September 2011, Vol. 238 No. 9


How Weather Reporting Affects Petroleum Prices

Dr. Jim Siebert, chief meteorologist, FOX26, Houston

It has already been an historic year in 2011 for weather disasters. The National Climatic Data Center (NCDC) is reporting that the economic cost from these events already surpasses $32 billion and the height of hurricane season is approaching. Typically, the year-to-date total is near $6 billion. The damages come from a combination of major weather events including severe weather outbreaks, winter storms and flooding.

These storms and weather scenarios impact natural gas and oil markets. Prices are obviously affected by pipeline system damage, but markets can also be affected simply by how the storms and subsequent damage are reported. Even something as simple as an altered forecast can create a swing in market prices.

Many energy traders and hedge funds anxiously await the weather computer models that forecast temperature and precipitation. Millions of dollars are made or lost depending on the timing of when the traders get the model data. If a trader is able to get the inside track that the Global Forecast System Model or (GFS) is about to run much warmer than a previous run, then that trader has a great advantage over others who are still awaiting the information.

Most days the markets fluctuate immediately after models are disseminated. Quite often, there will be “knee jerk” reactions when the models swing either much warmer or colder than before. Over time, as the models merge into better agreement, trading is less volatile and the markets tend to follow the trend and the swing is more or less averaged out. The accuracy of the models is not as important as the perception of them or the forecast by the users that have an effect on market prices for both oil and natural gas.

Meteorologist Dave Melita of Melita Weather Associates was among the first to forecast for energy trading entities in the early 1990s, and has seen first-hand how the timing of weather information is crucial. His particular specialty is seasonal forecasting for extended periods. Melita says, “Timing of weather forecasts is of critical importance over every time scale from intraday to several months lead time. This is because a weather forecast that consists of “new” information, or a markedly different temperature expectation compared to an earlier forecast, can and often does move energy markets substantially.”

This concept is even more pronounced when dealing with the release of storm forecasts. The trader who gets the report first will be able to move on their position in advance of their competition. In the short term, it may not matter if the weather report is accurate. The markets are going to react. For a hypothetical example: a hurricane is moving into the northern Gulf of Mexico. It has not made landfall yet, but the latest forecast calls for the storm to approach the Louisiana coastline. That scenario could affect natural gas and oil delivery as well as pipeline systems inland.

Hedge funds and traders are searching for any forecast data that will help strengthen their trading position and they are looking at computer models, talking to meteorologists and even watching the Weather Channel. They anxiously await the release of any new weather data. There is a high likelihood that prices are already on their way up as the markets close for the day. The evening run of the forecasting models indicate the storm is going to move off to the west and away from infrastructure, but the models are wrong.

However, nobody will know it (unless they are an excellent forecaster themselves) at least until the next model run or forecast advisory. In this case, the next model package from the early morning also indicates the storm will move away from sensitive areas. In reality, the storm will continue to present course, natural gas delivery is going to be affected and most people will not know it until the models start to correct themselves later after the markets have opened.

At the opening of the markets that day there is a good chance prices will fall until the model correction. The markets may over-react and a price spike can occur when it’s learned that the storm is going to be bad. All of this happens simply based on the fluctuation of the models. How valuable would it be to a trader to know in advance how the models would move back-and-forth?

“This type of fast market response typically occurs each day that the GFS computer model radically changes its forecast,” according to Melita. “Since the hour 12 GMT run of this computer model is distributed to the market during late morning hours when markets are open, this can cause energy markets to move up or down within a minute or two of release.”

The perception of what the models may forecast can be as important to moving prices as the actual verified forecasts. The effects can be multiplied when dealing with longer range or seasonal forecasts. “This market moving affect is most pronounced when the forecasts are considered reliable, and are predicting either a significantly colder winter or hotter summer that will generate higher energy demand than the current market consensus,” said Melita.

Timing is a critical element that regularly alters the markets, whether the forecast is for an extended period or the current day. Meteorologists are as concerned about being the first to relay important weather information as the traders are to get it. This is true for private forecasters, government meteorologists and even broadcast meteorologists. Literally a few minutes can make a difference and the data doesn’t even need to be accurate. This can be explained by a recent weather event in Houston.

