THE EUROPEAN CENTRAL BANK reviewed 9 different models for oil price forecasting and concluded that there is no single model that could give a good prediction. The ECB concluded that a combination model provides a better predictive power. The ECB recommends a combination of four models: (i) Futures; (ii) Risk-Adjusted Futures; (iii) BVAR Model and (iv) DSGE Model. See attached file.

BACK TO THE DRAWING BOARD. If there are too many models, it means that there might not be any concrete answer. If there are so many forecast models and none seems to have the final say, then this area is still a target rich field for research. There is no final answer in the forecasting model for oil price. There is still room to grow in the research literature.

THE EUROPEAN CENTRAL BANK reviewed 9 different models for oil price forecasting and concluded that there is no single model that could give a good prediction. The ECB concluded that a combination model provides a better predictive power. The ECB recommends a combination of four models: (i) Futures; (ii) Risk-Adjusted Futures; (iii) BVAR Model and (iv) DSGE Model. See attached file.

BACK TO THE DRAWING BOARD. If there are too many models, it means that there might not be any concrete answer. If there are so many forecast models and none seems to have the final say, then this area is still a target rich field for research. There is no final answer in the forecasting model for oil price. There is still room to grow in the research literature.

Low: Prices can touch a low of $30. This low would be achieved because Saudis in an about turn have reason to drive some shale producers out of business as US supply has hit long term demand in both US and China. China will support this move as long as stockade capacity lasts in China as they store up for more expensive times ( They have stopped buying in the intervening rally from $44 

Modelling: It would be interesting to see and probably it would be entirely new models that emerge in this spell. Markets will however hold the following as sacrosanct in each model:

Costs of producing US shale (Marginal) $45-50 Plus long term investments and overcapacity in shale(likely) 

In Econometric/Fundamental based modeling:

Decreased capacities and absolutely no substitution from LatAm(Brazil/Venez/Colombia) or Algeria Iraq and Iran

Reduction in growth of Green Energy plans with increase in basic Crude factor consumption esp in India China and the USA

Agreed, augmented by Chinas desire to move to natural gas...  I have heard estimates of 'rock bottom' Canadian oil sands costs in the $20s.  Does anyone think $25.00 for WTI is possible for a short period of time?

See:

Brent and WTI forecasts by US EIA - http://www.eia.gov/forecasts/steo/

WTI futures - http://www.cmegroup.com/trading/energy/crude-oil/light-sweet-crude.html

Jennifer -

As Frank inferred, the US EIA is the correct source for US energy data, but having worked there for about a quarter of a century until my retirement last August, I can tell you that you should look for all metadata you can, and question all assumptions. Modeling, as Paul noted, is always a concern for forecasting, but the relatively short term forecasting you seek here is not nearly so bad as longer term forecasting. However, to this I would add the concern of uncertainty in the input data. I have seen people use models where they implicitly assume 100% accuracy for the input data when there is already likely more inaccuracy there than may be hinted at for the model output! The EIA does publish some technical/statistical notes, and similarly titled appendices to the now online successors to the old printed reports. These documents are pushed into the background as ever more superficial online 'news'-like content seems to fill the EIA webpages, dominated by an office that takes what statisticians produce and then has their nonstatisticians purposely 'dumb it down.' The result is often overstated accuracy, reported by an office that really does not understand the subject matter or statistics. So ... question everything.

On the bright side, for your purposes, although I did not work directly on forecasting very much, it was my impression that the STEO model was fairly well vetted. However, beware of the possibility of bureaucratic mismanagement with the input and/or output. There are too many people involved who really are clueless. Unfortunately, it will not be likely that you can do better elsewhere. The bad influences on the integrity of US EIA data come for the most part from outside sources. Pressure from private industry that is often even less competent is often to blame. Also, management at the EIA has become more and more concerned with timeliness and less and less concerned with accuracy, when it should have been the other way around. Both timeliness and accuracy are necessary to as reasonable a degree as possible of course, but the balance between the two is not good. Often the pressure to produce more than can reasonably be produced is too great. (This reminds me of this adage: "They wanted it in the worst way, and that was how they got it.")

Sometimes you can obtain raw data from the EIA, though much of it is confidential and thus unavailable. Graphical analyses - such as simple scatterplots - that you might do on the least aggregate data you can find might help you to see data limitations to the STEO. However, I am not sure that much such information is available. I worked mostly on sampling and estimation and not in a forecasting office. As for the STEO itself, you may be able to obtain the code and analyze its strengths and weaknesses for yourself.

One forecasting model with which I did work years ago - gone now - tested to me to be way overfitted so that two-month ahead forecasts became less and less accurate as time passed. But my impression of the STEO was that it was not supposed to be so bad, though there has seemed to have been too much tendency to 'adjust' things in some way at the EIA to get what someone thinks they expected to see, and for all I know, that could have happened here also ... so beware.

With current volatilities, I would not expect much accuracy anyway.

Best wishes - Jim 

Most banks and investors look to the forward strip curves on NYMEX as being a market based and unbiased forecast of prices.  This is also the price that companies can hedge at and lock in prices which is useful for predictability of future returns.  Some banking arrangements require a minimum amount of production hedged based on the forward curves.

I would not give exact number. Except that there is downward trend caused by lifting sanctions from Iran and its desire to expand oil production.

Recently I heard a presentation where the forecast of future oil prices by different consultants and there future revision have been reviewed. Even World Bank or OPEC make errors. That means that prices of oil are just a bit less chaotic than stock exchange index. The reason is that there is not only high price sensitivity to small changes in demand and supply, but also high influence of geopolitical factors and perhaps speculation on paper oil.

If you really want some numbers, I personally believe that the average price in 2015 might be close to $50. Indeed, after expansion of Iranian production price may drop to $45 (this number I read in one of today's newspapers), but then it might be irrational for many producers to suffer such low prices and some new OPEC agreement may follow.

Crude Oil prices are likely to drop a more in 2016 as the increase in supply is expected to be more than demand. 

In the OPEC 2015 December meeting indications were also not clear so it also indicates chances are low

As per Total Boss- the price is supposed to drop more

Another important factor is the global storage of Finished Oil is expected to be full by Q1 of 2016

The edible oil market is also having a down slop 

Based on these info I believe the oil market will be down for 2016

low as $32 and high $48

https://www.investing.com/commodities/crude-oil