Description

At current trading levels, I believe the CVR Energy bonds are attractive. The 5.75s of 2028 traded today at 93 cents for a yield of 7%. I realize this is no man’s land on VIC because it’s not an equity, preferred or a distressed bond, only a coupon clipper. However, the bonds provide a good total return opportunity. By this time next year as the recovery continues, I expect the yield to be inside of 6% for a return of nearly 12% when factoring in the coupon. Debt is low when excluding cash on hand, 2021 expected CARES Act recoveries and the value of Delek (DK) shares, and then you get the two refineries and other assets. Eventually, I think there will be a catalyst to drive bonds higher even if credit spreads don’t tighten. The plan is to acquire an asset, grow the footprint and then sell to a bigger player.

Now let me get the rub out of the way first. The backdrop is very challenging with demand declines from Covid and excess supply. The refineries are located in PADD II and profitability is linked to a structurally tighter Brent-WTI differential. Additionally, Icahn is the majority shareholder (71% stake) which many would not interpret as a particularly bondholder friendly fact. However, I would make a few comments on the last point. Despite sizable cash, the company cut its dividend to zero and on its most recent call, management made it clear they don’t intend to lever up above 2x. Even assuming no cash, leverage is reasonable.

A brief overview: The focus here will be on the Petroleum Segment but CVR Energy also owns a Fertilizer Segment via its 35% stake in CVR Partners (UAN).

The two refineries are located within the Mid-Con in Kansas and Oklahoma (100-130 miles from Cushing). Total nameplate capacity of 206,500 bpd is split between the full coking, medium-sour 132,000 bpd Coffeyville unit and the smaller 74,500 bpd Wynnewood plant. The units have an average Nelson Complexity ratio 10.8. Gathering assets provide a crude sourcing advantage due to its proximity and access to the Anadarko Basin. "Midstream” or logistics assets (pipelines and storage) historically added $70-$80 million of annual EBITDA per management. Despite onshore production declines, the company continues to gather 120,000 bpd and expects to be able to source additional barrels within the basin to offset overall declines. Space on the Keystone and Spearhead pipelines provide access to Canadian crude – more on this later. Takeaway flexibility allows the company to access major cities, such as Oklahoma City, Kansas City and Tulsa. Products are split between gasoline (54%), distillate (41%) and pet coke, NGLs, slurry, sulphur and gas oil (6%).

The capital structure here is straightforward. Besides a small amount of capital leases ($62 million), the only other debt consists of the two senior bonds – 5.25% due 2025 and 5.75% due 2028 aggregating $1 billion. Cash is $624 million and CARES Act recoveries should be $50-$75 million. Additionally, the company owns 10.5 million DK shares worth $163 million. As such, I calculate net debt of ~$210 million. Total shares of ~100 million at the current price results in $1.5 billion of equity market cap beneath the bonds.

So the backdrop stinks. A lack of jet fuel demand is causing diesel inventories to swell and gasoline remains off from 2019 levels. Planned supply additions in Asia and the ME are another overhang.  It’s getting better if you listened to PBF Energy’s bondholder call yesterday as forward cracks to recover and 2.75-3.25 million bpd of capacity were mothballed or temporarily shut.

Beyond this, CVR has several other positives.

1. A major turnaround at Coffeyville was completed in 1H 2020 at a cost of ~$154 million and the next turnaround at Wynnewood isn’t until 1H 2022. The cost of this turnaround is expected to be $80-$100 million but management indicated they could simplify the refinery to cut the cost by 20%-30%. Therefore, 4Q20 and 2021 turnaround spending will be limited.

2. A renewable diesel project is planned at Wynnewood to convert a 19,000 bpd hydrocracker to 100 million gallons/year of washed and refined soybean oil processing capacity to produce renewable diesel and naptha with an in-service date of 6/30/21. The estimated cost of $100 million would be quickly recovered via the Blender's Tax Credit of $1/gallon that continues through 2022 and possibly longer with Biden in office. Furthermore, the project generates Low Carbon Fuel Standard credits and RINs (1.7 D4 biodiesel RINs per gallon produced). The last point is key. RINS are a major impediment to CVR’s capture rate. On a pro forma basis, CVR’s RINs obligation would decline from 250 to 80 million. For example, the cost of RINs in 2020 is expected to be $110-$115 million. Other renewable diesel projects at competitors CVR’s should cause RINs to decline further.

3. Potential PADD IV acquisition. A purchase of Delek seems unlikely as the CVR is unwilling to pay a big premium and Delek’s management doesn’t seem interested in selling (i.e. didn’t respond to CVR letter and put in a poison pill). However, an acquisition in PADD IV, while using cash would add another set of assets and diversify the company away from PADD II. As mentioned earlier, the plan is to bulk up and then sell the entire company.

4. Access to WCS crudes. The company has pipeline access for 35,000 bpd of Canadian crude. This is apportioned to 30,000 bpd. Every $1 increase in the WCS-WTI spread is worth $11 million to CVR. For example, looking at the 2021 curve, this is worth >$50 million to the company at a spread of $14/bbl net of the pipeline tariff of ~$6/bbl and $2.50-$3.00/bbl quality discount.

5. The cash balance plus CARES Act recoveries and DK shares is a positive but the company also expects to recover additional WC in 4Q especially with higher crude prices.

As mentioned above, total debt is ~$1 BN (slightly higher with cap leases) and net debt is ~$210 MM. CVR’s cracks are linked to Group 3 2-1-1 within PADD II.  Over the past 5 years, these margins have ranged from $9/bbl (in 2020) to almost $19/bbl, with an average of $15/bbl. Capture rates swung from 56% to 84% and averaged in the low 70s. Over the past 10 years, EBITDA ranged from this year’s basically zero to as high as $1.2 billion. Over this timeframe, the average EBITDA was slightly below $600 million. Prior to 2020, the low point was $223 million in 2016 (Brent-WTI of $0.80) and this obviously doesn't capture the benefits of the renewable diesel project. As mentioned above, the key differential here of Brent-WTI is impaired relative to prior years due to new pipelines, reduced shale production and low prices. The days of high single digit differentials have been replaced with $2.50-$3.25 spreads per the the futures curves.

The continuation of Covid into 2021 pushes out “normalized” EBITDA as does the partial year for the renewable diesel project. For this reason, I look towards 2022 to build up an estimate of a more “normalized" EBITDA. I assume Group 3 cracks of ~$12/bbl and an improved capture rate due to lower RIN costs. Including estimates for the renewable diesel project, I get to EBITDA of ~$400 million. If we assume the blender’s credit is cut by 50% post 2022, this EBITDA declines to $350 million. In either case, net leverage is <1x. If we assume that $400 million of cash is used for a PADD IV investment at 5x, this implies incremental EBITDA of ~$80 million, or $430 MM ($350+$80) of EBITDA for PF net leverage of ~1.5x. Either way, I think the bonds are well-covered as the sector trades at 4x-5x 2022E EBITDA. Note: My leverage estimates assume 2021 cash from operations covers capex including the renewable diesel spending.

Adding it all up, I think a year from now the bonds tighten to <6% (mid-Feb 2020 levels). This results in 6 points of capital appreciation and when summed with the coupon of 5.75 gives the investor a return of nearly 12%.

Catalyst

-Vaccine distribution 

-Crack spread recovery 

-Renewable diesel plant start-up

-Acquisition