Mar 252015
 March 25, 2015  Posted by at 10:41 pm Finance Tagged with: , , , , , ,  

Wyland Stanley Golden Gate Bridge under construction 1935

This is another essay from friend and regular contributor of The Automatic Earth, Euan Mearns at Energy Matters.

One comment on my part: Euan says ‘This has lead to speculation that weak global demand, stemming from masked economic woes, may also be playing a key role.‘ I don’t think the use of the term ‘speculation’ is appropriate here, because it seems overly obvious that China’s economic slowdown has played a major role in the oil price crash (and continues to do so). Even if there’s no ‘scientific’ proof, and even if the main media narrative remains OPEC overproduction and the inane meme of the cartel’s refusal to cut production, it certainly goes way beyond mere speculation.


Two of the factors in the oil price crash are well constrained: 1) oversupply of expensive light tight oil (LTO) in North America and 2) the decision of OPEC to not cut production. The third possible factor of weak global demand is not so easy to constrain but the current oil price crash bears many of the same hallmarks as the 2008 finance crash. This has lead to speculation that weak global demand, stemming from masked economic woes, may also be playing a key role.

In response to this, commenter Javier sent me a collection of 10 charts that he had collected from various internet sources together with his commentary that forms the basis of this joint-post. These charts tell a clear story of a major economic slowdown in China. This most certainly will be implicated in the ongoing oil price weakness. The $10,000 question is will China make a cyclical rebound like it has done in the past?

Figure 1 GDP growth. YoY = year on year % change. Note many charts are not zero scaled. China’s economy is still growing at 7% per year but has slowed down dramatically from 12% 5 years ago. Such change has happened before, notably between 1994 and 1998 linked to the Asian currency crisis. The oil price hit $10 per barrel in 1998. And in 2007 to 2009 an even more sharp fall related to the financial crash was also accompanied by a crash in the oil price.

Javier points out that in a country with rapid population growth a higher GDP growth rate is required than in a country with stable or declining population and he suggests that 7% is in reality approaching recessionary levels.

Figure 2 Decline in the growth rate of industrial production mirrors the decline in the growth rate of GDP (Figure1).

Figure 3 Fixed Asset Investment is a technical measure of investment in hard assets, infrastructure, property and plant and machinery. The graph tells the story of a country growing at phenomenal and increasing rates of growth up to 2005, that is the definition of exponential growth. From 2005 to 2009 the growth rate was flat, i.e. the growth was linear. From 2010 China is investing in fixed assets at decreasing rates of growth.

The change in 2005 is coincident with a change in growth and oil consumption in many OECD countries and therefore indicates that a global source of economic distress took place at about that date. China is changing the way it grows as it is not possible to grow exponentially forever.

Figure 4 Retail sales is a measure of national consumer expenditure. 2008 was the year of the Beijing Olympic Games, so we can pretty much discount the strong peak that year and see in this graph a strong growth in consumer expending until 2010. Since then retail sales have being growing at a slower rate, and current rate of growth is the slowest in ten years.

Retail sales growth will be driven by two factors. 1) the number of individuals economically active which in China grew at a phenomenal rate with the great migration to the cities and 2) the prosperity of those economically active.

Figure 5 Unlike in the previous graphs, China home prices have recently gone into an actual negative rate of change, which means that home prices are actually decreasing in China. If unchecked this could become a serious problem for China since real estate represents about 75% of household assets. So home prices are important for how Chinese perceive their own wealth. Falling house prices also lead to the risk of negative equity where the asset value falls below the amount of debt secured against the asset.

Figure 6 Rail freight is now falling at 16% per annum having gone into negative territory in mid 2014. In the case of China this is a measure mainly of national trade. This mirrors the picture of falling international trade as indicated by the sharp fall in the Baltic Dry Shipping Index.

Figure 7 Consumer Price Index (CPI) measures the average change over time in the prices paid by consumers on a representative basket of goods and services, while Producer Price Index (PPI) measures the average change over time in the selling prices received by domestic producers for their output. PPI shows an abrupt worsening for producers in 2011, since then the timid recovery ended badly in mid 2014 and PPI is now at levels found during recession. CPI shows that China is also flirting with deflation, which is bad news for banks and all individuals and organisations that have debt.

