Oil Price Crash of 2014 / 15 Update

Towards the end of last year I had a couple of posts, the first explaining the oil price crash of 2014 in terms of a simple supply – demand model and the second using this model to anticipate where the oil price may head in 2015 and 2016. In light of the supply, demand and price action of the last six months both of these posts now need to be updated and revised.

The 2014 Oil Price Crash Explained
Oil Price Scenarios for 2015 and 2016

In my Price Scenarios post I forecast a Brent price of $56.50 for December 2015 and with Brent spot currently around $60 this is looking quite good. So far this is panning out in the right direction but for the wrong reasons which does not count as being correct in my book.

Figure 1

The raw oil price and production monthly data that lies behind the model can be divided into 7 legs. 1) Jan 2002 to April 2004 oil supply was elastic allowing demand to grow with little impact on price……

Figure 2 The supply and demand curves are based on a model first proposed by Phil Hart in an Oil Drum post from 2009 (link no longer working). Supply, fitted to the data, changes from elastic (low gradient) to inelastic (steep positive gradient). Demand (conceptual) is inelastic with steep negative gradient, i.e. high price will suppress demand but not by a lot. The price is struck where the supply and demand curves intercept. The movement of the inelastic parts of the supply and demand curves relative to each other can result in large movements in price for relatively small movements in supply or demand. The grey line shows how demand increased from 2004 to 2008 against inelastic supply sending the price sharply upwards (the intersection of the two curves, supply and demand, should always be balanced by price.)

2) May 2004 to July 2008, OPEC spare capacity shrank and demand continued to rise against inelastic supply sending the monthly average price up from $34 to $133 / bbl (Figure 2)

Figure 3 Between August 2008 and February 2009 demand collapsed, in part due to the unwinding of speculative positions, sending the oil price back to from whence it came.

3) Aug 2008 to February 2009, the financial crash, in part caused by high energy prices, saw demand collapse and the price retraced its ascent falling from $133 to $43 / bbl (Figure 3).

4) Mar 2009 to May 2011, QE, debt inflation, OPEC supply reduction and overall management of our not so free markets saw the oil price recover to $123 / bbl (Figure 1).

5) Jun 2011 to Jun 2014, the impact of shale drilling in the USA began to feed through to rapid supply growth which once again became elastic allowing demand to increase with little impact upon price (Figure 1).

So far so good, these are the sequence of events described in my earlier posts. It is what happened next that is seriously at odds with my forecast. My 2015 / 16 price forecast was largely based on assumption that we were witnessing history repeat and that price collapse was to a large extent driven by weak demand combined with OPEC’s decision to abandon the policy of restraint. What has happened can in fact be explained by over-supply alone.

Figure 4 Since 2008 oil supply grew significantly, hence the supply curves have moved to the right. A significant increase in supply between August 2014 and January 2015 (grey curve) not met by an increase in demand, sent the oil price into a tail spin.

6) August 2014 to present saw supply increase from 93.1 to 96.1 Mbpd. A 3 million bpd supply increase against weak / static demand sent the price crashing down. Where did this oil come from? In the past OPEC have cut supply by at least 3 Mbpd to support price and their failure to do so this time is sufficient explanation. If one needs further convincing of the over-supply argument, US production has risen by over 1 Mbpd from August 2014 to May 2015. There remains a weak demand component to the story in that in past cycles OPEC has managed weak demand by cutting supply. This time they have not resulting in over-supply.

Figure 5

7) The final part of the story is that demand has risen, stimulated by low price, to mop up  that extra supply and providing the recent support to the oil price, driving Brent back over $60 / bbl (Figure 5).

What Was Wrong With Earlier Scenarios

There were two main problems with my Oil Price Scenarios for 2015 and 2016 post. The first is that I anticipated weak demand but went further and assumed that demand would fall as in 2008 and this would be the principle mechanism for price collapse. While demand then probably was weak, it did not fall. The second is that while I anticipated that supply may not fall for over a year, I did not anticipate the momentum for supply growth. Hence, the price movement thus far is largely as anticipated, but it has come about by supply growth and not a fall in demand.

What next?

It has always been a fickle black art to try and forecast the oil price. One of the first points to recognise is that the demand curve shown in all my charts, derived empirically from the 2002 to 2008 data may no longer apply. The Global economy is littered with mine fields, the main one staring Europe in the face is GREXIT. The second is unsustainable debt levels in many major economies starting with Japan, followed by Italy, France, the UK, the USA and China. And there are geopolitical scars from the early skirmishes of WWIII all around the borders of Europe. These negative factors need to be weighed against the warm glow of cheap oil flowing through the global economy. I find it impossible to judge the balance between these forces.

On the supply front, it was widely anticipated that the collapse in US drilling would bring about a fall in LTO production by this summer that has so far failed to materialise. The US oil rig count is stabilising at the 600+ level and I am beginning to wonder if this might not be sufficient to sustain production levels. A significant drop might not occur. And thus far, significant falls in production have failed to materialise anywhere.

