How Likely Is Another Oil Price Crash?

|
 |  Includes: BNO, CRUD, DBO, DNO, USL, USO
by: Inefficiency Hunter

In 2008 oil dropped from $144 per barrel to $37 dollars per barrel. Could this ever happen again, and could prices levels like that ever be sustained?

The questions are:

  1. What led to this degree of volatility?
  2. Could this ever be repeated?

By fundamentals I am referring to factors such as oil supply and reserves, the demand created by the global economy, known political factors such as ongoing conflict in the Middle East and price shocks due to unexpected events such as attacks or natural disasters, although some authors include nothing but physical supply-demand factors in “fundamentals”. I am not including speculation, although some do.

The reasoning between the distinction between fundamentals and speculation is that it is likely that the former are more predictable than the latter. If all oil price movements are due to the “animal spirits” of the markets, we are left with a high level of volatility, and a “known unknown”. If fundamentals are still important, we can at least have some idea about what to expect.

Buyers of oil in the futures markets can be divided into two categories: hedgers and speculators. Buyuksahin and Harris (2009) make a distinction between “commercial” participants in futures markets, such as oil producers, and “non-commercial” participants, such as traders, who are interested in oil as an asset rather than an input or output. Hedgers use the futures market to offset the risk that crude prices will drop (for producers) or rise (for buyers). In his recent book, "Understanding Oil Prices," Salvatore Carollo claims that current oil price trends have less and less to do with fundamentals, and more to do with speculation, amongst other things. I will begin by reviewing the findings of various academic studies into the events of the first decade, to see whether they attribute much importance to speculation.

Hamilton (2008) explains the way futures prices lead spot prices. Refiners need to buy crude, either by buying crude forward with futures and storing it in an inventory, or by buying oil when they need it. If the price of futures goes up, there will be an increase in demand for spot oil. Whether the markets are contango or backwardated, the futures price leads the spot price in a self-fulfilling loop. Hamilton merely gives a theoretical background to the behaviour of oil prices, however; he doesn't make any statistical calculations regarding speculation.

Coleman (2011) regresses oil prices against a number of variables to see whether there is a correlation. He uses the variable “Paper: physical volume” as a proxy for speculation. This is the ratio of the face value of futures contracts to global oil production. It is assumed that if there is more speculative activity relative to hedging activity, the face value of the futures contracts will increase. Coleman finds this variable significant at the 1% level (there is more than a 99% chance that speculation impacts oil price). He reports that in late 2007, speculation contributed to over $45 of the increase in the price of oil, and that one standard deviation of the speculation variable affected oil price by $8.5. Coleman also finds that other variables are highly significant, such as corporate bond yield rates, global GDP and military activity in the Middle East.

Coleman (2011), however, conducts no test for causation. It is exactly this that Buyuksahin and Harris (2009) attempt. They do not dispute a correlation between speculative activity and oil price – the participants in the oil futures market increased dramatically during the oil price spikes of 2007-8 - but they do not assume causation. They carry out tests of Granger causality, to test whether the number of participants in the futures market lead changes in oil price, and find that they do not. Instead, they find that the activity generally follows from price changes, suggesting herding, possibly increased volatility, but not necessarily upwards pressure. They explain that their results confirm the findings of the "Task Force Interim Report on Crude Oil" (2008), which found that changes in oil prices leading up to the global economic crisis were due to fundamentals rather than speculation. They point out that it is flawed to perceive hedgers and speculators as completely opposing forces. When hedgers choose not to participate in futures markets because they believe prices will increase, they are withholding their downward pressure in the same way as speculators are putting upwards pressure on the market.

Fan and Xu (2011) take yet another approach. Firstly they list a range of studies that have found varying results on the importance of speculation, from none to some. Then they introduce their approach, in which they assume that the futures market underwent some structural changes over the period 2000-9, which they divide into three periods: the “calm” period between 2000-4 the “bubble accumulation” period between 2004 and mid-2008, and the “global economic crisis” period between mid-2008 to the end of 2009. They use two variables for speculation: the percent of positions held by non-commercial traders, and the percentage of long positions held by non-commercial trades. For the full period they find that the speculative variable has only a weak effect. They find it has a significant and stronger effect during the relatively “calm” period between 2000-4, during which market fundamentals aren't significant, proxied by the Baltic Dry Index, a measure of the cost of shipping. During the bubble accumulation period, speculative factors still appear to play an important role along with the gold price and SP 500, whilst the BDI is still not significant. In the “global economic crisis” period, speculation variables are no longer significant. The BDI, SP 500 and the gold price are strongly significant, and in some cases their direction has changed.

We can draw the following conclusions from these studies:

  • Speculation is likely to play some role in oil price movements in some cases.
  • Claims that speculation is the main mover of oil prices should be treated with caution when they concern only correlation and not causation.
  • The importance of speculation changes of different periods, and the latest period tested appears to give less importance to speculators.

We cannot assume that the price crash of 2008 was due to speculation, therefore. Although Fan and Xu (2011) acknowledge that the crunch on global liquidity lead to the withdrawal of speculative funds over this period, they explain that it is precisely this fact that enabled other factors to become more significant. Variables other than speculation are likely to be able to have a strong impact on oil prices at the moment, those such as the physical demand and supply of oil, the stock markets, the gold price and the USD. According to Fan and Xu, in the latest period these all had a positive effect on oil prices, although in the “bubble accumulation” period there was a negative correlation between stocks and oil. Importantly, if Europe hits a recession then this could well bring down the price of oil. Although I am not saying we will see a repeat of the crash of 2008, we cannot rule out a significant drop.

End Notes

  • Buyuksahin, B. and Harris, J. (2009), 'Do Speculators Drive Crude Oil Prices?' The Energy Journal.
  • Carollo, S. (2011), Understanding Oil Prices.
  • Coleman, L. (2011), 'Explaining crude oil prices using fundamental measures', Energy Policy.
  • Fan, Y. and Xu, J. (2011), 'What has driven oil prices since 2000? A structural change perspective', Energy Economics.
  • Hamilton, J. (2008), 'Understanding Crude Oil Prices', The Energy Journal.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.