Figure 8 This graph shows that as export growth has stalled and imports are actually declining since 2013 and specially since mid 2014. This is one of the main causes of the commodity price crash that includes oil. China is buying less raw materials which is bad news for commodity producers that depend on China, like Australia. Note exports are still well in excess of imports and China still runs a huge balance of trade surplus.

Figure 9 Growth in oil consumption in China underpinned the bull run in the oil price. This growth in consumption stalled in 2012. The reasons for this should be clear from the preceding charts.

Figure 10 The phenomenal growth in China has been fuelled in part by an equally phenomenal growth of debt. The chart shows private sector debt has gone from much lower than OECD countries to much higher in just two decades. Debt is a fantastic growth hormone, but it is subject to a very strong diminishing returns curve. When there is too much in the system, it becomes a growth inhibitor.


A wide range of economic measures shows that China is undergoing a period of rapid economic slow down and is flirting with recession. China has grown to become the world’s second largest economy and strong Chinese growth has underpinned global growth for many years. Without it, the world faces the risk of another global recession. The slowing of growth in China means a softening of demand for natural resources, including oil and softening of demand for consumer products made in Europe and the USA.

Low oil prices may help stimulate growth in China and without this stimulus the global economy may already have been in recession.

Many of the charts are simply thermometers of the Chinese economy. Three of the measures, however, give rise to more concern about China’s ability to climb out of the malaise as it has done before. Falling property values, risk of deflation and debt saturation. Like many of the world’s leading economies, China, appears to have driven into the same economic cul-de-sac.

Sources of charts

Dr. Ed’s Blog
Snake oil trading blog

Who is Javier?

Javier holds a PhD in Biochemistry and Molecular Biology and has been a scientist for 30 years in molecular genetics and neurobiology. He wrote a blog on macroeconomy and investments from a cyclic point of view for over two years and currently writes a blog in Spanish about the economic crisis, energy crisis and climate change under the pseudonym Knownuthing.

Dec 172014
 December 17, 2014  Posted by at 10:19 am Finance Tagged with: , , , , ,  

Harris&Ewing F Street, Washington, DC 1935

This is another article from our friend in Aberdeen, Euan Mearns. It was first posted on Euan’s own site, Energy Matters. I earlier posted Euan’s The 2014 Oil Price Crash Explained on November 24, when the price of oil was still quite a bit higher than today. WTI ended that day at $75.74, it’s now $55.15. Here’s Euan:

A couple of weeks ago I had a post titled The 2014 Oil Price Crash Explained that was cross posted to over 20 other blogs including The Automatic Earth and Zero Hedge. In this post I use the empirical supply and demand dynamic described in that earlier post (Figure 1) to try and constrain the oil price a year from now and in 2016. The outcome is heavily dependent upon assumptions made about supply and demand and the behaviour of OPEC and the banking sector. Three different scenarios are presented with December 2015 prices ranging from $45 to $100 / bbl. Those hoping for a silver bullet forecast will be disappointed. Individuals must judge the scenarios on merit and decide for themselves which outcome, if any, is most likely.

Figure 1 The blue supply line is constrained by monthly production – price data from 1994 to 2008 and shows how supply became inelastic to demand post-2004. As demand continued to rise, prices rose exponentially to $148 / bbl in July 2008 before crashing all the way down again. The blue supply line in this chart is shown as a faint blue dashed line in all other charts to provide a frame of reference.

But first a look at the recent response of oil price dynamics to fluctuations in supply and demand.

OPEC spare capacity

Part of the key to understanding how the global oil market performs is to look at OPEC spare capacity data which gives a picture of how OPEC have provided or withheld capacity to try and retain order in the oil markets. OPEC suspending their market interventions has caused the recent oil price rout.

Figure 2 OPEC have tried to maintain order in the oil market. Rather than allow price fluctuations to control supply and demand, OPEC have aimed for a price that suits them and tried to maintain it by reducing and increasing supply in tune to fluctuations in global demand and non-OPEC supply. The picture of OPEC spare capacity therefore reflects fluctuations in the global oil market.