My Oil Price Scenarios for 2015 and 2016 anticipated a significant drop in production in 2016 that would send the price back up towards $100. Now I’m not so sure. The momentum built in recent years on the back of high price may take more than 12 to 18 months to dissipate. The industry is doing all it can to cut costs in order to adjust to the lower price environment. Those companies locked into high cost developments will pour gasoline on the bonfire and may have to chew losses for a number of years.

Recent history has not repeated and that makes it nigh impossible to predict the future with so many unconstrained variables.

This entry was posted in Energy and tagged , , . Bookmark the permalink.

17 Responses to Oil Price Crash of 2014 / 15 Update

  1. artberman says:


    This is an excellent redux of your earlier assessments and I agree that prediction forward is difficult.

    The point that I question is that demand has roared back. Your chart shows that global production has stalled in May with a decrease of 100 kbpd compared with April, an insignificant amount that is well within the error of recent reporting. December 2014 fits dangerously closely in the April-May cluster but that was part of the supply-increase-not-met-by demand part of the curve.

    As I showed in my recent post (http://www.artberman.com/a-year-of-lower-oil-prices-crossing-a-boundary/), demand growth thus far in 2015 is down to a paltry 0.5% YOY.

    I don’t think the case is there for resurgent demand because of low prices…yet.

    The interesting question is, Why has production apparently stalled in May?

    As you know, there is quite an uproar here in the U.S. that EIA reporting is way off compared with Texas and North Dakota where most of the tight oil production comes from. I do not subscribe to that view but I do not agree that 600 rigs can stabilize U.S. tight oil production.

    Rather, I think that slightly improved cash flow from price increase February-May 2015 plus record capital supply from stupid money has allowed increased completions from the backlog of drilled-uncompleted wells. That backlog may last for some time because it is large (we only know the number for North Dakota and it is nearly 900 wells) but it is not sufficient to stabilize US production more-or-less permanently.

    All the best,


    • Euan Mearns says:

      Art, US had 1609 rigs drilling oil and this fell to 628 a week ago. The 1609 rigs gave production to 1) cancel declines, 2) grow production and 3) add to the inventory of uncompleted wells. To keep production constant the 628 rigs just need to cancel declines. Now I’m not saying that this is enough rigs to do so, but maybe 800 rigs is. But as you point out we have all those uncompleted wells – savings in the bank. I just don’t think we are going to see US production nose diving near term. Plateau / slow decline maybe.

      Recent action on Friday with rig count jumping from 628 to 640 sends cat among pigeons. WTI may be heading back to $40. This looks increasingly like a Mexican standoff between USA and KSA.

      I didn’t mean to give impression of resurgent demand for oil, simply that low price has fed through to increased demand. In order to comment more meaningfully here I need to assemble the IEA storage numbers.

      Not sure I know where the EIA are at. I stopped following there 6 months out of date stats. The very fact that the USA has allowed this to happen fills me with suspicion. Its like the FED reporting 3 months in arrears what the interest rate is / was / is going to be.



      • artberman says:


        We agree on most points, then.

        The issue with rig count is that of the 1,600, many or most were drilling crap plays (Mississippi Lime, Tuscaloosa Marine Shale, etc.) or were drilling marginal wells in the better plays. U.S. tight oil production will fall by several 100s of 1000s bpd but the decline will be spread over many months. If oil price stays in the $60 range, then more wells will be drilled and the decline will be deferred or muted.

        Things will hit the fan in Q4 2015 and Q1 2016 when credit redeterminations occur and when PUDs will be written down based on 2015 SEC oil price, greatly reducing company net asset value thereby potentially triggering loan covenants.

        I will look for your re-assessment of demand with great interest!



  2. dereklouden says:

    Hello Euan,
    That’s the best analysis I’ve seen of what has happened and why prices have ended up where they are. On the geopolitics one of the major imponderables is the US/Iran nuclear talks. If the talks succeed and Iran regains access to world markets the price will fall (rightward shift in the supply curve). How far it falls will depend on the capability of Iran and whatever co-venturers and suppliers it works with to boost production.
    I’m expecting the talks to break down as the shale producers in America explain to Washington what success in the talks means for them.
    Production will increase in the UK over the next few years as legacy developments from before the price crash come online and re-developments boost production. What happens in the medium term is a lot more uncertain
    Kind Regards

    • Euan Mearns says:

      Derek thanks for link to your quite stunning analysis of the UK oil and gas sector. For those of haven’t clicked, there are 75 slides there and a treasure trove of data.

      Iran coming back on may add about 800,000 bpd over a period of time. Oil price negative to be sure, but perhaps not the price killer that some may expect. Its the sort of thing that OPEC may unexpectedly respond to even though KSA and Iran are enemies.

      The production increase in the UK over the next few years will be more gasoline on the oil price bonfire. Norway too. And with the US shale drillers tooling up once more, its hard not to be an oil bear right now. Grexit and a shiver flowing through the € zone won’t help.



