Description

Thesis:

NS a midstream MLP limited partner with crude oil transportation and storage assets, refined products transportation and storage assets, an asphalt refining business, and a marketing business. For those unfamiliar with MLP’s, there is a limited partner (LP) and a general partner (GP). The general partner typically maintains a small economic interest; however, the GP’s main source of cash flow is its Incentive Distribution Rights which give it a disproportionate right to cash flow above certain defined per unit levels. NSH is an LLC HoldCo whose only assets are 100% of the NS GP and ~10.4mm NS common units. The NS GP has the following rights to cash flow: Up to an NS distribution of $2.64 / unit, the GP gets ~$0.075 / unit; above $2.64 / unit, the GP gets 25% and the LP gets 75%. See endnote 1 for an example calculation of the NS GP’s current distribution.

NS’s financial results have been suffering for more than a year. In FY 2012 NS generated $220mm of distributable cash flow (DCF) and paid out $375mm in distributions, filling the cash flow shortfall with the sale of  commodity price sensitive assets (50% of its asphalt refining business as well as a small crude oil refinery purchased in 2011). The cash flow shortfall has not gotten much smaller in 2013. In Table 1 , the “Actual” columns show NS and NSH's 1Q 2013 DCF and declared distributions annualized.  For 1Q 2013, NS generated only 66% of the cash flow it needed to cover the current distribution level. There are three options to close this shortfall: 1. Increase DCF / unit, 2. Issue debt to cover the cash flow shortfall and 3. Decrease the distribution / unit to a sustainable level. Before getting to options 2 and 3, let’s evaluate NS’s ability to generate incremental DCF / unit in the near future. Please note that throughout this discussion “DCF” refers to distributable cash flow (EBTIDA less interest expense less maintenance capex less cash taxes plus unconsolidated JV distributions plus/less other cash items).

Option 1 and Business Outlook:

NS has four business segments: storage, transportation, asphalt refining and marketing.  Until 2013 NS reported asphalt refining and marketing results together, so I will discuss them in the same section.

Storage Segment:

Roughly half of NS’s 1Q 2013 DCF came from oil and products storage. This is a great business and NS is well positioned; however, currently demand (and rates) for crude oil storage is down because the forward curve is backwardated (future prices are lower than present prices). This decreases demand for storage as there is no longer value in storing and selling oil at a future date. At the same time, 41% of NS’s leased capacity comes up for recontracting this year.  Recontracting this capacity in today’s depressed market will hinder NS’s ability to achieve its YoY growth projections of $10 to $30mm in 2013, even with contributions from its capital program. In fact, NS’s most recent guidance update projected 2Q 2013 storage segment results to be even lower than 1Q 2013.

In addition, NS’s storage segment includes its crude oil rail terminal business at St. James, LA. This business has been very profitable over the past couple of years, (~$20mm of DCF / year); however, NS needs a large WTI-LLS spread for this terminal to continue to be successful. Not only does this spread drive customer demand for the capacity, but NS also has profit sharing contracts with some of its existing customers. These contracts need a >$7.00 / bbl WTI – LLS spread to be profitable. With the Seaway/Keystone expansion(s) online and coming online, cheap ships from Corpus Christi to St. James (~$2.50/bbl), as well as the Ho-Ho reversal providing more than enough light sweet crude to the Gulf Coast refining market, there is no reason to think the spread will blow out very far past $5.00 - $7.00 / bbl again and therefore no reason to rail crude to St. James. Incredibly, $45mm of NS’s 2013 capital program (~11%) is an expansion of this rail terminal (supported by 1 customer). I believe NS will be lucky to cover its cost of capital on this facility by year end 2013 as marketers look to rail crude to the higher priced East and West Coast demand centers.

Transportation:

NS’s bright spot is its crude oil and product pipeline business which in 1Q 2013 accounted for ~42% of NS’s DCF. NS has good assets in the Eagle Ford, some of which are homegrown pipeline reversals and conversions (therefore relatively cheaper), but the lion’s share of crude oil pipeline cash flow comes from its 4Q2012 TexStar Eagle Ford pipeline acquisition which was quite expensive (>12.0x unlevered DCF multiple) and is projected to barely cover its cost of capital for 2013. The TexStar acquisition will become more accretive in coming years as it gives NS the ability to put more capital to work (addt’l capital projects) in a growing resource play; however, 1Q 2013 TexStar opex was higher than expected and volumes are barely keeping pace with projections (which were very optimistic).

The remainder of the transportation segment consists of refined product pipelines. These provide steady, reliable cash flow; however, refined product supply and demand centers have not changed in decades and are well serviced. As such, this business will provide few near term growth opportunities to help fill the cash flow shortfall. 

