Oil price forecasting by EIA and Merrill Lynch not going so well

Forecasting is a tricky business. There are many ways to do it wrong due to the many biases that affect the forecaster. Oil forecasting isn’t just a statistical problem solved with numbers, there has to be some domain expert knowledge applied and applied carefully.

Once a forecast is made, time will judge its accuracy. Use MAPE (Mean Average Percentage Error) as the yard stick, we can quantify the accuracy of past forecasts. Almost all forecasts will be off my some amount, but we really want to find a forecast method that gives the smallest error over time. Hence, we seek one with the smallest MAPE.

Below is a graph of the error of the EIA (Energy Information Agency), Bank of America’s Merrill Lynch, and a simplistic model called Naive.

The EIA and BoAML can't even beat the Naive model for forecasting 1Q, 2Q, 3Q, and one year out.

The EIA and BoAML can’t even beat the Naive model for forecasting 1Q, 2Q, 3Q, and one year out.

The Naive model simply looks at the price of oil the day before the forecast and says that is what it will be 3 months from now, 6 months from now, etc. There are 4 simplistic models in forecasting theory that set the bar low. To illustrate how simple they are, two of them are the mean price model, and the naive model. If your complex proprietary model can’t even beat (achieve lower MAPE) the simple four, you should give up.

What this graph is saying, is that a sophomore in high school could have made a better oil forecast than the EIA or Merrill Lynch. Except forecasting 18 months out, in which case Merrill Lynch redeems themselves.

In another post we’ll review EIA’s performance over a longer period of time than just 5 years.

WTI oil price forecast for May 2015

Oilfield-Intel, using methods developed by a team of analysts and petroleum engineers, has constructed a prediction model to forecast global petroleum prices. By analyzing supply and demand data from a number of sources and making predictions on key geopolitical events (particularly, OPEC meetings, major political changes in high-production geographies, and key dates on the financial calender) Oilfield Intel has established a model for forecasting the international price of West Texas Intermediate crude oil.

We then compare our own price forecast against several others from government agencies, financial institutions, think-tanks, and editorial publications. We weight these models according to their accuracy when tested against historical data. The product of these weighted averages is the forecast shown below.

The resulting forecast of WTI crude spot prices (in current US Dollars, not adjusted for inflation) are for the price average of the quarter.


2QTR 2015 – $56.67
3QTR 2015 – $58.24
4QTR 2015 – $58.36
1QTR 2016 – $60.50
3QTR 2016 – $63.44

EIA predicts US oil production to decline beginning of June 2015.  Currently, there are several West coast refineries down for seasonal maintenance that will soon be back online further consuming oil stocks.

Even with still growing oil inventories, prices have trended higher due to fundamentals on the demand side coupled with increasing hostilities in OPEC nations.

West Texas Intermediate Spot Crude Price Forecast

Oilfield-Intel, using methods developed by a team of analysts and petroleum engineers, has constructed a prediction model to forecast global petroleum prices. By analyzing supply and demand data from a number of sources and making predictions on key geopolitical events (particularly, OPEC meetings, major political changes in high-production geographies, and key dates on the financial calender) Oilfield Intel has established a model for forecasting the international price of West Texas Intermediate crude oil.

We then compare our own price forecast against several others from government agencies, financial institutions, think-tanks, and editorial publications. We weight these models according to their accuracy when tested against historical data. The product of these weighted averages is the forecast shown below.

The resulting forecast of WTI crude spot prices (in current US Dollars, not adjusted for inflation) for the first day of the specified quarter is as follows:

2QTR 2015 – $49.58
3QTR 2015 – $52.40
4QTR 2015 – $56.45
1QTR 2016 – $59.94
2QTR 2016 – $62.39

LIBYA – Oil Production Forecast for 2015

Libya Oil Production Forecast – 2015

Analyst: M Taylor

Libya is currently producing 482,000 bpd. We forecast Libya will increase production to 975,000 bpd by Dec 31, 2015 and 1.1 million bpd by June 2016. This is barring any unusual escalation in hostilities introduced by third parties.

