This paper highlights several ways to stress testing companies within the energy industry, particularly with regards to impact of COVID-19 and the current oil price shock. We have conducted a statistical analysis in order to assess the impact of persistent low oil price on Q2 2020 Probability of Default (PD) in the energy industry. In addition to this paper, S&P Global Market Intelligence is also working on the creation of an oil price scenario tool factoring in various aspects from a supply and demand perspective.
Assessing the Impact of COVID-19
The impact of COVID-19 on the energy industry is unprecedented. According to the International Energy Association (IEA), in the first quarter of this year, global demand had decreased by 2.5 million barrels per day.[1] Aside from the price war impacting supply, demand has also suffered. An example of this is the drop in fuel demand from airlines globally due to government-imposed travel restrictions designed to curb the spread of COVID-19.
Governments have also tried to relieve the impact of mandatory business closures by using monetary and fiscal stimulus. However, a negative impact will likely be felt. As of March 31, 2020, S&P Global Ratings downgraded approximately 340 corporate entities globally and put another 370 on Negative Outlook or CreditWatch with negative implications. Almost 80% of these ratings actions apply to entities in the speculative-grade category.[2]
One of the industry approaches used to measure the potential impact of market events like the oil price shock, and COVID-19 on a company’s creditworthiness is the adjustment of annual revenue figures by simply dismissing an entire revenue tranche. In other words, completely erasing the revenue from a given quarter while keeping all other financials the same (e.g. same figures from the previous year quarter). One of the major assumptions here is that the fixed costs remain the same while the variable costs are reduced with the revenue decline.
2014-2016 Oil Price Crash
The recent oil price shock was triggered by Russia and Saudi Arabia’s oil price war. The two countries failed to agree on coordinated production cuts following diminishing demand from the biggest oil importer, China, due to COVID-19 pandemic. Oil prices fell dramatically, leading the oil industry into uncharted territory, and a number of energy companies are likely to be negatively impacted. Although we are in an unprecedented situation with regard to the energy industry, the previous oil price crash can provide a proxy to stress the financials.
In 2016, oil prices declined by 70% from their peak in 2014. There were a number of factors that contributed to the decline in oil prices during 2014-2016 period, including: China’s slowing growth rate impacting demand, and efficiency gains in United States (US) shale oil impacting supply. In the current 2020 downturn, oil company costs are comparatively lower and annual cash flow after dividends may remain positive for many large companies, but debt is still relatively high compared to the 2014-2016 downturn.
Figure 1: Average Annual Crude Spot Prices
Source: United States Energy Information Administration, as of April 6 2020, retrieved from: https://www.eia.gov/dnav/pet/pet_pri_spt_s1_a.htm
S&P Global Ratings base case assumptions for Brent oil prices are: $30 per barrel for the remainder of 2020, $50 per barrel in 2021, and $55 per barrel in 2022. Working on the base case assumption of $30 per barrel for 2020, this represents a drop of 53% from the 2019 average Brent spot price of $64 barrel.[3]
The observed change in the financials due to the oil price drop between 2014 and 2016 is also utilized as a stress testing approach. For example, between 2014 and 2015, the average crude spot price change for Brent was 48%. When looking at a company’s key financial metrics change over the same period, one could determine the deterioration change percentage. This change can be used to stress the input factors required for the Credit Analytics scoring models. This approach could also be combined with the revenue-erasing approach to better reflect the current business environment.
Impact on the Probability of Default
Assessing the full extent of any relationship between oil price movements and the subsequent stress on financials requires substantial time and resources. With the purpose of providing a high level and timely view, we conducted this quantitative analysis to test for a statistically significant relationship between the oil price change and the change of the PD in the energy industry.
The study was performed on quarter-over-quarter (QoQ) changes of the last twelve months (e.g. for Q1 2019: January 1, 2019 to March 31, 2019) PDs between Q1 2011 and Q4 2019. The QoQ change of the PD has been analyzed with respect to QoQ changes in oil price (oil price quarter value is calculated as the average daily value of the three consecutive months within the quarter). This study focuses on the results over a short time horizon, in particular the oil price change effect on the PD within the same quarter.
The sample data selected for the analysis consists of public companies in the energy industry with the PD pre-scored by PD Model Fundamentals (PDFN) between 2011 and 2020. We used logistic regression to regress PD on QoQ changes of oil price and the model parameters were trained via maximization likelihood estimation. The sole explanatory variable is the QoQ change of the oil price and the dependent variable is the PD in the same quarter. The PD in the previous quarter is also added to the model as a fixed feature. The QoQ oil price change is defined as the percentage change of the average oil price from one quarter to the next one. The PD is calculated by using Last Twelve Months (LTM) financials and PDFN.
Figure 2 shows the analytical process and the results structure. Two different test cases were created to maximize the statistical significance: 1) all companies (i.e. defaulted and non-defaulted companies); 2) non-defaulted companies only. We find that the majority of defaulted companies have a PD at the speculative grade level prior to default.
Figure 2: Analytical process of the study
Source: PD Model Fundamentals,S&P Global Market Intelligence, data as of April 1, 2020. For illustrative purposes only.
With the model trained on the historical data, one can apply the oil price forecast to predict the PD in Q2. In Figure 3, there are three suggested scenario cases where, after Q1 of 2020, the oil price falls either at $20/$30/$40 per barrel for Brent or $15/$25/$35 per barrel for West Texas Intermediate (WTI) in Q2, where the average Q1 oil price is $50 and $45 for Brent and WTI, respectively, as of April 1 2020. The results are shown for different groups, under Brent and WTI oil price, including and excluding defaults. These two cases represent different levels of vulnerability for investment grade and speculative grade companies. The impact of the oil price changes is more material for more vulnerable companies. Note that the change in PD is not linear and constant as logistic regression was utilized. The values in Figure 3 represent the approximate percentage change in PD with respect to the PD level in 2019.
Figure 3: Proxy impact of oil price levels on the PD under different scenarios
Note: further information is available upon request.
Source: PD Model Fundamentals, S&P Global Market Intelligence, data as of April 1, 2020. For illustrative purposes only.
In summary, various unique approaches are being conducted towards stress testing, given the exceptional nature of the circumstances. To comprehensively understand the impact on PD, a combination of these approaches may be required to factor in the various dynamics at play.
At the time of authoring this paper Saudi Arabia and Russia among other oil producing nations reached an agreement on production cuts. However, there are many concerns about the drop in oil demand, and it is hard to gauge the extent of this drop in demand as different countries experience different periods of ‘shutdown’ during the COVID-19 pandemic. Frequent stress tests, like the ones we’ve suggested in this paper, can be helpful to keep pace with the changes.
Our Credit Analytics products provides analytical tools and solutions for scenario and stress testing when running risk analysis on rated, unrated, public, and private companies.
[1] Iea. (2020, March 1). Global oil demand to decline in 2020 as coronavirus weighs heavily on markets – News. Retrieved from https://www.iea.org/news/global-oil-demand-to-decline-in-2020-as-coronavirus-weighs-heavily-on-markets
[2] “COVID-19 Weekly Digest”, S&P Global Ratings, April 1, 2020.
[3] “Harsh Downturn Prompts Rating Actions On Multiple European Oil And Gas Companies”, S&P Global Ratings, Redmond et al., March 25, 2020, retrieved from RatingsDirect® on the S&P Capital IQ platform, https://www.capitaliq.com/CIQDotNet/CreditResearch/SPResearch.aspx?DocumentId=44345397&From=SNP_CRS
Read it from SPGlobal – Photo as posted on SPGlobal