Asphalt Refining and Marketing:

NS’s asphalt refining business and its marketing business are not currently profitable. The decline of these businesses is the main reason for NS’s current dire situation. In 2011, the asphalt refining and marketing businesses combined accounted for $100mm of EBITDA. In 2012, these businesses combined produced an EBITDA loss of ~-$20mm due to lower refining and marketing margins.  The outlook for both of these businesses remains negative. In 2012 NS took a $266mm non cash impairment charge on the asphalt refining business, sold half of the business for ~$175mm + working capital (used the proceeds to fill distributions shortfall and pay down debt) and deconsolidated its results. The implied carrying value of NS’s remaining half of the asphalt business after the impairment charge is $52mm. Although the formation documents of the joint venture entity have not been made public, NS indicated in its 2012 2Q 10Q that its new partner will have a distribution preference and liquidation preference over NS in the JV. This explains the difference between the carrying value of NS’s 50% ($52mm) and the purchase price of the sold 50% ($175mm). All signs point to management not expecting much from the asphalt refining business in the future. For the marketing business, management guidance is $20-40mm of DCF for 2013; however, as this business lost money in 2012 and 1Q 2013 and management explicitly stated that 2013 will present marketing “challenges”, I expect the low end of this guidance at best.

2013 Capital Program

Taking a step back from specifics, NS has ~$400mm of growth capex budgeted for 2013. How much of the $132mm 1Q13 annualized DCF shortfall (Table 1 Line 7) can these projects possibly fill? Let’s assume that on average these projects are completed halfway through 2013 and are financed at the same time (as NS doesn’t generate enough cash flow even to cover their distributions, they must finance 100% of all growth projects with debt and/or equity). Assuming a 7.0x unlevered DCF multiple (~14% ROC), they will generate ~$57mm of unlevered DCF (400/7), half of which ($28.5mm), will benefit 2013. Assuming 50% debt and 50% equity financing NS has a blended cost of capital of ~9.1% (50% equity at 11.2%² and 50% debt at 7%³), once again only half of which (.091*400/2 = $18.2mm) will burden 2013. Under these assumptions, the 2013 projects will generate ~$10.3mm ($28.5 - $18.2) to help fill the cash flow hole. If we assume also that NS’s marketing business returns to profitability and hits the midpoint of management's DCF guidance at $30mm, NS will have an incremental $40.3mm to pay its distributions for 2013, still ~$92mm short of 1.00x distribution coverage.

After examining NS’s business prospects, it seems exceedingly unlikely that either an “uptick” in business or incremental DCF from capital projects will be enough to avoid an NS distribution decrease in 2Q or 3Q 2013 as the negatives in the storage, rail, and asphalt businesses weigh down the capital project and marketing contributions.

Option 2:

In order to bridge the gap to better times, NS or NSH could issue debt to cover the DCF shortfall.  I do not believe they will do this for three reasons. First, NS is already heavily leveraged with a debt to EBITDA ratio of ~6.6x and an EBITDA to Interest ratio of ~2.9x. S&P downgraded NS’s credit rating from investment grade (BBB) to junk (BB+) in July 2012. Any increase in leverage would only damage NS’s financial health and be seen by the investor community as a desperate attempt to delay the inevitable.

Second, NSH has negligible debt and could certainly lever up to bridge the gap to better times; however, I believe this would destroy value for NSH rather than preserve it. If NSH attempts to “bridge the gap”, it will have to issue debt every quarter until NS DCF / unit grows enough to support the current level of distributions, paying out the majority of the cash raised to the NS public unitholders. Once that DCF / unit is reached, NSH unlevered cash flow will be exactly what it would have been if it had not issued debt to ”bridge the gap”; however, NSH’s equity value will be lower by the value of the debt that it issued to help maintain distributions. Therefore, a long term investor would prefer to allow the distribution to fall in the short term to preserve greater value in the long run.

Third, William Greehey, former chairman and CEO of Valero, is chairman of the board of NS and NSH and owns 18.7% of NSH. Greehey has continually bought NSH units since its IPO in 2006, especially during downturns in unit price, proving himself to be a long term investor. Therefore, although it would sacrifice near term unit value and distributions, Greehey is incentivized not to issue NSH debt to maintain NS’s distribution level, as illustrated by the previous paragraph. Furthermore, any decline in NSH unit price would present Greehey with an opportunity to buy more of an asset he clearly favors at a reasonable price allowing potentially greater returns when NS returns to distribution growth.

Option 3 and NSH Valuation:

Now to Option 3 and resulting NSH valuation. The "Revised" column in Table 1 shows the effect of reducing NS’s distribution to $2.99/ unit (the level necessary for an NS cash coverage ratio of 1.05x). This ~32% decrease in NS distribution cuts NSH’s DCF by 55% due to the structure of the GP’s right to cash (see endnote 1 for an example GP distribution calculation; this asymmetry is why I recommend shorting NSH and not NS). Line 14 in Table 1 shows the revised implied unit prices of $30.67 and $11.89 (decrease of 32% and 55%) for NS and NSH respectively after reducing the NS distribution, assuming their current distribution yields remain constant. These revised unit prices are directionally correct but simplistic.