A strong tropical wave was moving across the western Caribbean and was expected to strengthen into a tropical depression or tropical storm as it swept into the Gulf of Mexico. The forecast called for whatever would form to possibly make landfall either on the Texas or Louisiana coast. The storm had not formed yet, but at 2:50 p.m. Central Daylight Time a local television affiliate station sent an e-mail and iphone alert stating that Tropical Depression #4 had formed. Soon after, the Houston Chronicle daily newspaper sent a similar alert with the same headline. However, it wasn’t correct.

Even of more concern was the National Hurricane Center’s never issuing an advisory about a tropical depression. The system went from being a tropical wave, skipping the depression status, straight into Tropical Storm Don. The markets were still open and assuming traders were paying attention (most traders have multiple screens open on their computers, and they subscribe to multiple e-mail, texts, and other online alerts), they had 10 minutes to trade with that information.

Why is it so important to know exactly when a depression or storm forms? The timing of when a system becomes better “organized” can have a huge influence on how strong it may become, where it might go due to its size, and consequently how much an impact it may have on lives, property and the prices of natural gas or oil. In this case, the fact that the disturbance skipped the depression status brought critical implications for Don to possibly become a hurricane as it was moving over extremely warm waters of the Gulf. This did not happen, but the perception that this could be a serious storm was already heightened.

At 3:32 p.m., another alert was sent out retracting the first claim saying there was a slight misunderstanding due to “internal government website” information. The Houston Chronicle followed suit and the headline was changed, but now the markets were closed. At 3:53p.m. the National Hurricane Center released the first advisory on Tropical Storm Don.

It is rather difficult to verify if this sequence of events had a significant effect on natural gas or oil prices, but there was some movement on prices. The point here is the markets could be affected by these kinds of reporting events and the timing of the release of weather information is critical.

Storm damage reports after a major weather event can also have a dramatic influence on the markets. Within minutes of a storm, pictures, video and live interviews saturate news programs, websites and even You Tube. These images can lead to a false sense of the amount of devastation simply because the visuals are so graphic. Emotional reactions alone can have an impact on oil and gas markets. For example, how was your perception of Hurricane Katrina altered after you saw dramatic images?

While the images of these events take only seconds to disseminate, it can take much longer to receive official storm damage reports which contain more facts from which to properly judge how long it may take to recover. However, at times, these reports may not be fully reliable and may not have enough critical data necessary to make good trading decisions.

The NCDC is a nerve center for receiving and posting damage reports. However, their website includes a disclaimer page indicating that the reports may contain inaccuracies beyond their control. It reads, “Data contained in Integrated Surface Data and other NCDC datasets are received via numerous sources. Though a great deal of quality control is performed, some errors will remain in the data.”

The numerous sources consist of damage reports from local National Weather Service offices, storm damage assessment teams, SkyWarn spotter networks and others. The reality is there can be many holes in damage reports and it is difficult to know when a fully trustworthy report may be issued. In many cases the best option may be to have an eyewitness report from a trusted source onsite.

This article barely scratches the surface on the possible implications weather reports can have on market prices. Weather will continue to have an immense impact on trading. It doesn’t matter if it is real or perceived weather events, the markets will react. Whenever there are storms resulting in damage, timing issues on the release of forecast data, or simply gaps in damage information; somehow prices will be pushed either up or down.

When it comes to weather reports – including forecasts, storm events, tropical advisories or damage reports – it may be more important to ask the right questions about the weather data itself, including questions concerning its reliability, the degree of possible misperceptions and how the markets may respond.

Dr. Jim Siebert
is FOX 26 Chief Meteorologist in Houston. He can be seen at 5 p.m. and 9 p.m. He holds degrees in meteorology/earth science from Baylor University in Waco, communications from Brigham Young University and science education from the University of New Mexico. He teaches both meteorology and geography at the university level. He also has published research and presented his findings at national conferences. One area of focus for his research has been in the implementation of new graphics technologies for broadcast meteorology.

Related Articles


{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}