Volatility in the Price of Oil since Hubbert's Peak and Investment Risk

Is volatility in the price of oil a cause of the "bad behavior" of the financial industry?
The former chairman of the US Federal Reserve Alan Greenspan provided the following response during questioning about his “ideology” before a committee of the US congress on the 23rd Oct 2008:

Greenspan said: "I have found a flaw. I don't know how significant or permanent it is. But I have been very distressed by that fact...."

A reflexive media and commentators jumped on a bowed Greenspan for the usual depressing reasons. However, was it possible that Mr Greenspan had let his mask slip for a moment ? The attention grabbing part of his sentence was “…how significant or permanent…”. Could it have been that Mr Greenspan was not saying he’d been wrong ? Instead, was he implying that conditions in the economy had changed unexpectedly and fundamentally in a manner that now made his old way of thinking flawed ?

This idea explores the possible causes for Mr Greenspan’s distress. In short, it proposes the outline of a mechanism for how volatility in the price of oil might have contributed to the apparently irrational actions of the financial industry. In particular, it speculates how recognition of the implications of this volatility might have rationally and quickly changed the incentives of certain financial insiders. The idea also proposes that in coming years, unpredictability in the price of oil will  become broadly understood (i.e., infamous) as one of the most serious and pressing threats to our economy and democracy.

(Note to Therramus - connect intro back to Greenspan's flaw of unknown significance and permanence)

Consider the following hypothetical sequence of events:

1. Global oil production peaks around the yr 2000 as predicted by Hubbert.

2. A new pattern of volatility in oil prices emerges shortly thereafter and this pattern continues to build through the present (and into the future).

3. This type of volatility is in the nature of being on the production down slope of a finite resource in great demand (i.e., oil). Variance and occasional very large movements were certainly evident in the upslope of the production curve. However, the frequency and severity of oil price movements that characterize the new pattern on the downhill side of the production bell curve are envisaged to be of a different order of magnitude.

4. Owing to the singular role of oil in the economy, volatility in its price begins to propagate in variable degrees into the volatility of the price of nearly everything else.

5. A general increasing volatility in prices translates into increases in risk of investment - indeed, due to unknowns in future prices and most especially of oil itself, real financial risks across the board are probably rising exponentially relative to price volatilities.

6. A smart, knowledgeable and initially small number of financial insiders anticipate and confirm the implications (as outlined broadly in 1 through 5) of being on the down slope of the global oil production 4-8 yrs earlier than the rest of us. These people don't need to know each other, but do need to have specialized knowledge and be capable of uncommon insight.

7. With financial risk increasing, and a still small, but growing number of insiders coming to understand what is going on, the incentives within markets shift rather quickly from protecting shareholders interests, to figuring out how to quickly “cash out”. This shift in incentive is the fundamental change (of uncertain permanence) that may be the real cause of Mr Greenspan's distress.

8. The behavior of the primary few spreads within the financial industry. Most of these secondarily affected individuals are probably oblivious to the ultimate cause (i.e., oil price volatility) of their actions. A fin de seicle ethos becomes pervasive. There is an unspoken or perhaps even quietly discussed urgency that you should make your “nut” (i.e., sufficient money to retire wealthy) and get out. Greed, self-interest and/or stupidity do the rest.

How the "cashing out" occurred is where my idea goes somewhat fuzzy. But the story does seem to conclude with tens possibly and hundreds of trillions dollars in worthless assets in the "shadow banking system". We know the rest - credit crisis, stock market crash, bailouts and a looming severe recession.

Why it is all about oil

A common reaction to this idea, is that its just another liberal attempt to blame everything on oil - the Iraq war, Global warming and so on. Ponder this. The tank of gas that powers your car for a week contains more energy than 6 people doing manual work for a year. It is silly, but imagine paying 6 people to drag your car around rather than using the car's engine. At minimum wage your "car team" would set you back more than $100,000 a year (vs ~$2000 a year for gas), and be slow to boot. You'd need another person  to wave a fan at you to replace your air conditioning. Now multiply what machines powered by inexpensive gas do across the entire economy and you start to get the picture. Think about this $100,000+ next time you feel bummed about paying $40 to fill up. Oil is cheap and makes life easy. It so important that we all but eat the stuff. This essential role in powering modern life is why fluctuations in the price of oil have the potential wreak havoc and also why it is front and center factor in the hypothetical mechanism above.

