With interest rates and bond yields at extremely low levels, it's not surprising that many yield-seeking investors have piled into income producing assets such as REITs, high yield debt, and high dividend equities. For example, the Financial Times recently reported that yields on junk bonds hit an all time low of 5.61% in late January and that sophisticated investors are starting to short the asset class.
Not only are investors assuming greater risk in their search for yield, but in many cases have bid up the value of some securities to outrageous levels. Nowhere is this phenomenon more evident than in the shares of Ferrellgas Partners LP (FGP).
Horrifyingly, a deeper dive on FGP reveals that not only are the shares extremely overpriced, but that the "income" investors are receiving is sustained only by an ever increasing mountain of debt and perpetual issuance of new shares. The "yield" is a mirage. An overvalued stock, crippling debt load, unsustainable distributions, and never-ending dilution are bad enough, but it gets worse. My analysis reveals that the company is in a state of secular organic decline that is being masked by an acquisition binge of over 200 companies.
Putting it all together, I have a difficult time valuing the shares at anything more than half of the current share price.
Ferrellgas, L.P. is quite simply a propane distributor with two segments: 1) supplying propane to residential, industrial/commercial, and agricultural customers, and 2) operating a propane tank exchange program (under the brand name "Blue Rhino," primarily for the barbecue market. The Ferrellgas partnership was formed in 1994 in conjunction with the LP's initial public offering, but the company began operations in 1939 as a propane distributor in Kansas.
(click to enlarge)
FGP is the 2nd largest propane distributor in the U.S. (behind Suburban Propane), operating in all 50 states but concentrated in the Midwest, Southeast, Southwest and Northwest regions. The company's website states its business "strategy" is as follows:
Through operational excellence, Ferrellgas is the leading propane company in the United States, with focused expansions through acquisitions and internal growth.
Below is a chart of FGP's share price performance since the 1994 IPO.
As you can see, the stock, despite periodic bouts of panic and euphoria, has gone nowhere for 19 years. FGP has been public, currently trading in the $19 range, 10% below the 1994 IPO price.
In conjunction with the IPO, FGP initiated a distribution at a rate of $0.50 per quarter, or $2.00 annualized. 19 years later, the distribution is still $0.50 per quarter. The fact that a company is unable to raise its distribution at all, let alone keep up with inflation should be a red flag to investors. Using the inflation calculator provided by the Bureau of Labor Statistics, distributions have actually fallen by 1/3 in real terms ($2.00 today equates to $1.28 in 1994 dollars).
A $2.00 distribution on a $19.00 stock appears attractive in this environment, offering investors a "yield" of 10.5%. However, a double-digit yield should serve as another red flag that something is amiss. There is.
Residential propane use is in structural decline. As a natural gas liquid, the price of propane most closely tracks that of oil. Thus, the shale gas bonanza that has driven natural gas prices to extremely low levels has been of little value to beleaguered propane customers. To illustrate just how costly propane is in comparison to natural gas, the U.S. Energy Information Administration (EIA) maintains an updated Heating Fuel Comparison Calculator. Using data as of March 5, 2013 on a per million BTU basis the cost breaks down as follows:
One can see from the data above (which include delivery costs etc.) that propane is one of the costliest ways to heat a residence. Not surprisingly, propane use is generally limited to rural areas where fewer alternatives exist. As natural gas infrastructure continues to be built out, and alternatives such as heat pumps and geothermal become more prevalent, propane use in FGP's core market can be expected to decline over the long term.
This gradual decline in propane use amongst residential customers is confirmed in EIA data. The EIA's March Short-Term Energy Outlook shows the number of households using propane for primary heating, broken down by region, and how this has changed over time:
You can see in the table above that the propane heating business is in a state of long-term decline at a rate of approximately 2% per year. Given the enormous cost disadvantage of propane heat there is little reason to think this trend will reverse. There is also evidence that homes are becoming more energy efficient and using less energy for heating purposes so this decline rate may be understated.
It is worth noting that retail sales to end customers accounts for the vast majority (77% in FY 2012) of FGP's gross profit. Strangely, the company does not break out BBQ tank exchange metrics, even though this is a distinct business. Instead, FGP lumps in BBQ tank exchange customers and bulk propane sold to wholesale customers into a catch-all line item named "Wholesale - Sales to Resellers." Thus some portion of wholesale revenue will be tied to heating. Little data exists on propane barbecue demand (available data lumps natural gas and propane grills into a single category of "gas grills"), however the increasing popularity of natural gas BBQ hookups in urban centers leads me to conclude that it is not a growth business either.
Adding it all up, I find the EIA data a reasonable approximation for the organic growth rate of FGP's addressable market - holding prices and market share constant I would estimate the organic erosion in FGP's top line in the order of 2% per year.
