Description

Summary: Just like HIF-U, SGU is a classic misunderstood event-driven situation where the market leader in a stable, profitable, high cash flow generating industry has experienced a temporary market dislocation as its recent suspension of distributions has driven away its core base of yield-seeking investors and uncertainty around operating and liquidity issues have weighed on the stock price. We present evidence below that suggests that the underlying economics of the business have not materially changed and that the issues of primary concern have already begun to show evidence of resolution:

General uncertainty & anxiety over the fear of a potential hedging disaster should be reduced by information contained in the recently released 10-K. The primary operational concerns (massive attrition and open-ended margin decline) have also been mitigated by recent trend data found in the 10-K. Liquidity appears substantially more certain than it has at any point since SGU temporarily suspended distributions on October 18 and the company currently has multiple independent options for working capital financing viability.

The opportunity here is not only to benefit from a generous cash flow yield once distributions are fully reinstated (probably in 2006), but also to benefit from unit price appreciation of 25% to 85% (to $8.00 – $12.00) over the next 12 to 18 months as yield-seeking investors once again return to SGU in anticipation of those distributions.


Prologue
Before you continue, please read the write-up on Heating Oil Partners Income Fund (HIF-U CN) posted (by us) on VIC just yesterday (12/14/04). Because the situations are so similar and the data in the HIF-U posting is still completely fresh, we are going to refrain from just repeating any information here. There are, however, a few important differences between the two companies which were highlighted in the HIF-U thesis and we will attempt to provide new answers to those issues herein.

We are posting this idea so soon after HIF-U for two reasons:
1) SGU just filed its most recent 10-K yesterday, which answered a lot of questions that investors have been asking and in our minds positively resolved a lot of the uncertainty surrounding the key issues. Yet the stock price is up less than 9% since Monday’s close (before the 10-K was released), implying to us that the market either has not fully processed or has incorrectly handicapped the new information provided.
2) There are a number of positive near-term catalysts that we expect to play out literally over the next few days.


Company Description

• Description: Star Gas Partners, LP (“SGU”) is a Connecticut-based US Master Limited Partnership (“MLP”) and is the largest residential heating oil distributor in the US. SGU also operates a propane business which serves about 345,000 residential & commercial customers in 26 states.

• Master Limited Partnership: In short, this is a flow-through entity that allows SGU to avoid paying taxes at the corporate level. MLPs typically pay regular dividends (taxable to unitholders) and are valued based on a yield of those distributions (very similar to HIF-U). The propane assets are what qualifies SGU to receive MLP status, whereas HIF-U is currently ineligible (and thus chose the Canadian income trust route).


Recent Events

• Suspension of Distributions: On October 18, SGU announced the temporary suspension of its dividend due to liquidity concerns arising from operational difficulties. The stock plummeted 80% in one day.

• Credit Amendments & Debt Facilities: On November 5, SGU announced the negotiation of certain waivers and amendments under its working capital facility (with Wachovia) that would allow it continued access to funds through December 17, 2004. SGU also announced that it entered into a Commitment Letter with JP Morgan for a $300M asset-based senior secured revolving credit facility (later revised to $350M or $260M if the propane segment is sold) and a $300M senior secured bridge facility (in anticipation of an estimate $300M public or private offering of debt securities).

• Propane Sale: On November 18, SGU announced that it would sell its propane business to Inergy LP (NRGY) for $475M or an estimated 9.5x EBITDA. There were questions as to why SGU chose to sell these very solid assets given that they had already made such strong initial progress toward alleviating their liquidity concerns. The speculation is that the price offered by Inergy was too compelling to pass up (9.5x EBITDA is a very rich multiple for propane). The transaction is scheduled to close on or about December 17 and the sale would lower SGU’s net debt from roughly $520M to approximately $88M (after transaction & financing fees), likely substantially reducing near-term liquidity concerns.