The hypothetical mechanism imagines a cause for our current economic collapse beyond ordinary human failings such as greed and selfishness. Human frailty is a necessary, but not sufficient factor. Speculating on the cause of this crisis is a weakness of this idea... as it is not so hard to be off track when there are so many unknowns. On the one hand, the workings of complex systems are ineffable....and well...complex. However, even in the most emphemeral of cases, the proverbial butterfly in South East Asia still has to beat its wings to provoke that storm two years later in North Carolina. Some causes are cryptic, but turn out not so hard to be resolve with the right perspective. For example, though not evident until Jared Diamond pointed it out, how the axial geography of Eurasia conspired to give a leg up to certain technological civilizations now seems blindingly obvious to any thinking person. Sadly, the actual mechanisms for many events are lost to the past - owing to various factors including acceptance of the conventional ignorance (i.e, wisdom), the tendency of victors to write the history and so on. Nicholas Nassim Taleb has written beautifully on this topic. Though often hard to locate or hidden there from view, there are always real and ultimate causes for events in the natural world. This we have learned.

The hypothesis posed here does NOT require an organized conspiracy to work. Simply put, the idea suggests that anticipation and/or confirmation of increasing volatility in the price of oil following its global peak in production provides a key to understanding a rationale  for the current financial crisis. Indeed, pricing unpredictability in this most fundamental of energy sources is suggested to be ultimately the necessary and sufficient factor in the present debacle (and perhaps future ones as well). All else, bad behavior of the Wall Street included, self-organizes and flows downstream from this consideration. Time will tell whether growing unpredictability in oil pricing is directly tied the "bad behavior" of certain parts of the financial sector or whether the musings provided here are a long-winded canard.

Data and Analysis

Events 1 through 8 of the “hypothetical sequence” outline  propositions for which a certain amount of data can be derived. For example, the “hypothetical sequence” suggests the following:

A. Volatility in the price of oil will have increased since Hubbert's predicted peak in global oil production yr 2000.

B. This increase in the volatility of the price of oil should propagate volatility into the price of other goods and services.

C. In turn, increasing volatility in the price of oil and pricing in general should result in increases in investment risk.

D. Certain financial insiders anticipated and confirmed the implications of being on the down slope of the global oil production prior to the rest of us.

(Note to Therramus - D needs to ref investors and investment risk)

In the following an attempt is made to derive data to support the propositions that are outlined in A, B, C, and D. It is probably fair to point out at this stage that although propositions B and C propose mechanistic linkages, the actual data presented will be correlative. This is one step better than providing an opinion not backed up by any facts. However, we are all aware that correlation does not imply causation. Proposition D will also be correlative in nature and even more loosely so than B and C. Here, I will explore how anticipation and confirmation may have occurred, paying special attention to what may have been learned from the peak in US oil production in the 1970s. These caveats being said I will proceed and hopefully, you will read on.

Propostion A: Volatility in the price of oil should have increased since Hubbert's predicted peak in yr 2000.

Getting hold of a free source of historical oil pricing data on the internet is surprisingly difficult. There appeared to be packages that could be purchased, but eventually what seemed to be a reliable and free data base was identified at the Illinois Oil & Gas Association website.

http://www.ioga.com/Special/crudeoil_Hist.htm

This site provides complete monthly data from the mid-1980s up until the present on the “HISTORY OF ILLINOIS BASIN POSTED CRUDE OIL PRICES”. Not perfect, but a start.

The chart in Figure 1 is a simple plot of monthly crude “oil price” over a period from 1986 to 2009.



The sharp rise and fall in oil in 2008 is a part of this story that is all to familiar to motorists.

The next problem was how to calculate an index in the volatility of oil price based on these data. It was reasoned that this index of volatility should reflect the spread or variability in price over successive months. To acheive this a straightforward approach based on parametric statistics used together with Microsoft Excel. The index was calculated as follows. The monthly oil prices for Jan, Feb and Mar of 1986 were $22.50, $16.00 and $14.00 respectively. First  the standard deviation (SD) of the first 2 numbers (i.e., $22.50 for Jan 1986 and $16.00 for Feb 1986) was calculated as an index of their spread. This SD was 4.60. Next, the standard deviation of the 2nd and 3rd numbers for (i.e., $16.00 for Feb 1986 and $14.00 for Mar 1986) was estimated to give an SD of 1.41. These calculations were carried out for successive pairs of months for all 276 months from Jan 1986 down to Dec 2008.