After the Arab Spring in 2011, Libya’s oil production recovered from near 0 to 1.4 million bpd in under 7 months. This implies a 200k bpd recovery capability as a base case. However, the current civil war is more complex. In the Arab Spring case, there was minimal damage to the oil infrastructure. In the standing Libyan conflict, the disabling oil infrastructure by particular actors denies income to competing actors, so there will be more damage and a slower recovery of production. There are a series of vulnerabilities to shut down production. From upstream to downstream: the oil field itself, the pipeline, the terminal at the port. If either side can shut down any one of these, then the oil will stop flowing. This is leading to raids on pipelines, oilfield control buildings, and terminals.

The two main warring factions:
New General National Congress (GNC) in Tripoli is a claimed continuation of a now-defunct government body. It is made up of Muslim Brotherhood politicians that lost elections last year, and has militia allies in the Libya Dawn, LROR, the ALQ-linked Libya Shield Force.

House of Representatives (HoR) in far-east Tobruk, is the parliament of Libya disolved by the Libyan Supreme Court in November 2014. They control the weakened Libyan National Army and are still recognized by the international community as a legitimate government body for brokering a peace.

Two minor factions:
Shura Council of Benghazi Revolutionaries based in Benghazi
Islamic State of Iraq and Levant (ISIL) and similar IS radicals, as seen peppered throughtout the Arab and North African geography.

The GNC and allies do not possess modern aircraft weaponry. They do have some planes, but they are reported as unweaponized training craft. These planes have some form of hardpoints for bombs but the GNC have neither the talent to fly them as bombers, nor targeting equipment to do so successfully. While some aerial raids have occurred, the damage was minor and the raid mainly achieved fear. It is not expected they will keep much oil off the market in the future from the GNC air assets.

The pro-goverment in Tobruk has shown little interest in damaging Libyan oil assets held by the GNC. That may change if the GNC were reported to be selling oil via tanker.

The two minor Islamic groups have been disruptive, but not able to inflict lasting infrastructure damage. At this time, They are not a massive disruptive force to future production infrastructure, only security destabilizer. ISIS currently controls oil-rich Sirte and surrounding geography. There have been reports of explosive damage to control rooms in this oil field, which is relatively short-lasting damage. We expect this pattern to continue.

“A more recent phenomenon, with some implications for the market, particularly in the short term, has been the surge in terrorist attacks and politically motivated sabotage acts against energy infrastructure in several MENA oil and gas producers. We distinguish these kinds of attacks from the general, political instability, which primarily influences a government’s decision-making ability and its capability to implement projects, although the two factors are strongly interrelated and are mutually reinforcing. Non-state actors opposed to current government policies have, in the past, seen oil and gas infrastructure such as production plants and pipelines as a convenient target. An attack on such
infrastructure can disrupt government revenue streams while drawing outside attention to issues of local discontent – such as an inequitable distribution of oil and gas export revenues between regions,
tribes, ethnicities, or sects.” – Oxford Institute for Energy Studies, March 2014: The Arab Uprisings and MENA Political Instabilities.

Current production:
Even in peacetime, obtaining accurate production numbers can be difficult. Under civil war conditions, sources are few and numbers are more guesswork. Because there are often competing numbers from different sources, we have recorded both peak (or highest reported numbers) and an estimated average. Based on open sources, and reports from field personnel, Libya is currently producing oil from only the following fields:


El Feel (Elephant) 78,000 bpd
Intisar 40,000 bpd
Amal 15,000 bpd
Al Wafa 30,000 bpd
Sarir & Messla 180,000 bpd
Al Sarah 30,000 bpd
Bouri 39,500 bpd
Al Jerf 40,000 bpd
Bahr Essalam 30,000 bpd (condensate)

Source: Platts, MEES, Libya-today.com, Websites from Operators, Libya-businessnews.com, Openoil.net, libya-analysis.com

Future production increases will occur as power is restored to oilfields, terminals are by capable actors allowing multiple fields coming online, and pipelines repaired. Based on location of oilfields, previous production, previous history recovering from security shutdowns, and speed of pipeline repairs we have made three cases of bringing production back online. We’ve presented these as optimal, average, and poor cases.