Investment Thesis

• What does SGU have in common with HIF-U? (Read HIF-U write-up for details.)
- Well-positioned market leader in stable, high cash flow industry
- Current fundamental investor mismatch
- Liquidity concerns brought on by increased working capital requirements
- Record high heating oil prices
- Recent adverse weather conditions

• What about the negative SGU-specific issues raised in the HIF-U write-up?
- Customer attrition: For the three fiscal years ending September 30, 2003, SGU experienced an impressive customer attrition rate (excluding the impact of acquisitions) of only 1.3% (versus HIF-U’s rate of approx. 5%). When SGU warned on October 18 of “the effects of unusually high customer attrition,” no one was sure how bad the numbers would be and investors placed a hefty uncertainty discount on the units as a result. The recent 10-K filing clears this issue up for us by providing an attrition rate of 6.6% for fiscal 2004 (ending Sept. 30) and 2.2% for the fourth quarter of 2004 (implying an 8.8% annualized run rate). While certainly representing a meaningful decline from prior years, this figure was probably well below the expectations of most investors as measured by the initial 80% fall in the unit price. More importantly, however, is that in the notes to the 10-K financial statements, SGU provides further data for the months of October and November 2004. Whereas SGU lost 11,000 customers or 2.2% during the fourth quarter of 2004 (vs. only 1,000 or 0.2% for the same period in 2003), the attrition slowed to a loss of only 1,000 customers or 0.2% for the first two months of the second quarter (vs. a loss of 500 or 0.1% for the same period in 2003). In simplest terms, for the fourth quarter, the attrition rate was 11x the 2003 rate, but for the first two months of the second quarter, it was only 2x the 2003 rate. Although the full effects of the customer losses on profitability will not be felt until after the heating season, this is clearly evidence that the attrition rate has shown a marked decline from the levels that lead to the suspension of the distribution.
- Margin compression: On October 18, SGU also warned of “the inability to pass on the full impact of record heating oil prices to customers”. Once again, the uncertainty surrounding this statement was reflected in the discount applied to the units. The 10-K reveals that SGU’s per gallon margin actually increased by 0.8 cents (1.3%) during fiscal 2004 as a result of increases early in the year. However, the fourth quarter shows a per gallon margin decline of 2.3 cents (or approximately 3.9%) as a result of the 38% increase in heating oil prices versus the prior period. As with the customer attrition data, this figure seems quite manageable and is likely well below expectations. It is also important to note that since the end of the fourth quarter, wholesale heating oil prices have fallen 17%, likely alleviating much of the margin pressure going forward as we enter the crucial winter heating season (when ~75% of volumes are typically sold).
- Fixed price contract hedging: The speculation that SGU may have taken unusual amounts of unhedged heating oil price risk on fixed price contracts will likely be laid to rest by the magnitude of the margin decline mentioned above. During a period in which wholesale heating oil prices rose 38% versus the prior year, an aggressively unhedged position would have likely resulted in a fourth quarter margin decline of much more than 3.9%.
- Transparency and uncertainty: Perhaps the largest detriment to the stock price since the Oct. 18 announcement has been the uncertainty surrounding the issues mentioned above and the speculation that unknown others may exist. As a result of the dividend suspension and the resulting unitholder lawsuits, management has been completely inaccessible to both investors and analysts and the flow of detailed information has been anemic. As evidenced herein, that all changed yesterday with the filing of a 200+ page 10-K providing detailed annual, quarterly and subsequent period quantitative and qualitative information. Although the data is still subject to interpretation in the absence of any guidance from management and the heating season has only just begun, the fact that fresh information has been released and that it is apparently free from any major unexpected negatives should be a very strong catalyst for a reduction in the uncertainty discount that has been weighing on the units.