The chart in Figure 2 plots the index of “Oil Price Volatility” in red in the left hand Y-axis along with the monthly oil price plotted in blue (now in the left hand Y-axis). Eyeballing the "seismograph-like twitchings" on this chart suggests that volatility has been increasing, particularly since 2002. Something that is particularly intriguing about this chart is a progression of increasingly larger "peak twitches" that appear to rise notably above a lower background of variance in the volatility index (see small arrows on Figure 2). This could be noise, but these higher amplitude pulses do seem to have pattern which looks as though it may be meaningful.



A few notes on Figure 2 and its underlying assumptions. First, one has to worry about deviance from normality etc. The approach used is not perfect. However, the aim was not to clear the forest, but to cut a narrow path through it. For this purpose, using SD as a proxy for spread seemed sufficient for now. Second, it is understood that variance (i.e., SD squared) is the preferred method by statisticians for estimating the spread of a sample population. Variances were calculated for the list of SDs and these calculations tended to accentuate the trend seen in red in figure 2. However, it was decided that the simpler SD calculation did the job, so it was stuck with. Third, calculating SD for 3 or 5 successive months was also carried out. The same overall pattern resulted from these calculations as was found for the 2-monthly calculations in figure 2. As this was the case, the calculations from monthly pairs was used as it was surmised that what would be lost in noise by using bimonthly figures would be made up with increased signal from a calculation based on higher temporal resolution. The thinking was that the robustness of approach was not so much in the individual bimonthly calculations of SD, but in the trend that emerged from them over the period between 1986 and 2009. Fourth, it would have been great to get one's hands on daily, rather than monthly, oil price data to do the calculations – but as mentioned, such data is hard find on the internet. Finally, coefficient of variance (i.e., average for the 2 monthly numbers over their SD x 100) was considered as another parametrically based index of volatility. However, on reflection  it was deduced that an SD-based index would be better as it would reflect absolute, rather than mean normalized changes, in the price of oil. Absolute volatility in price is surmised to be a best estimate of the factor propagating instability into the prices of other goods per the “hypothetical sequence”. Mean normalized SD (i.e., coefficient of variance) would be a dampened index relative to a measurement based on the absolute variability in oil price. And in turn, it was surmized the SD of oil price volatility (not coefficient of variation) would be more proximally related to increasing investment risk also as per the “hypothetical sequence”.

Qualitatively, the underlying trend in oil price volatility over time in Figure 2 appeared non-linear. Microsoft Excel was used to calculate a 3-factor polynomial fit to the scatter plot. This trend as represented in the pink non-linear regression line overlaid the “red” price volatility data in Figure 3. The R squared of the regression line is statistically significant and so on as provided in the chart. Interesting features of the pink regression line include that it starts to move notably from around year 2000 (i.e., Hubbert’s predicted production peak as in the "hypothetical sequence"). It also climbs in what appears to be an exponential manner from this time to the present (i.e., Jan 2009).



The trend in oil price volatility is isolated in Figure 4. From this it can be seen that volatility based on the calculation approximately doubled between yrs 2000 and 2004 and then approximately doubled again between 2004 and 2006. What happens next is an interesting question ? A clue to this might come from a smaller bump in volatility occurring during the 1990s (see Figure 4). Based on this pattern it is speculated that the current exponential rise will flatten as time proceeds and eventually becoming a second bump, albeit of higher amplitude than that occurring during the 1990s. After completion of the current bump, one imagines that a rise and fall in volatility has the prospect to occur over again, if and when economic recovery occurs. With a tightening supply, cycling variance in oil price could be a stomach churning ride.



The regression line in figure 4 was recalculated for different time frames in an attempt to figure out when in the past the trend line would have indicated a significant uptick in oil price volatility. This was done to test a speculation in point 6 of the “hypothetical sequence” in which it was posed that data on rising oil price volatility and its implications on investment risk  may have been available to financial insiders 4-5 yrs ago. This was done by removing yr 2008 and then calculating the regression line, and then doing the same with yrs 2008 and 2007 removed from the regression and and so on. Doing this, it emerged that the signal for the rising trend in oil price volatility was faintly evident (in an empirical sense) from the end of 2003 and continued to build over successive years from that time. Prior to 2003 it would have been very tough to pick up the rising trend signal from the approach taken here.

'In conclusion, looking back from 2008 there is quantitative evidence for a rising trend in the volatility of the price of oil that commenced from around year 2000 coinciding with Hubbert's predicted global peak in oil production. The signal confirming the start of this upward trend in volatility would have been evident 4 to 5 years ago from around the end of 2003.'

B. This increase in the volatility of the price of oil should propagate volatility into the price of other goods and services.

