed appear to be strongly backward-looking in the oil industry. In the first few years of the 21st century, painful memories of the long period of low prices in the 1990s held back plans to expand production even as prices surged.
More recently, the production and investment plans of the major oil companies and U.S. shale drillers appear to have been based on the assumption the period of ultra-high prices experienced between 2011 and 2014 would be sustained forever.
When prices have been high for some time, it becomes an entrenched assumption that high prices will persist for the foreseeable future, and vice versa.
At the same time, changes in investment and production take a long while. It can take a decade or more to train an experienced driller or seismologist, and at least that long to bring many complex offshore oilfields into production.
Feedback Mechanisms
Oil markets are usually analysed with the language and concepts of economics but they also share many of the characteristics of complex systems ("Complexity: a guided tour", Mitchell, 2009).
Complex systems are "large networks of components with no central control and simple rules of operation that give rise to complex collective behaviour, sophisticated information processing and adaptation".
Complex systems often exhibit highly non-linear behaviour in which small changes in the initial conditions can generate large changes in outcomes.
Non-linearity makes the behaviour of complex systems chaotic and extremely hard to forecast over anything other than a short time horizon.
Thinking of the oil industry as a system helps explain other features. Borrowing from the language of control engineering and population biology, systems can be analysed in terms of feedback loops.
In a feedback loop, an initial change or shock to the system results in a change in system outputs which in turn influences the input conditions.
Negative feedback mechanisms tend to dampen the impact of the initial shock and are therefore stabilising and promote a rapid return to equilibrium.
But positive feedback mechanisms exaggerate and amplify the impact of the initial shock and are therefore destabilising and delay return to equilibrium.
The concept of feedback was popularised in the 1920s by communications engineers at the Bell Telephone Laboratory ("A history of control engineering 1800-1930", Bennett, 1979).
But feedback mechanisms have a long history in economics and were implicit in
Adam Smith's invisible hand and David Hume's price-specie flow mechanism.
The oil industry is characterised by multiple feedback mechanisms on the supply and demand sides of the market.
Feedback mechanisms operate both when prices are high and rising (2004-2014) and when they are low and falling (2014-2016).
The various feedback mechanisms all operate at different timescales, in some cases with a long lag, so the balance between them varies over time.
In the aftermath of a shock, positive feedback mechanisms may dominate, delaying the process of adjustment, while negative feedback mechanisms are more powerful later, promoting convergence to a new equilibrium.
Complex Systems
Thinking of the oil market as a system is also useful because systems can often be analysed as a set of separate sub-systems which interact with one another in complex ways.
In the case of oil, there is not really one "oil market" but a series of separate but closely related markets for crude, fuels, refining, oilfield services, engineering construction, drilling equipment, skilled labour, raw materials etc.
Each of these markets is subject to its own feedback mechanisms, operating at different speeds, the balance of which is constantly changing.
Rebalancing the "oil market" actually means rebalancing each of these markets, which is a complicated and lengthy process.
Economists and policymakers are enthralled by the concept of equilibrium but experience as well as the theory of complex systems suggests the oil industry has never actually been in equilibrium.
Oil markets are in a perpetual state of disequilibrium; at any given point they may be moving further away from equilibrium rather than towards it.
The industry has often been hit by a new shock while it is still adjusting and rebalancing from the last one.
The fact real oil prices have tended to fluctuate in a (very wide) band suggests equilibrating negative feedback mechanisms do dominate in the long run.
But in the short run, destabilising positive feedback mechanisms can and do prolong the adjustment ("Why is oil market rebalancing taking so long?" Reuters, 2016).
(By John Kemp)