• How real are the liquidity concerns?
- JP Morgan transaction: The credit facility and debt placement bridge facility currently under negotiation with JP Morgan are expected to close on or about December 17. JP Morgan has already provided a Commitment Letter for these facilities and has specifically stated that they are not subject to or predicated upon the successful sale of the propane business. These facilities would provide SGU with up to $650M ($560M if the propane transaction closes) of liquidity for working capital purposes and to facilitate the extension of other near-term debt maturities. There is no information that would indicate that this transaction is in jeopardy for any reason. According to SGU’s 10-K, the transaction-related investment banking fees and expenses are expected to total $20-$35M, quite an incentive for the banks as well.
- Propane sale: The common perception is that Inergy is still very enthusiastic about the propane transaction and is working expeditiously to close the deal by the December 17 target date. According to SGU, it was Inergy that initiated the propane sale dialogue and Inergy’s CEO has been quoted in the press as saying how much Inergy likes the assets and how complementary they are. It is unlikely, therefore, that Inergy would choose to walk away from the transaction and nothing in the recent 10-K filing suggests that they would have any material reason to do so. In order to finance the purchase, Inergy is currently in the midst of a road show to sell $400M of 10 year senior unsecured notes in a private offering. We have heard that the road show is going very well thus far and that the transaction is expected to price on or about Friday, December 17. Given that the stated use of proceeds is the purchase of SGU’s propane assets, the impending pricing of the notes should provide further comfort that the completion of the propane sale is imminent.
- High quality receivables: As discussed at length in the HIF-U write-up, it is difficult to envision a situation where the current (Wachovia) or prospective (JP Morgan) lending group would rather see SGU unable to meet its ongoing working capital obligations than to marginally increase a profitable loan to a very stable company with high value secured assets.
- Politics: In the unlikely scenario that all of SGU’s financing safety nets should fail (propane sale, bridge loan, bond offering, credit waivers) and working capital requirements force SGU into an untenable financial situation, it is possible that there could be the threat of regulatory intervention of some sort to ensure that SGU received financial assistance adequate to continue operating. This is because nearly one in three residents of the domestic northeast rely on heating oil and the 103% rise in home heating oil prices from October 2003 to October 2004 has already become a politically charged issue. This stems from the fact that home heating is more of a necessity than a discretionary purchase and that its inherent price inelasticity disproportionately affects those in the lower income brackets.

• Where should the units trade?
- Historical yield: Since its 1995 IPO, SGU has primarily traded at a yield of between 9% and 13% or between $17.50 and $24.00 per unit, when the partnership was generally paying an annual distribution of $2.30 per unit.
- Projections: Annualizing the attrition rate based on data from the most recent five months (July – November) would yield a run rate of 5.7% However, this figure is likely quite conservative for two reasons: 1) Subsequent period data from October and November suggest that the attrition rate has leveled off substantially, so allowing the fourth quarter data to represent 60% (3 of 5 months) of the annualized figure is probably overstating the attrition rate. 2) Although the attrition rate (expressed as the number of customers dropping service, excluding the impact of acquisitions) was 6.6% for fiscal 2004, the total volume decline related to net customer attrition was only 18.2M gallons or 3.2% of total heating oil gallons as per the recently filed 10-K (not much greater than the attrition-related decline of 14.5M gallons in 2003 – a more normal year). This suggests that recently defecting customers are below-average volume consumers and therefore using the implied annualized attrition rate to forecast the expected decline in gallons sold is once again potentially overstating the attrition-related decline. Nevertheless, assuming an onerous 5.7% decline in volumes from 2004 levels and that the entire 2.3 cent margin decline from the fourth quarter remains in place for the entire upcoming year (even though wholesale heating oil prices have declined 25% since the end of October), would imply a trough gross profit on liquid sales of $297M (versus $327M in 2004). Assuming an expense structure otherwise similar to 2004 (constant per gallon delivery expenses and constant other fixed operating expenses), the projected EBITDA for 2005 (as a fully-impacted trough year) would be approximately $36M (versus $57M in 2004 and $70M in 2003). This figure gives no benefit to cost reduction efforts currently underway and in fact conservatively adds $3M of corporate expense to the heating oil segment figures. As such, we believe this projection to be extremely conservative. After giving effect to the improved capital structure, an additional $1M of potential local taxes and a conservative 50% increase in segment capital expenditures (from $4M in 2004 to $6M), the expected trough per unit distribution (assuming the reinstatement of dividends at a 100% payout ratio – a figure around which management has provided no guidance) would be approximately $0.51. For a normalized year, once the customer service issues have been resolved (returning to 1.3% net attrition) and the margin has stabilized (returning to the 2004 annual level of $0.593 per gallon), similar conservative assumptions would imply a distribution of $1.00 per unit (once again assuming the reinstatement of dividends at a 100% payout ratio).
- Asymmetric pay-off structure: Given the distributions implied by the scenarios above valued at the midpoint of the historical yield range (11%) would result in a 12-18 month floor value of approximately $4.65 using the above conservative downside scenario (although such an onerous trough scenario would almost certainly trade on a much lower yield – perhaps 8% or $6.40 per unit) and a value of $9.00 in the normalized scenario. This translates into a likely potential pay-off structure of –2% on the downside and +40% in the normalized scenario, before even considering any potential growth opportunities.
- Potential upside: As the negative operational factors mentioned herein once again stabilize, there is no reason to believe that SGU cannot return to or exceed the $70M in EBITDA generated in 2003 (implying a distribution of $1.45 per unit or a unit price of $13 using the assumptions above). In fact, that figure was achieved with a smaller customer base than the company currently has (as a result of acquisitions) and under a much less efficient cost structure (further improvement initiatives are underway). Furthermore, as discussed in detail in the HIF-U write-up, the current pricing environment is creating an industry atmosphere that is ripe for consolidation. This fact, combined with SGU’s anticipated much-improved liquidity situation, could lead to substantial dividend growth (through acquisitions) well above the numbers cited herein.