As a first step, the monthly year-to-year inflation rates between 1986 and 2009 were used to derive an index of general price volatility in goods and services. Unfortunately, there is evidence that the Government may "monkey" with statistics on inflation. One would imagine that under-reporting inflation has benefits for the "powers that be" that would include making them look good or "rigging the table" on payouts of inflation-proof bonds and so on. While it may be easier to manipulate absolute levels of an index, it is assumed that cloaking the pattern of a second order variable such as volatility is more difficult for would be cheats - more on this later. So, one won't be ungracious. The authorities are taken at their word on their monthly inflation calculations - at least to start with.

Monthly numbers on historical inflation rate (calculated from the so-called CPI-U) were obtained from:

http://inflationdata.com/Inflation/Inflation_Rate/HistoricalInflation.aspx

Figure 5 shows a co-plot of monthly inflation (orange as calculated from the consumer price index - CPI-U), the index of inflation volatility based on bimonthly SD (green), and the trend in the inflation volatility index (light green line) over the period between Jan 1986 and Jan 2009. Bimonthly SD was calculated as an index of volatility of CPI-U/inflation exactly as was described above for the index of oil price volatility. As was the case for oil price volatility, a trend line was also calculated by Microsoft Excel as a 3-factor polynomial fit to the CPI-U/inflation volatility data (equation and R square value are provided on the chart).



The pattern for CPI-U/inflation volatility is different in subtle ways from that of oil price volatility. Nonetheless, a commonality of increasing volatility from the late 90s rising until the present is evident. The dark green line indicating the CPI-U/inflation SD switches with increasingly greater amplitude over this period. The lighter green regression line confirms the trend.



In Figure 6 the trends in monthly oil price and inflation (CPI-U) volatility are plotted together on the same time scale (i.e., Jan 1986-Jan 2009). A rise from the late 1990s to the early 2000s in the trends for both indices is apparent. However, an important problem is that the inflation SD trend appears to reach it lowest point in 1997, slightly earlier than the oil price SD trend (which as discussed above bottoms around 2000). This feature is referred to as a "non-coincident bottom" on the plot. This lack of coincidence in the lowest trough of the two plots is problematic for the "hypothetical sequence" in which it is implied that oil price variance should causally underlie inflation volatility. It is also an issue from the point of view of suggesting that some sort of coincidence occurs between the timing of Hubbert's prediction of global peak oil (i.e., yr 2000) and inflation volatility kicking off. One can perhaps account for this "non-coincidence" as a statistical error, the effects of more than one time-dependent factor on inflation, the result of government manipulation and so on. Nonetheless, this discrepancy is a concern and further clarification is required.

There is much question as to whether the "official" US inflation figures reflect the actual rate of inflation in this economy. An interesting website that provides alternate calculations of inflation is "John Williams Shadow Government Statistics":

http://www.shadowstats.com/

At this site, inflation numbers are calculated using methodologies that were utlized in 1980s. A site newsletter indicates that this "Alternate Consumer Inflation measure, reverses the methodological gimmicks of the last 25 years". In other words, the site suggests that Government "cheating" on inflation has been removed from estimates so that the true variance in inflation over time can be seen. Interestingly, a strong advocate for the current "monkeyed with" Govt. version of inflation calculation was none other than Mr Alan Greenspan of the current distress. Unfortunately, shadowstatistics.com does not provide its numbers free, and as there was no money or desire to purchase  said data, the numbers were interpolated from a plot helpfully provided at the website. I might add that if I been caught by my partner paying just under $200 "for just some numbers off the web!"... there would have been serious repurcussions. Though the coming "the end of the world as we know it" is being reflected upon here, the author has no wish to sleep alone with the cats during the interim. The plot below in Figure 7 provides the fruits of the effort to (re)generate data on "honest" inflation and inflation volatility from publicly available information. And MAN... it is a doozy !



What we see from the "honest" numbers in Figure 7 is that volatility is now picking up slightly later into the 2000s than calculated from the official Govt. stats. Indeed, it appears that shadow inflation rate volatility does not start a new pattern of more vigoruous "twitching" until after 2000. The light green plot of the volatility trend confirms a low point for the shadow numbers occurs from around the year 2000. Figure 8 co-plots the trend for oil price volatility and shadow inflation (CPI-U) volatility and from this it can be ascertained that the pattern of variance of the two indices since the mid 1990s is very similar. The non-linear exponential rise in the inflation volatility curve is now apparently lagging a little behind the rise in oil price volatility. Most importantly, the low point of the two trends curves now share a "coincident bottom" and a coincident bottom is a lovely thing.