Over the past 10 years there have been three market cycles. Two of those have had roughly 3 years duration and amplitude of roughly 2 Mbpd (Figure 2). These sit either side of a larger cycle of 4 years duration and amplitude of 4 Mbpd caused by the 2008 financial crash. These cycles represent OPEC responding to global demand and non-OPEC supply changes. The smaller cycles may be viewed as “normal” and the larger cycle as rather extraordinary. OPEC intervention provided price stability of sorts. Without it we have price volatility that requires production to be balanced by varying demand and varying non-OPEC supply.

The spare capacity data suggests that demand / supply imbalance may last three years, requiring 18 months to work through to the mid-cycle point where over-supply turns to under-supply. It is by no means certain that the market will respond to the same time dynamic when we are now dependent upon natural production capacity wastage to occur as opposed to OPEC simply closing the spigot. But this is all I have to go on. The downturn in the current price cycle began last July and we are therefore just 6 months in. Another year of pain to go for the producers, that is unless OPEC decides to intervene.

Supply or Demand Driven Markets?

It is also difficult to discern if the current over-supply state is down to excess production capacity or a reduction in demand. Both are likely but my opinion is that the price rout is demand driven with many parts of the global economy performing badly – Japan, China and the EU to name but a few. These are about to be joined by OPEC, Russia and Canada.

The graphic below from the newly published December IEA OMR (oil market report) tends to confirm this view. 4Q14 supply is flat while demand is down. By 1Q15 a 2 Mbpd supply-demand gap is beginning to open tending to confirm the position laid out above.

Figure 2b The oil supply-demand view from the December IEA OMR.

Scenario 1

In 2015 demand falls by 2Mbpd relative to summer 2014 peak. New 2015 non-OPEC production capacity of 1.4 Mbpd (OPEC forecast) does not materialise leaving the supply curve as it is today.

This leaves the oil price around $60 a year from now (Figure 3). In the interim the price may go a lot lower as production capacity continues to rise before falling back to current level at year end; and because of short term trends driven by speculation.

Figure 3 Scenario 1 December 2015. Supply capacity grows and then falls back to where it is today. Underlying ills in the global economy sees demand drop 2 Mbpd from the July 2014 peak. The price ends up at around $60 / bbl, where it is today. But in the interim may go on an excursion to lower prices followed by recovery. Note that the 2014 demand line is retained in other charts to provide a frame of reference.

In 2016, low price causes a fall in global oil production capacity of 1 Mbpd and an increase in demand of 1 M bpd. These very small adjustments see the oil price rebound to $105 / bbl by December 2016 (Figure 4). Every cloud appears to have a silver lining if you are an oil producer, but global oil production capacity is cut by 1 M bpd in the process.

Figure 4 The low price of 2015 gives the oil industry a hangover in 2016 and supply drops 1 Mbpd. At the same time consumers party and falling supply collides with rising demand sending the oil price back up to $105 / bbl by December 2016.

Scenario 2

Under Scenario 2, the oil price rout causes high cost, high debt producers to default on loans creating a new banking crisis that spills over to the main economy. Re-run of 2008/9 though perhaps worse since most banks and national government balance sheets have not recovered from prior crisis.

Demand falls by 4Mbpd relative to summer 2014 peak, but supply capacity is also cut by 1 Mbpd owing to shale and other high cost operators going out of business. In December 2015 the oil price stands at $45 / bbl (Figure 5).

Figure 5 The fall in demand experienced so far causes trauma to many global producers that default on loans with knock on to banking sector and the broader economy resulting in further falls in demand during 2015. The price rout continues but is tempered slightly by non-OPEC supply being reduced by 1 Mbpd.

The low oil price works its magic on the global economy that rebounds strongly in 2016 pushing demand up by 2 Mbpd. But the $45 oil experienced in 2015 seals the fate of another 1Mbpd production capacity that is lost. Rising demand collides with falling capacity sending the oil price soaring back to $100 / bbl by December 2016. But the world has lost 2 Mbpd oil production capacity as a result of the price rout.

Figure 6 The price rout sees supply fall by a further 1 Mbpd. But the ultra low price causes a major rebound in demand of 2 Mbpd in 2016 from an “oversold” position. The oil price recovers to $100 / bbl.