Optimal: Production will come back online at the same speed it did after Arab Spring shut ins. This is unlikely to happen, as the conditions are very different.

Likley: Some infrastructure damage must be repaired. Operators must invest in infrastructure and the security must be hardened to keep said infrastructure on line.

Poor: Islamic extremism remains rampant and wide-spread, and civil conflict between governments continues or intensifies. Operators are only able to get partial oilfields back online, and do so sporadically. Fields and assets that are put online may be lost again without costly security measures.

OFI prediction of Libya Oil2





PDF of this report:




SAUDI ARABIA: All Eyes on The Kingdom – A Look at Oil in 2015

Where We Are:

For the last three months, the quick and steady decline in the price of oil has dominated newspaper headlines the world over. Saudi Arabia’s unwillingness to cut production, despite resounding pleas from allied and unfavorable nations alike, left crude oil sliding from over $100 USD (where it had hovered for the past two years) to under $50 in a matter of months. Countries with policies banking on expensive oil (like Venezuela, Russia and Iran) felt the sting of dropping prices the most. Companies in other up-and-coming oil producing nations (like the United States, for instance) with similar policies came next. As the price slid, different fields, geographies, and technologies in the oil extraction industry held their breath as their profitability came into question.

Despite being worrisome for a few producers, Saudi’s refusal to curtail production in the face of such pointed requests and vocal media felt like it was going “according to plan”, whatever that plan may be. Statements from princes and ministers abounded on who would do what if the price reached such-and-such a number. Mostly, the message was the same: we are not cutting production, period. Empty rhetoric from heads of state in Iran and Russia demonized the falling prices as a globally detrimental conspiracy. Venezuela all but cursed the Kingdom directly. And still, the Saud’s calmly pumped the price steadily downward.

This price-fall was beneficial for Saudi Arabia directly in two ways.

Firstly, oil at $60 per barrel or less renders much of the United States’ recently discovered production in the shale fields of North Dakota, Texas and elsewhere, economically unfeasible. Lifting cost (or, the final price per barrel a producer must pay to extract it) for these newer fields is much higher, and relies on oil prices remaining high for projects to remain profitable from beginning to end. This will drive down the daily production capacity of the US, which had just passed Saudi Arabia as the world’s largest producer in July of 2014, as well as exploration. Cooling the US oil industry helps Saudi in maintaining their position as the world’s “price-release valve” for a few more years, at least.

Secondly, Saudi benefited from cheaper oil by slashing the budgets available to Iran and Russia for actions in Syria, Yemen, and other areas of interest to the Kingdom. Both Saudi Arabia and Iran have been known for playing contrasting roles in their surrounding nation’s affairs. By driving the price of oil down directly, Saudi hurts their regional adversary by limiting their financial resources in proxy conflicts around the Middle East.

The lock-step precision and timing in which Saudi Arabia executed the production policy that sustained oil’s slide also seemed to be in concert with American foreign policy. The US exploration of relations with Cuba (and Cuba’s subsequent rejection of Venezuela’s foreign aid in crude), our Congress’ quiet approval of a bill for defense spending in Ukraine while Russia reeled from lost oil revenues, and our government’s relative silence on the Kingdom’s production policies, have all spelled out tacit approval from Washington. Although oil’s slide has hurt some American producers, the lower gasoline prices provided some relief to a weary American consumer, and US foreign policy makers welcomed a reigning in of Russian influence in Eastern Europe.