These findings are gratifying in the sense that a case can now be made from the data that variance in general pricing (as reflected by the shadow inflation (CPI-U) volatility calculation) is plausibly downstream from oil price volatility, at least since peak oil. However, there are dangers here. First, close and convenient correlations in timing and curve shape between 2 trends do not imply causality. Second, by accepting the shadow statistics inflation over those of the Government, multiple errors are possible. For example, the mistake of confirming one's own preconceived ideas has to be watched for. When building a story from a tableau of similar, but subtly differing facts it is all to easy to assume that the data variant that best seems to fit the narrative is the data that must be right. Also, the Govt. may NOT be lying. Yes, yes, I know.... but one must keep an open mind. However, even if the inflation numbers have been "monkeyed" with, it is hard to conceive that the intent of the change in the 1980s was to uncouple the appearance of a connection between volatility in inflation and oil price that would occur some 20 years later. Even  Maestro Greenspan could not have such prescience. The "monkeying" was likely done for more mundane reasons. All the same, the effects of this messing with the inflation numbers may provide a salutory lesson. Never fuck with the fabric of reality, as it may stop you from seeing something else that is unexpected and really, really important !

In conclusion, looking back from the present it is cautiously suggested there is quantitative evidence for a rising trend in the volatility of the prices of goods and services as reflected in the CPI-U that coincides with and perhaps lags the emergence of increasing oil price volatility from year 2000.

Please note that the following sections are work in progress. Additions will be made on an ongoing basis
C. In turn, increasing volatility in the price of oil and pricing in general should result in increases in investment risk.

D. Certain financial insiders anticipated and confirmed the implications of being on the down slope of the global oil production prior to the rest of us.

Assuming that the mechanics of price volatility does turn out have actual or perceived adverse effects on investment risk, questions emerge. 1. Did certain individuals anticipate this problem ? 2 If so, was this information used by certain individuals to enrich themselves and/or influence the ethos of the financial sector wittingly or unwittingly ? If the "hypothesis sequence" has some semblage of basis in reality... and it probably is nothing more than an entertaining speculation... it seems unlikely that any culpable actors will be called to account.

Speculations on other Future Consequences of Volatility in the Price of Oil for Politics and Freedom

There is a another set of worrying questions concerning how we deal with rising oil price volatility in the future. What becomes of our present capitalist model with its emphasis on the invisible hand of the market, maximising individual freedom of choice and relying on enlightened self-interest ? This model does not seem equipped to deal with the unpredictable consequences of rising oil price volatility ? In this "brave new future" "will realizing your dreams", "having it all", "reaching for the top" etc be possible for most. Probably not, at least in the material sense. How will we organize ourselves politically ? Will cycling waves of volatility-induced economic disorder result in more centralized and regulated control by the authorities or chaos. What becomes of political freedoms in such a world ? Perhaps in the same manner oil price volatility could propagate instability into prices and increases investment risk, could other knock on effects include lability in long held social mores and ethical principles ?

Recent manifestations of oil price volatility induced instability in "democratic principles" might include acquiescence by elected representatives to the use of unprecedented amounts of tax-payers money in so-called bailouts,   unilateral preemptive acts of war by democracies (e.g., US in Iraq, Russia in Georgia, Israel in Gaza), increased use of unregulated surveillance by the state and the justification (and/or obfuscation) of the use torture as a tool by the elected authorities. Admittedly, the timing of a number of these manifestations do not fit the time line illustrated above in Figures 3 through 4. Nonetheless, the shocks caused by the ongoing effects of fluctuations in oil pricing could soften us up to accept increasingly draconian impositions. I heard a politician say "never waste a good crisis" recently. Unfortunately, this individual may be spoiled for choice in future emergencies available to him for charraling the flock.

On the more hopeful side, a realization of the consequences of oil price volatility in economics could provide individuals who consider themselves "conservative", in the sense it is understood in US politics, a rationale for a fundamental change in their thinking. It also would give urgency to the majority of self-identified "moderates" and/or "liberals"  who pay lip service to the idea of "addiction to oil" but don't do anything meaningful about this addiction. When it becomes understood that that volatility in the price of oil will absolutely ruin our economy and disrupt our governance systems long before oil runs out or we burn sufficient amounts of it to kill our  beautiful planet, then calculations  in all parts of the political spectrum could change suddenly and fundamentally. To put this another way, a beneficial effect of the widespread anxiety induced by recognition of the clear and present danger posed by price volatility could prompt a collective choice to urgently explore  how end to our dependence on oil. This recognition could be far more efficient in changing behavior than less immediate threats such as global warming and the eventual depletion of the resource.