Q&A with GLJ on Forecasting Commodity Prices
Editor’s Note: While VISAGE rebranded to VERDAZO in April 2016, we haven’t changed the VISAGE name in our previous blog posts. We’re proud of our decade of work as VISAGE and that lives on within these blogs. Enjoy.
Interview Guest: Tyler Schlosser, P.Eng., Chief Commodity Pricing Analyst, GLJ Petroleum Consultants
The GLJ October 1, 2013 commodity price forecast was recently released and is available at http://www.gljpc.com/commodity-price-forecasts. This is the first in what will be a regular quarterly blog on commodity pricing and how VISAGE is used to analyze pricing data when building GLJ forecasts. I sat down with Tyler Schlosser, the Chief Commodity Pricing Analyst at GLJ Petroleum Consultants, to ask a few questions about how he uses VISAGE when developing his commodity price forecasts.
VISAGE: What goes into building a price forecast?
GLJ: In commodity price forecasting, we examine the “fundamentals”, things like regional supply and demand, infrastructure development and geopolitical pressures, and the “technicals”, which include recent pricing trends, comparative ratios and differentials, historical averages and mean reversion tendencies. We use VISAGE to analyze the “technicals”.
What are the most important commodity pricing benchmarks to follow?
An oil or gas “benchmark” is simply a product that serves as a reference point for the many other available products. The primary world oil benchmark is Brent crude, which is a light, sweet crude located in the North Sea. In North America, the primary oil benchmark is West Texas Intermediate (WTI) crude, and the primary gas benchmark is Henry Hub. Both are heavily traded in the form of futures contracts and have price quotes freely available online. There are contracts specifying delivery every month for the next several years, with a unique price for each.
Prices for other products in North America, including those in Canada like Western Canadian Select (WCS) oil and AECO gas, are often expressed relative to these primary benchmarks. VISAGE allows us to see a daily history of all this benchmark price data in a single plot, like the one below for Henry Hub natural gas, binned annually.
What role do historical prices play in forecasting future prices?
Historical prices are vital to generating forecasts of future prices because they tell us the value that market participants were actually willing to buy/sell for in the past. Although historical price patterns will never be exactly repeated in the future, they provide an excellent starting point in understanding and predicting price behavior going forward. This can be especially true when comparing the price of one product to another – barring a major change to market fundamentals, the price ratio between two products is often relatively stable with a tendency to revert to a comfortable average value. For example, VISAGE can quickly display the ratio between any two price series, and shows us that the price of light crude at Cromer, MB was fairly stable at 98% of light crude at Edmonton, AB between 2010 and 2012. So, unless a significant disruptive fundamental change is expected, a reasonable forecast for 2013 is that the ratio should probably be close to 98%.
So far in 2013, this logic has served us well, since the average ratio so far has been 99% – a very close match.
For comparison, the differential between WCS and WTI crude (below) is much more volatile, and may require consideration of seasonality and current heavy/light crude fundamentals to arrive at a higher confidence near term forecast.
What happens when a forecast is wrong?
As with the weather or the stock market, it is very difficult to predict future commodity prices, and one should be wary of anyone claiming to be highly confident about the future. Actual prices will vary from the forecast price almost immediately. If a forecast appears to be highly accurate, it will likely only be that way for a very short time. However, it is important to note that while perfect accuracy is unattainable, a forecast can and should still be realistic and internally consistent.
GLJ publishes commodity price forecasts every quarter. After a quarter has elapsed, if the previous forecast deviated too far from expectations, there are two options: change it, or stick with the original forecast. New information becomes available every day and unpredictable events can shift reality relative to prior expectations. In commodity price forecasting, a common practice is to adjust a near term forecast based on new information to ensure a close match with the current reality, but leave the long term forecast unchanged if the big picture fundamentals haven’t changed. Below, VISAGE shows us the actual current price data for the WTI-Brent differential, along with a previous GLJ forecast (red) and our updated forecast (black). In this case, our forecast for Q3 2013 turned out to be more pessimistic on WTI prices relative to Brent prices than reality has shown. To better match current and expected pricing, we’ve modified our Q4 2013 forecast based on new information from the last quarter and left the forecast unchanged after that.
For more information on GLJ Petroleum Consultants and their commodity price forecasts, visit http://www.gljpc.com.
In case you missed our last blog series:
- Drill Bits & Bytes Part 1: Curious about drilling activity in Western Canada
- Drill Bits & Bytes Part 2: The Decline of Gas Drilling in Western Canada
- Drill Bits & Bytes Part 3: A Shift in Western Canada’s Gas Production
- Drill Bits & Bytes Part 4: Oil Drilling Activity in Western Canada
- Drill Bits & Bytes Part 5: Western Canada Oil Keeps Rising
Thanks for reading. I welcome your questions and suggestions for future blogs.
Some other blogs you may find of interest:
About VISAGE – visual analytics for the petroleum industry
VISAGE analytics software equips operators and analysts in the petroleum industry to make the most valuable and timely decisions possible. VISAGE brings together public and proprietary oil and gas data from multiple sources for easy to use interactive analysis.