Additional Risks / Issues

• Tax on phantom income: The sale of the propane assets is expected to trigger a taxable gain to unitholders of an estimate $150M. Although the taxable gain to each unitholder is a function of unit purchase price and aggregate distributions (including return of capital dividends), it is our understanding that the taxable gain to the new marginal unitholder can be roughly estimated by simply dividing the total taxable gain by the 36.08M units currently outstanding, for a total of ~$4.15 per unit. At a capital gains tax rate of 15%, the implied tax burden would be $0.62 per unit for the new investor (or ~10% of the total investment). This tax is particularly onerous because no distributions are currently forthcoming from SGU, so the burden would have to be satisfied through unit sales or from funds on hand. This tax is applicable to unitholders of record as of the close of the propane sale transaction, which is expected to be Friday, December 17. If the transaction closes on Friday as planned, the units should trade ex-taxable gain on Monday, December 20 and one would expect them to trade approximately $0.62 higher to represent the absence of the tax burden for the marginal purchaser in addition to some likely substantial amount higher as the concern surrounding SGU’s potential liquidity issues would be resolved. In this event, there would also be a minor timing benefit as a result of the fact that the investor should immediately receive the benefit of the units trading ex-gain before year-end, yet the actual taxes payable would not be due until 2005. Furthermore, should the propane sale transaction not occur for some reason, the units should still trade up to offset the absence of the $0.62 tax burden, netted against some potential incremental discount for liquidity concerns (if any). In essence, an investor in the units prior to the resolution of the sale stands to benefit no matter what the outcome of the propane sale transaction.

• Loss of MLP status: There is some speculation that because SGU will likely succeed in selling its propane assets, it will no longer qualify for MLP status and would therefore lose the benefits of its current tax-efficient structure. However, Petro (the remaining non-qualified heating oil asset) is already currently held through a C-Corporation with net operating losses, so as in the case of HIF-U, SGU would still be able to offset taxable income for years to come. This would also give management plenty of time to research other tax-efficient structures and solutions.

• Structuring for hedge funds: Given the unique tax attributes of MLP distributions, holding SGU units outright may not be appropriate for all types of fund investors. Rather than relying on interpretations herein, you should consult your own tax advisors with respect to your specific situation. In the event that your tax advisors are not comfortable with the tax consequences of outright ownership, one straightforward (and quick) solution is to purchase the units through a swap arrangement with your prime broker.

Catalyst

• Pricing of Inergy bonds (increased propane sale certainty)
• Closing of propane transaction (reduced liquidity uncertainty & units trading ex-taxable gain)
• Close of JPM credit facility or extension of Wachovia waivers
• Continuing retreat of heating oil prices
• Bell-curved winter
• Reinstitution of dividend distributions
• Discovery by value / event-driven hedge fund investors in absence of core base of yield-seeking investors