Where We’re Headed:

On 22 January 2015, King Abdullah bin Abdulaziz Al Sa’ud, the sole monarch of Saudi Arabia, died of pneumonia related to a lung infection in an undisclosed hospital just weeks after being admitted. He was immediately succeeded by his eldest half-brother, Salman, age 79, who was serving as the Minister of Defense. The designation of Crown Prince was expectedly passed to Muqrin, currently serving as second deputy prime minister and previously as a ranking member in the Saudi intelligence service.

The passing of Abdullah and Salman’s succession of the world’s largest oil exporter has added uncertainty in the future global oil production. King Salman has already announced a plan to stay the course of maintaining current production; the worry isn’t for imminent policy changes. The uncertainty comes in the future succession of the Kingdom: Salman is nearly 80 and is in poor health, himself. He underwent spinal surgery in 2010, has suffered a stroke, and is rumored to suffer from dementia and possible Alzheimer’s disease. The newly appointed Crown Prince Muqrin is the eldest of the last two able-bodied candidates among Saudi Arabia’s founder’s (Ibn Saud) grandsons. Muqrin is not widely known as an effective leader which leaves only a single candidate (Ahmad, former interior minister) in that generation of aiers to the throne.

After Muqrin and Ahmad (both about 70 years of age), the line of succession falls to the next generation of princes. This generation numbers in the hundreds in a variety of upper-level positions of the current government. Former head of intelligence and brief Ambassador to the US Prince Turki bin Faisal, the eldest son of the late Abdullah, and the sons of King Salman are among possible considerations. How this line of succession would look in the event of a string of royal deaths is still difficult to determine, but Abdullah’s passing was a long-awaited shift that allows these political movements to begin taking shape.

Future key indicators for the situation in the Saudi royal family will include cabinet appointments and shuffling (revealing political preferences of the current reigning King), changes in direction of social policies within the Kingdom (as Abdullah was known as a social reformer, this would be a change from the status-quo), proxy-conflict policy shifts (including Syria, Lebanon, Morroco, Libya, Afghanistan, and Muslim portions of eastern Europe), and direct changes in Saudi oil policy.

IRAQ: POTENTIAL Impacts and Effects of Instability in Iraq on Global energy markets

The following is a report generated for an client in late February of 2014. With their permission, it has been republished here as a sample of our custom analytic bulletins:

SUMMARY: Due to regional instability in the security situation in Iraq, there is a large increase in likelihood for meaningful attacks that could result in a civil war in the next (24) months. In expectation of several clearly defined indicators, we believe the likelihood for a civil war in that timeframe is 66%. This will result in approximately (1.2) million barrels of oil per day production capacity to go offline, resulting in an estimated increase of $8.82 in the global price of oil.

DEVELOPMENT: Based on patterns of previous attacks, and an increase in unrest in particular regions of Iraq(particularly Anbar province and surrounding), we expect (est. 80% likelihood) significant attacks to Sh’ia holy sites during Ashura holidays. Such attacks (on Najaf & Karbala, in particular) would strain or publicly end the relationship between Muqtada Al-Sadr, commander of the Mahdi army / Promise Day brigades and a religious and political leader of Iraqi Shi’ite, and Nouri Al-Maliki, the moderate Sh’ia prime minister of Iraq. A halt of these relations would result in effective civil war in the state of Iraq, severely affecting their oil production and distribution capacity. Based on production figures achieved during periods of similar unrest (2006 through 2007), we estimate a decrease of 1.2 million barrels of daily production, resulting in a total of 1.8 million barrels per day production for Iraq.

Internal Iraqi security forces will continue to lose effectiveness to maintain sectarian peace and control over Anbar province and the surround territory. Maintenance issues from an aging pipeline system, withdrawal of foreign oil personnel due to security issues, and minor pipeline attacks will result in a fall in Iraqi production from 3 to 2.7 million barrels per day.