To build a more accurate view on NSH valuation we must break it into two parts: the NS units owned by NSH and the NS general partner (see Table 1 for current unit prices and unit counts). The current market value of NSH's NS units is $468mm (10.4*$44.96=$468) implying a GP valuation of $653mm (42.6*$26.32-$468=$653). A GP valuation of $653mm implies a GP distribution multiple of 12.8x (653/51=12.8)⁴. At first glance, a 12.8x multiple seems very reasonable for an MLP GP in today’s market. However, after the decrease in NS distribution to $2.99 / unit, the NS GP distribution falls from $51mm to $15mm⁵. If we generously assume that the value of the NSH owned NS units remains constant after the distribution reduction, then to maintain the current NSH market capitalization of $1,121mm, the GP distribution multiple would have to increase to 43.5x ($653/$15). Pretty optimistic. In today’s frothy MLP market a reasonable distribution multiple for an MLP GP with decent growth prospects might be ~20.0x, giving an implied GP value of $300mm (15*20). This implies a market capitalization for NSH of $768mm (468+300) and an NSH unit price of $18.03 (618/42.6), a 32% decrease from today's unit price of $26.32. Of course, NS’s units will likely also fall in value on the distribution decrease, providing a larger margin of safety.

Sensitivity:

What happens if NS’s results turn out better than expected? Keeping the NS unit count constant, for every $10mm of incremental DCF, $7.5mm goes to the LP and $2.5mm goes to the GP. Assuming 100% of this incremental DCF is distributed (and the NS unit count remains constant), the GP distribution increases by $2.5mm, which at a 20.0x multiple adds $50mm of value to NSH; the NS distribution / unit goes up by $0.096 (7.5/77.9), which at an assumed 6.6% NS yield⁶ adds $1.46 to the NS unit price (.096/.066) and $15mm of value to NSH (10.4*1.46).⁷ The total value increase to NSH of $10mm of incremental NS DCF is $65mm or $1.53 / unit ($65/42.6). Therefore if 2Q 2013 NS’s annualized DCF is $30mm higher than projected, the NSH unit price should still conservatively decline to $22.61 ($18.03+$1.53*3), a 14% decrease from today’s unit price.

Risks

I see two more unaddressed risks to this thesis. The first is the unpredictability of the asphalt business. Historically, asphalt has been a seasonal business, with business picking up significantly in the summer. However, the downturn of the sector over the past couple of years seems to have attenuated this effect as illustrated by NS’s 1Q and 2Q 2012 asphalt and marketing results. 1Q 2012 NS’s asphalt and marketing segment had a cash operating loss of -$9mm. In 2Q 2012 instead of recovering, results actually declined even further to a cash operating loss of $-19mm. Keep in mind at this point NS owned 100% of the asphalt business. Today, with a subordinated 50% stake and another asphalt operating loss in 1Q 2013, it does not seem likely that the asphalt business could improve enough to make a material contribution to NS’s DCF in 2Q or 3Q; however, it is a risk.

The second situation I see that could mitigate the decrease in value of the NS GP and by extension NSH, would be for NS to issue a large chunk of equity in making a large non-dilutive acquisition. Issuing NS equity increases the NS GP distribution because it receives the IDR’s on the new units issued even if per unit cash flow does not increase; however, I do not believe NS has this option. NS’s cost of equity and debt is far higher than its competitors’ meaning it is no longer competitive in the M&A market.

Conclusion

NS has run out of time and money. Although their most recent 10Q shows $116mm of cash on hand, this balance does not reflect the payment of the 1Q13 distribution of $98mm which occurred after 3/31/13. After making the 1Q2013 distribution, NS only has $18mm of cash to help cover the shortfall for their next quarterly distribution, which we know from the above analysis is not nearly enough. NS will have to decrease its 2Q 2013 (3Q at the latest if they try to bridge the gap with debt) annualized distribution to $2.99. The distribution reduction will cause a massive decrease in the value of the NS GP and therefore the value of NSH. I expect the NSH unit price to drop by 30%-50% on the announcement of the NS distribution reduction in 2Q or 3Q 2013.

 Endnotes:

¹ NS GP Distribution =  ((4.38 - 2.64)/.75*.25 + .075)*77.9 = $51mm

² NS Equity Cost of Capital =  4.38 + (4.38 - 2.64)/.75*.25 + .075 = 5.035/44.96 = 11.2% (assumes no issuance discount)

³ NS’s most recent debt offering (1/15/13) priced at 7.625%. These are fixed to floating notes which stay fixed at 7.625% until 2018 and then begin to float. 7.0% is a generous assumption for NS’s cost of debt.

⁴ For this discussion, we will ignore the ~$4.0mm of G&A that NSH incurs annually. Because the revised GP distribution is so low, inclusion of this expense materially skews the implied multiples. Furthermore, the vast majority of the ~$4.0mm is related to public company expenses; as such, a private buyer would not incur them and likely not include them in its valuation of the NS GP

⁵ The revised GP distribution at the revised LP distribution of $2.99 / unit is $15mm = ((2.99-2.64)/.75*.25+.075)*77.9

⁶ 6.6% is the yield necessary for NS units to maintain current valuation at the revised distribution of $2.70; See Table 1 line 16

⁷ If we assume that NS and NSH yields stay constant after the distribution decrease, the unit price change of an additional $10mm of NS DCF is only $0.99 and $0.99 for NS and NSH respectively (Table 1 Line 19)

Catalyst