    • Aslangeo says:

      Derek – Thanks for your treasure trove of data – I do however have a few points

      1. You mention the rough gas storage facility as the largest producing field on the UKCS – this is not new gas, but pre produced or imported gas which is taken out to meet peak demand having been re-injected in the summer – more information here – http://www.centrica-sl.co.uk/files/operational_guide.pdf – this facility accounts for the majority of the UK’s storage capacity in case of peak demand or outages

      The UK’s gas storage capacity is significantly smaller than that of continental European countries because until about 10 years ago the UK was not a gas importer

      2. There are significant maintenance issues with many older UK fields – I do not think that a significant increase in production is likely but a temporary end to the decline is going to happen for the next few years

      3. DECC do give an honest assessment of what they think might happen in the future – they are given full updates on a regular basis by the operators – I do not think that they – under-exaggerate for political reasons (Scottish independence)

      Overall however a very good piece of work

  3. Euan:

    After staring at your Figure 1 for a while I drew another line through it:

    It defines what might be termed the “lower limit of elastic supply”, and in at least the last 13 years of ups and downs the production/price plot hasn’t broken through it to any significant extent. Does this give us any kind of guide to the future behavior of oil prices, or is it just a statistical artifact?

  4. Willem Post says:


    Much depends on the definition of demand. Is demand defined as the total production? There are time lags between production and use. Some of the production may be added to existing storage and some of what is used likely comes from existing storage.

    One needs to know total world storage of oil, including quantities en-route, and their ups and downs.

    Purchasing agents of various entities will buy for various reasons. Either they have storage capacity and like to fill it with low-cost oil, for later use, or they are running low and need to buy to replenish, no matter what the price.

    The actual use creates most of gross world product, as some of that GWP is generated by the energy sector, which has a leading function, i.e., energy comes before other goods and services.

    This is a very complicated subject that may require a supercomputer to process the data, i.e., similar in complexity to a global warming computer program.

  5. Askja Energy says:

    Very interesting article and excellent charts.
    I find the discussion if oil markets are mainly demand driven or supply driven, quite fascinating. Such as discussed by Steven Kopits here:

  6. Javier says:

    By far the best approach to projections when dealing with complex issues to me is to define a series of scenarios that cover the major variables.

    For example the economy variable. We know that OECD to a large extent has not recovered from the 2008 crisis, and that recovery has come mainly from developing markets, many of whom are doing poorly (Brasil, Russia) or reducing rate of growth (India, China). Not recovering from a previous crisis after 7-8 years is unprecedented since the 30s-40s (70-80 years ago).

    So economically we can define two most likely scenarios:
    A. The inevitable next world economical crisis comes within the next couple of years.
    B. We continue to recover very slowly for the next 2-3 years.

    None of this scenarios is conductive to a significant increase in oil demand regardless of price.

    Regarding production we can also define two most likely scenarios:
    1. The decline in North-American oil production is offset by increased production from other producers, mainly Arabia Saudi, Iran and Libya.
    2. The decline in North-American oil production is not offset by other producers, so global C+C production starts to decline.

    Now we can see the effect of the combinations:
    A1. Reduced demand plus maintained production should crater the prices again
    A2. Reduced demand but reduced production should sustain mid-level prices
    B1. Slow increase in demand plus maintained production should slowly increase the prices. By far the best scenario.
    B2. Slow increase in demand but reduced production will again trigger a return to high priced oil.

    Now we can assign probabilities. This is very personal. Let’s say that A has a 60% probability over B, and 2 has a 60% probability over 1. That means:
    A1: 0.6 x 0.4: 24%
    A2: 0.6 x 0.6: 36%
    B1: 0.4 x 0.4: 16%
    B2: 0.4 x 0.6: 24%

    So I see a 48% chance of price volatility due to unbalanced reduction of demand or supply, 36% chance of price stability due to reduction of both demand and supply, and only 16% chance of the good outcome of slow increase in prices due to slow increase in demand and sustained production.

    Can’t be very optimistic, even not counting any possible black swan.


  7. Jim says:

    All the lements of Peak Oil are still there. Being an E&P is getting harder than ever. What is new is Canadian heavy/bitumen and shale exploitation. Canadian bitumen is somewhat limited (as ever, the good bits were taken in the 1970’s) and black oil from shales is also restricted to a few fields. The Marcellus ia dry gas, I believe). The balance of liquids production is gas liquids, condensate and lighter, which we don’t use much. I’m guessing that the big storage builds are propane etc.
    So…I am more optimistic than you are, that the “surplus” is mostly illusion.

    • Euan Mearns says:

      Jim, the storage part of the story I have not followed but it is suddenly important. Its a lot of work for me to compile the data from pdf IEA reports, but I’ve been meaning to look into it. Perhaps JODI or EIA have XL tables.

      The IEA don’t seem to understand the storage story. What you suggest about C3 to C5 perhaps makes sense – the storage build should all be in USA in that case.


Comments are closed.