SGU (Star Gas Partners) is the largest retail heating oil distributer in the US.  It is extremely cheap trading just over 4x EBITDA and 6x FCF (including the seasonal WC release SGU trades at 3x EBITDA and 4x FCF).  The company pays a modest distribution now but is buying back shares and looking for ways to use its large (and growing) cash hoard.

Management is successfully migrating to a more stable, higher margin customer base.  The company has some economies of scale and ample growth prospects through its ability to buy competitors at cheap multiples.

---------------

History:  Star Gas is the largest retail heating oil distributer in the US.  SGU grew through acquisitions and commanded a premium valuation until 2004 when significant hedging losses and a failed transition to a centralized call center cratered the stock price and earned management a famously scathing letter from Dan Loeb.  In 2005, SGU reorganized and recapitalized, replaced management, abolished its cash distribution, sold its higher margin propane business, and wrote off $69m of goodwill.  Shareholders sued (the lawsuit was dismissed but is awaiting judgment in appeals court) and without a dividend, the stock has languished under $5.

Since 2006, SGU has generated significant EBITDA and FCF.  The company was reluctant to distribute cash while it stabilized operations and the recent spike and volatility in energy prices grew working capital requirements and increased customer risk.  SGU's dividend was reinstated in 09Q1.  Current Management is migrating to a higher credit and less price sensitive customer base and margins have grown from $.56 to $.87 per gallon from 2005 to now.  SGU's service and installation business improved significantly and contributed $8m gross over the LTM. 

Between the customer transition, volatility in the energy markets and recession, SGU has had some grizzly customer losses and churn (offsetting losses and gains).  However, customer metrics in the most recent quarter improved (net losses were only 2,600) while the company impressively maintained its pricing (heating oil prices tend to be sticky on the way down inflating margins, margins get squeezed as energy prices rise).  Management attributes some of the 2009 customer attrition to credit losses and talked about improving organic growth in some branches on the last call.  An improving economy and lower energy prices should continue to ameliorate the customer churn going forward.  Both Dan Donovan (CEO) and Rich Ambury (CFO) are longtime industry veterans, speak intelligently about the business and have done a good job balancing customer losses and margin growth to maximize cash.  I estimate that roughly 85% of SGU's costs are scalable limiting operating leverage.

There were several other boosts to SGU in 2009 (SGU's fiscal year ends in September).  Last winter was significantly colder than two years ago.  In the current environment, heating oil has regained its historical price advantage to natural gas which, combined with the significant capital outlay required, should deter gas conversion.  SGU has historically lost roughly 50% of customers who move and home sales are down significantly.  SGU has $80m of NOLs still available.

Analysts miss SGU because GAAP accounting is misleading in two ways.  1) SGU does not use hedge accounting and takes large (very, very large in 07-08) mark to market gains and losses on its derivative book.  The hedges are put in place as SGU negotiates fixed and capped price customers during the year and then offset (at expiry) the difference between the fixed or capped price of sales and the spot price at which SGU buys heating oil.  2) SGU should be looked at on a margin per gallon basis.  When prices exploded in 2008, revenue blew out while gross margins shrank (per gallon margins look small compared to bigger revenue base).  Revenue shrank and gross margins expanded as heating oil prices normalized.

Follow the Cash:   As of 10Q1 (fiscal 09Q3), SGU had $99m of cash and $133m of long term debt (due 2013).  Cash has accumulated through operating FCF and the release of working capital as energy prices fell from peak levels two summers ago.  Winter is the height of SGU's working capital requirements (hedges, receivables and inventory) and at current heating oil prices I estimate that the seasonal WC release is worth ~$100m.  Adding the WC release to the balance sheet, SGU is trading at 3x EBITDA and 4x FCF.   SGU has a $290m credit facility that sits untouched and has also announced that they are now getting trade credit.       

So, what is SGU going to do with all the cash?  Management is in the middle of a 7.5m (~10%) share buyback, raised the distribution by $.02 and called $50m of long term debt at a slight premium.   The debt buyback netted $5m a year in interest savings but was disappointing given the grossly undervalued equity.  

