The effect of low oil prices on hotel industry
 
The effect of low oil prices on hotel industry
28 OCTOBER 2015 7:50 AM

Low oil prices have a negative impact on the hotel sector in states such as North Dakota, where the oil industry is a prominent source of income and employment.

BROOMFIELD, Colorado—The oil and gas boom has been a boon for the hotel industry in several states, particularly rural areas of Texas, Oklahoma and North Dakota. With the drop in oil prices in the latter portion of 2014, oil production plateaued in spring 2015, and STR Analytics noted a slight decline in demand for the United States as well. (STR Analytics is the sister company of Hotel News Now.)
 
Following a full year of subdued oil prices, the number of rigs in operation has declined 56.2% (from September 2014 to September 2015). Consequently, hotel performance has suffered in many of those areas. After enjoying a couple of peak periods in late 2011/early 2012 and mid-2014, in May of this year the number of oil rigs in operation hit a level as low as the trough in 2009 in terms of amount of working oil rigs.
 
As expected, Texas is the state with the largest amount of working rigs, followed by Oklahoma, Louisiana, North Dakota, New Mexico and Wyoming. 
 
The interactive chart below allows you to select or hover over each state to view the historical number of rigs in operation.

 


A valid question at this point is: Does this truly affect the hotel industry?
 
Many people believe that lower oil prices have a positive impact on travel.  
 
However, low oil prices have a negative impact on the hotel sector in states such as North Dakota, where the oil industry is a prominent source of income and employment.
 
With an average of 187 open rigs in September 2014 and only 69 in September of this year, North Dakota has reported a shutdown of 118 rigs within 12 months, with about 70% of those ceasing operation in 2015. 


Sources: Baker Hughes, STR

As the chart above illustrates, there appears to be a relationship between the average number of rigs open and trailing 12-month revenue per available room for North Dakota over the past 15 years. 
 
Using a regression model, it is easier to understand the impact of the number of open oil rigs on hotel industry metrics and to confirm that the relationship between those factors is statistically significant. 
 
One metric that was affected the most was the 12-month moving average for RevPAR. (For data geeks, an r-square score of .97 confirmed this was a reliable model and that the correlation between RevPAR and the number of open rigs is strong.) 
 
To make this easier to digest, take a look the chart below; the closer the predicted values are to the actual values, the stronger the correlation between the variables. In this case, RevPAR predicted by the number of open rigs is similar to the actual RevPAR.

 
 

The maps below illustrate the locations of working rigs at the previous peak period (May 2012) and current trough (October 2015). (The number of rigs shown in the maps differs slightly from the number in the regression model because the maps were generated from different sources.)

North Dakota
Open Rigs May 2015

Source: Rig Data
 
North Dakota
Open Rigs October 2015
Source: North Dakota State Government
 
As a state, North Dakota has become synonymous with the oil industry in recent years. So, it is probably not much of a surprise that the faltering oil industry has hurt hotel performance in North Dakota. 
 
However, North Dakota is not alone. Areas of Texas, Oklahoma, New Mexico and Colorado are seeing similar trends in performance declines. Should oil prices remain low, it is highly likely that the newly constructed hotels in these regions will become distressed in a short time, and the oil boom will quickly turn into an oil bust. 
 

2 Comments

  • Nat Holland October 28, 2015 6:21 AM

    The correlation of 0.97 is likely due to autoregression of time series data.

  • Steve Hennis October 29, 2015 5:35 AM

    Hi Nat, Yes, autoregression is always something to be aware of when working with time-series data. By definition, this is an issue when the next term can be predicted by some combination of the previous terms. However, in this case we have 2 series of data (RevPAR vs. Date and Rigs Open vs Date) where the date is the commonality between 2 otherwise random variables, and even though the series follow each other across time, there is not regression of a variable against itself. Therefore, since we are using two different variables to find the correlation, autoregression is not occurring here.

Comments that include blatant advertisements or links to products or company websites will be removed to avoid instances of spam. Also, comments that include profanity, lewdness, personal attacks, solicitations or advertising, or other similarly inappropriate or offensive comments or material will be removed from the site. You are fully responsible for the content you post. The opinions expressed in comments do not necessarily reflect the opinions of Hotel News Now or its parent company, STR and its affiliated companies. Please report any violations to our editorial staff.