Scenario 3

OPEC blinks first and with both Qatar and Kuwait cutting production in November, there are signs that this may be possible. Much depends upon Saudi Arabia who could conceivably raise production to counter the cuts made in other Gulf States. In Scenario 3, OPEC cuts production by 2 Mbpd by December 2015. While demand falls by 2 Mbpd from the July 2014 peak. The oil price recovers to $100 / bbl by next December 2015 (Figure 7).

Figure 7 Early in 2015 OPEC succumbs to pressure from several members and cuts supply progressively for a total of 2 Mbpd over the year. The net demand fall from the July 2014 peak is cancelled by the supply cut and the price recovers to $100 by December 2015.

This effectively means re-establishing the status quo of recent years. In this scenario, it is not necessary to look beyond 2015 since OPEC have re-adopted their strategy of market stability at a price that suits all – including the high-cost producers.


Each of the scenarios see strong recovery in oil price to the region of $100 come 2016. The main differences are in the extent and duration of short term pain and in the global production capacity. Scenarios 1 and 2 sees production capacity falling by 1 to 2 Mbpd come December 2016 and this would mainly be non-OPEC capacity that is destroyed handing greater control of oil markets to OPEC. Scenario 3 sees capacity maintenance and with re-establishment of status quo and high oil price, further expansion of N America. We need to wait and see if OPEC does what OPEC says.

My estimation of probabilities goes something like this:

Scenario 1: 40%
Scenario 2: 50%
Scenario 3: 10%

So my weighted forecast for December 2015 goes like this:

($60*0.4)+($45*0.5)+($100*0.1) = $56.50

Every year around this time I have a bet on the oil price a year from now. A year ago I bet $125 and my friend $110. Rarely have we both been so badly wrong. I lost again :-(

Ilargi: And of course Euan’s friend Roger Andrews has the first comment again:

Euan: Excellent piece of work. My weighted prediction for December 2015 is ($60*0.45)+($45*0.5)+($100*0.05) = $54.50/bbl. I’ve halved your already-low scenario 3 probability because I just don’t see OPEC – and certainly not the Saudis – cutting production first.

I think the Saudis have decided that preserving their historic market share is worth a few years of privation, particularly when they can live off their accumulated fat in the meantime. I say “a few years” because I don’t think it’s out of the question that the price slump could go on for that long.

Each of the scenarios see strong recovery in oil price to the region of $100 come 2016. Will this be the next “oil shock”?

Nov 242014
 November 24, 2014  Posted by at 2:33 pm Finance Tagged with: , , , , , , , ,  

NPC Capitol Refining Co. plant, Relee, Alexandria County 1925

This is an article by our good friend Euan Mearns at the University of Aberdeen. It was originally published here .

  • In February 2009 Phil Hart published on The Oil Drum a simple supply demand model that explained then the action in the oil price. In this post I update Phil’s model to July 2014 using monthly oil supply (crude+condensate) and price data from the Energy Information Agency (EIA).
  • This model explains how a drop in demand for oil of only 1 million barrels per day can account for the fall in price from $110 to below $80 per barrel.
  • The future price will be determined by demand, production capacity and OPEC production constraint. A further fall in demand of the order 1 Mbpd may see the price fall below $60. Conversely, at current demand, an OPEC production cut of the order 1 Mbpd may send the oil price back up towards $100. It seems that volatility has returned to the oil market.

Figure 1 An adaptation of Phil Hart’s oil supply demand model. The blue supply line is constrained by data (see Figure 4). The red demand lines are conceptual. Prior to 2004, oil supply was fairly elastic to changes in price, i.e. a small rise in price led to a large rise in production. This is explained by OPEC opening and closing the taps. Post 2004, oil supply became inelastic to price, i.e. a large change in price led to marginal increase in supply. This is explained by the world pumping flat out. Demand tends to be fairly inelastic and inversely correlated with price in that high price suppresses demand a little. Supply and price at any point in time is defined by the intersection of the supply and demand curves. 72 Mbpd and $40 / bbl in 2004 became 76 Mbpd and $120 / bbl in 2008 as demand for oil soared against inelastic supply.

Figure 2

Followers of the oil market will be familiar with the recent evolution of oil supply and price shown in Figure 2.

Figure 3

What is less widely appreciated is that a cross plot of the data shown in Figure 2 results in the well-ordered relationship shown in Figure 3. Oil supply and price are clearly following some well established rules. This relationship led to Phil Hart developing his model shown as Figure 1.