A large scale sectarian attack by ISIS during Ashura in Shi’ite holy cities of Najaf or Karbala will trigger a breakdown of the Al-Sadr-Maliki alliance resulting in vigilantism by Promise Day Brigades. Civil order will break down, sectarian unrest will result in national sectarian violence. Iraqi security forces are already thoroughly penetrated by Sadr’ist sympathizers will conduct reprisal attacks leaving oil and gas infrastructure targets undefended. ISIS will heavily target O&G infrastructure to demonstrate the weakness of the Maliki administration. Pipelines, Salah al Din refinery, Doura refinery, Basra refinery & oil terminal, and Diwaniyah refinery will be under highly elevated risk of attacks. The pipelines should expect repeated debilitating attacks which will reduce output and reliability.

Overall Iraqi daily production will drop from 3 to 1.8 million barrels per day. This will create a geopolitical Risk Premium on oil of $8.82 based on internal models.

Oilfield Intel is Now Accepting Bitcoin

Oilfield Intel will now be accepting Bitcoin at current market prices with Coinbase (www.coinbase.com). All of our products and services will be priced in US Dollar, but will be issuing invoices through Coinbase for payment in Bitcoin upon request. If you have any questions, please contact us at brads@oilfield-intel.com or at (512)784-9758.

Update on Ukraine Pipeline Shutdown Forecast

Ukraine pipeline shutdown forecast 7mar14The Russian invasion of the southern part of Ukraine continues to fester tensions in the region.  We’ve updated our forecast to show developments as we see it:

  • The Ukraine government realizes it can’t get involved in open hostilities. This lowers the chance of war.
  • Racial tensions in Ukraine between Russian speaking Ukrainians and regular Ukrainians do not appear to be fostering violence. This lowers the chance of ethnic civilian attacks.
  • The occupied Crimea shows signs that parts of the populous are pro Russian, but even the Soviet troops are reasonably polite to Ukrainian troops. We’ve encountered stories where soldiers of opposite sides know each other, and clearly don’t want hostilities to breakout. This lowers the chance of war.
  • Russia has begun to threaten to close the gas pipeline based on ‘accounting’ as predicted in our earlier forecast.  This raises a shut down probability.

Overall, the net effect is to lower the chances of a natural gas pipeline shut down to 64%.

Ukraine Gas Pipeline Likely To Be Shutdown – Report

Russia has invaded the Crimea in the Ukraine. This will put Russian natural gas deliveries at risk to the EU as 80% of Russian gas flows though Ukraine.

We have completed our Ukraine report for a client and they have been kind enough to permit us to release it to the public. Under normal conditions our reports are confidential.

The PDF is available at Ukraine Russia March 2014 situation PDF


Ukraine: Potential disruption of Russian natural gas supplies to the EU



Russia supplies roughly 25% of all natural gas used by the EU. Of this amount, 80% is in transmission lines that traverse Ukrainian territory. Due to rapid regional instability in the security situation in the Ukraine, there is a large increase in likelihood of a civil war or a Russo-Ukrainian war in the next (6) months. This will result in a major disruption of at least half the natural gas capacity to go off-line. Resulting in a significant increase in the regional price of natural gas to eastern Europe. The disruption may last from 1 week given a political skirmish to 9 months in the case of a full blown shooting war.

Modeling the political situation with 9 clearly defined indicators, the forecast on a major disruption of the Ukrainian pipeline is 66% as of today.



The Naftohaz UkrainyGazprom pipeline has been a contentious problem since the fall of the Soviet Union. Based on multiple gas shutoffs during the last 24 years, the naïve chances are roughly 25% that a disruption will occur any given year.


Further nationalistic complications have arrived, Russian troops have invaded the Crimea and Kiev is considering it a declaration of war.