There is the potential for a special dividend.  SGU is set up as an MLP where the GP is Kestrel Heat, a unit of Kestrel Energy Partners.  Kestrel is a small PE shop run by Paul Vermylen, the former president of Meenan Oil which was bought by Star Gas in 2001.  Kestrel's only other investment is Downeast LNG, a LNG import facility in Robbinston Maine that has not commenced construction, is mired in regulatory issues and lacks financing.  As the GP, Kestrel sets the board and manages Star Gas.  Kestrel also owns 17% of the LP or common.  The GP also gets a generous split of 90/10 (LP/GP) on any cash over $.0675 (per quarter) up to $.1125 (per quarter) and 80/20 pro rata above $.1125.  As part of the recapitalization signed in spring 2006, the GP agreed to use 50% of its share of excess distributions to fund a long term management incentive comp plan which allocates participation points to key management.  Vermylen and management also own small amounts of stock outright and bought this spring when prices were in the $2-3 range.

Management has been equally conservative and opportunistic in using cash for acquisitions. There were only $3.9m in acquisitions in 2009 and $2.6m in acquisitions in 2008.  Management is rigorous about buying good businesses at low multiples (2.5-4x EBITDA and 4-5x FCF) and then slowly and carefully integrating.  Because of its relative size in a fragmented industry, there are few competing buyers when mom and pop businesses want to sell out.  Management has intimated that part of the reason for the cash hoard is the potential to make acquisitions and refresh the customer base.

Growth and Economies of Scale:  As the largest player in a highly fragmented industry, SGU has several big advantages.  There are some economies of scale in the advertising, storage, delivery and hedging operations.  SGU also has the opportunity to make highly accretive acquisitions at low multiples.  As mentioned before, these are primarily mom and pop operators which rarely have competing buyers.  While management has been conservative making acquisitions and says it evaluates each business on a case by case basis, they generally buy businesses at 2.5 - 4x EBITDA (of course making acquisitions at those multiples means that SGU should be buying back shares).  Acquisitions are integrated slowly and SGU usually retains the acquisition's personnel and brand.

Comps:  There are no other public heating oil distributers and the best comps for SGU are the propane distributors APU, FGP and SPH which trade at EBITDA multiples of 8-12x.   Propane is considered a slightly better business because of the more diversified customer base (SGU agreed to sell its propane business as part of the 2005 recap and will look to reenter this year as its non-compete expires). 

Corporate Structure, MLP and Taxes:  SGU is a master limited partnership that sits directly above an operating C corp.  The C corp. pays federal and state taxes, but taxes have been offset by NOLs at the operating level.   Cash is dividended up to the parent in order to make cash distributions, pay a very small amount of corporate overhead, and make debt and unit repurchases.   Unit holders are taxed on the amount dividended up to the parent rather on the cash distributions.   However, the excess tax is eventually offset by a negative return of capital that writes up your cost basis when a unit holder goes to take capital gains.   In effect, because the SGU parent pays out less cash than it earns (dividends from the sub), you have a tax acceleration rather than the tax deferral intended by the MLP structure.  The tax difference was material in 2009 as the company bought back significant amounts of debt and stock, but should be smaller going forward.

However, the company has no UBTI and will not have any UBTI going forward.   All of SGU's operations are related to heating oil distribution and qualify for tax exemption.   Worrying about UBTI is a little arcane when the operating company is paying taxes, but the absence of UBTI should make tax exempt investors more comfortable.   Even better, SGU is talking about ripping off the MLP structure, saving ~$500k - $1m in annual filing costs, harvesting ~$2m of deductions at the parent level, and making everyone's lives much, much easier.  My understanding is that this could be messy, but will only proceed in a case where there is no phantom income for unit holders.  In my discussion with CFO Rich Ambury he put the most probable timeline as next tax year (2011).

Risks and Problems:  The largest risk is that continued customer attrition will outpace margin growth.  Although net attrition has slowed, the high level of churn makes it difficult to evaluate how successful management has been in migrating SGU's customer base.   Also, the significant sales lost to conservation may be sticky, similar to the late 1970s when customers permanently nudged their thermostats from 72 to 68.  Variance in yearly temperatures also leads to volatility in annual revenues. 

Because SGU is unable to perfectly predict customer usage, it can't perfectly hedge.  SGU has to estimate how much and when it will actually need to deliver heating oil when it puts on its hedges during the year.  In the case when prices rise and demand exceeds expectations, SGU could find itself underhedged and forced to deliver heating oil at prices below which it can purchase it in the spot market.  Overhedging in a declining price environment is not as pernicious.  Only 14% of customers are on fixed price contracts hedged by swaps vs. 44% which are on capped price contracts where SGU can only lose the option premium on the downside.  Management has said that they are conservative with their hedging, buy extra calls to remain overhedged and build in excess margins on fixed and capped price customers.  A very rough back of the envelope calculation says that if heating oil prices rise by $1.00/gallon and SGU finds itself underhedged by 10% for the entire year (extreme conditions), SGU loses $19m.