Figure 4

Separating the data into two time periods brings more clarity to the process at work. The data define a fairly well-ordered time series beginning at January 1994 at the bottom left rising slowly to January 2004 and then steeply to the Olympic Peak of July 2008. The financial crash then caused the oil price to give up all of its gains returning to 2004 levels by December 2008.

Figure 5

The second time period from January 2009 to the present shows some different forces at work. Starting in 2009 some new production capacity was built. This was not in OPEC and is concentrated in N America where the light tight oil (LTO) boom took off supplemented by steady expansion of tar sands production. Prior to 2009, the production peaks were of the order 74 Mbpd. Post 2009 peaks of the order 77 Mbpd were achieved. About 3 Mbpd new capacity has been added. In May 2011 there is a significant and curious excursion to lower production not accompanied by a fall in price. This coincides with Libya coming off line for the first time and the loss of 1.6 Mbpd production. It seems possible that this coincided with weak demand and the fortuitous loss of production cancelling weak demand leaving price unchanged.

The EIA are always running a few months behind with their statistics these days, not ideal in a rapidly changing world. Thus we do not yet have the data to see the recent crash in the oil price. But we know the price has fallen below $80 and production is unlikely to be significantly changed. So, how do we explain production of roughly 77 Mbpd and a price below $80?

Figure 6

Figure 6 updates Phil Hart’s model (Figure 1) to take account of the oil supply and price movements of the last 5 years. Capacity expansion is achieved by adding 3 Mbpd to the former, well-defined supply-price curve (blue arrow). There is no a-priori reason that this curve should hold in the new supply-price regime, but for the time being that is all I have to work with. The red lines, as described in the caption to Figure 1, conceptually represent inelastic demand where high price marginally suppresses demand for oil. The recent past has seen oil priced at $110 with supply running at about 77 Mbpd as defined by the right hand red coloured demand curve. Reducing demand by about 1 Mbpd brings the price below $80 / bbl (red arrow).

The Recent Past and the Future

Old hands will know that it is virtually impossible to forecast the oil price. The anomalous recent price stability of $110±10 I believe reflects great skill on the part of Saudi Arabia balancing the market at a price high enough to keep Saudi Arabia solvent and low enough to keep the world economy afloat. The reason Saudi Arabia has not cut production now, when faced with weak global demand for oil, probably comes down to their desire to maintain market share which means hobbling the N American LTO bonanza. Alternatively, they could be conspiring with the USA to wreck the Russian economy? But Saudi Arabia is not the only member of OPEC and the economies of many of the member countries will be suffering badly at these prices and that ultimately leads to elevated risk of civil unrest. It is not possible to predict the actions of the main players but it is easier to predict what the outcome may be of certain actions.

  1. If demand for oil weakens by about a further 1 Mbpd this may send the price down below $60 / bbl.
  2. If OPEC cuts supply by about 1 Mbpd at constant demand this may send the price back up towards $100 / bbl.
  3. Prolonged low price may see LTO production fall in N America and other non-OPEC projects shelved resulting in attrition of non-OPEC capacity. This may take one to two years to work through but with constant demand, this will inevitably send prices higher again.
  4. Prolonged low price may see many specialist LTO producers default on loans, risking a new credit crunch and reduced LTO production. This would likely lead to a major consolidation of operators in the LTO patch where the larger companies (the IOCs) pick up the best assets at knock down prices. That is the way it has always been.
  5. Black Swans and elephants in the room – with conflict escalation in Ukraine and / or Syria-Iraq and a new credit crunch, all bets will be off.

[Ilargi:] And this comment to the article from Euan’s friend and collaborator Roger Andrews certainly warrants attention as well. (check the grey dots!):

Hi Euan:

I put this XY plot of the data together from the data links you supplied. It shows the same trends as your Figures except that I’ve plotted all the points on one graph and segregated them into five periods, with trend lines and arrows showing the overall “direction of travel” for each (arrows in both directions for October 2004-May 2009, which as you noted goes zooming up and then comes right back down again).

I’ve also projected production data for the missing months since May 2014 (shouldn’t be too far off) so that we can at least get an idea of how the latest trend might compare with the old ones. We seem to be in uncharted territory.