We have modeled the current political situation using a Bayesian Network with the important events tracked as the following:


  • A pro-EU leadership comes into power in Kiev.
  • Russia annexes more area than the currently occupied ground.
  • Ukraine declares war and widespread hostilities begin.
  • Ukraine civilians attack Russian speaking civilians with more than 100 casualties.
  • Russian speaking civilians attack Ukrainian civilians with more than 100 casualties.
  • A Ukrainian civil war breaks out with over 1,000 dead per month.
  • Crimea declares independence.
  • Russia announces accounting problems with Ukraine over the pipeline.
  • NATO is asked to assist Ukraine.


Given the loose affiliation Crimea has with the Ukraine, as a semi-autonomous region, we expect Crimea to consider independence under pressure from Russia. Such a declaration enhances the chances of a civil war, but only slightly.

Russia has seized the main airport and blockaded ground routes to Crimea, we assume to impair Ukraine’s ability to transport troops to reinforce local units. Consequentially, Russia can now layout the plan on their timetable and not Ukraine’s.


Internal Ukrainian shaky leadership and questionable combat effectiveness will hamper their ability to deal with the Russian invasion of Crimea quickly. The Crimean support of the Russians will undermine Ukraine’s moral high ground with any international PR spin they put on this. Ukraine knows Russia has superior fighting force and will, this is Russia’s game to play.

Modeling the political situation with 9 clearly defined indicators, the forecast on a major disruption of the Ukrainian pipeline is 66% as of today.


The likely unfolding of events, based on our model is Ukraine declaring a limited war (61%), Russia occupying more of Crimea beyond the 02MAR14 boundaries (81%), Crimea declaring independence (64%), and no internal civil war occurs (22%). Given those events happen there will be a disruption of the Ukrainian natural gas pipeline (91%) for a few weeks.



The unfolding of the worst case would be Russia occupying more of Crimea beyond the 02MAR14 boundaries (81%), Ukraine begins a diplomacy-only option to extract the occupying Russians (20%), Crimea declares independence (64%), Russia propaganda flames past hatred to get civilian Crimeans populations to attack non Russian speaking Ukrainians and it spirals out of what the occupying forces can control. In the north, Ukrainians begin to attack Russian speaking Crimeans and the government becomes inept from infighting. Civil war breaks out (28%). Given those events happen there will be a disruption of the Ukrainian natural gas pipeline (95%) which will take months to repair.


All the probablistic nodes evaluated for the 02MAR14 report

All the probablistic nodes evaluated for the 02MAR14 report

(Probabilities given in this report are conditional to other events occurring and cannot be taken out of context.)

An Update for December!

The past several months have been very busy for us. We migrated a few of our data sources to a new single location to improve some performance issues, established several new data management protocols to streamline various parts of the site, and expanded our inventory and search capacity substantially. We have also begun using a new customer engagement tool, Intercom.io, which will allow our users to directly reach out to us through the site.

We’ve also opened up access to a few more locations for production data, bringing our list to include Alaska, California, Colorado, Louisiana,  Michigan, Montana, North Dakota, New Mexico, New York, Texas, Utah, and Wyoming. Ohio’s on the way!

These data are current to 3rd quarter of 2013 and are being made available in search over the next few days.

We’ve also got our National pipeline data-set to a usable place, and will be integrating it into our mapping features over the next few weeks. We’re excited to offer over 450,000 miles of pipeline data currently, and that’ll be growing all the time.


National Pipeline Data Set


On the docket for features is the inclusion of a Decline Curve tool which should be added by end of January, as well as a few odds and ends to be added to our mapping package, including a better view of production with mouse-over windows, surface ownership information for context and ownership reference, as well an Area of Interest tool to alert users to updates to any of our data sources in a particular region.

We’ve received a lot of interest in our Estimated Lease Hold data-set, which should become available by February 2014. If you’ve expressed interest, we haven’t forgot you! As soon as the data is available, we’ll be reaching out to you directly.

We appreciate the feedback from all of our beta-testers on what they’ve liked thus far, and are excited to be working on our next batch of updates. We are always interested to hear from folks as to what features they would like to see, and encourage anyone interested to drop us a line, or sign up for a demo.