There is virtually no competitive moat to be carved out in the propane supply industry. It is a cutthroat commodity business with fierce price competition and minimal customer loyalty. Furthermore, economies of scale offer only marginal opportunities- the main cost inputs are the commodity itself, labor (delivery), trucks, delivery fuel, and tanks. Scale matters little provided sufficient local customer density exists (thus even small mom and pop distributors are still competitive).
With a little background out of the way, it is time to turn our attention to FGP itself. To generate the following data and charts, I went through 19 years of 10-K filings - every one since FGP's 1994 IPO.
The chart below shows Revenue, Gross Profit, Operating Income (excluding stock compensation), and Distributions paid to unitholders over the past 19 years:
One can see in the chart above, that barring some recent weakness (in part attributable to a warm 2011/2012 winter season, the company's gross profit (a more stable measure of the topline due to propane price fluctuations - can be considered "net revenue") and operating income have increased steadily. The chart shows that total distributions have steadily increased as well, but given the per-unit distribution of $2.00 has not increased since 1994, 100% of the increase in distributions is due to increased shares outstanding.
It is only when one looks below the operating income line and digs into the SEC filings and cash flow statements that a true picture of Ferrellgas emerges. Before we can discuss the disease, I present some of the symptoms:
I found it odd that even though the business seemed to be growing revenue, gross margin, and operating income, the company was unable to pass on a single nickel of that growth in the form of a distribution increase. In inflation-adjusted terms the distribution has actually fallen by 1/3.
So while the company has never been able to raise the distribution, what has happened to the share count? It has ballooned:
(click to enlarge)
Looking at debt, we can see that FGP's debt load and interest costs have increased dramatically also:
Comin' home late smellin' like perfume / Things that make you go hmmmm...
-- C&C Music Factory
At this point we have established that:
- FGP's industry is in a state of secular decline
- The company has paid a distribution that hasn't increased since 1994 (declining in real terms)
- The share count has risen dramatically (about 3X)
- FGP's debt load has continued to climb
It should be obvious at this point that the company cannot organically support its distribution level. If the business were stable there would be little reason to issue shares or assume more debt. If the reasons for the share issuance and debt increases were for new growth opportunities then presumably after 19 years they would have been able to raise the distribution.
If the reasons for issuing shares and taking on debt are not to grow the distribution (shown empirically above), then I conclude it is only doing so in an attempt to preserve the distribution.
In the next section I will show that FGP's distributions are so far above what the business can support that it is almost entirely reliant upon the company's ability to acquire more and more businesses, issue more and more shares, and take on more and more debt.
FGP has told us that acquisitions are central to its "strategy." They aren't kidding. According to the company's most recent 10-K, FGP has acquired over 190 companies since 1986. This is, on average, one acquisition every 7 weeks for the past 25 years. The company's disclosures make it impossible to determine financial results over time with and without acquisitions - there is no "same-store sales" or similar metrics provided. I note that the word "acquisition" appears 89 times in FGP's 10-K, while the word "organic" appears zero times.
Fortunately we can go back and look at what the company has spent on acquisitions over the past 19 years as a public company. Below is a chart that shows, cumulatively, what FGP has spent on acquisitions in the form of cash, share issuance, and assumption of debt:
To put that $1.5 Billion number in context, it is roughly equal to the company's current market capitalization.
With the knowledge of how much FGP has spent on acquisitions, what is happening to the customer base? When I first went digging through the filings I expected to find at least a stable to modest growing customer base (presumably hundreds of acquisitions at a cost of $1.5B would at least stabilize the customer base). What I found was stunning:
It is interesting that despite the acquisition spree, the business is essentially the same size as it was 12 years ago (note propane volumes can fluctuate each year due to weather). Interestingly, for the first time ever (from what I can see), the company has only now just dropped the customer # disclosure above in the most recent 10-Q filing:
Past disclosures looked like this (from the prior 10-Q)"
One can see that the "approximately one million" has now vanished. I assume that FGP stretched the word "approximately" as long as possible and the number could be below 950,000.
Further evidence of the company's deterioration and shifting product mix can be shown in the company's gross profit margins:
With all of the moving parts I thought it would be helpful to zoom out and look at what the business has actually generated over its 19-year history, what has been spent to generate that cash flow (capital expenditures + acquisitions), what has been paid out in the form of distributions, and how any gaps have been filled. This involved poring over every 10-K filing since 1994, but it was the only way to get a complete picture. Due to disclosure and accounting rule changes over the years I have had to make certain (I believe minor) adjustments but I believe the numbers I present are accurate.
I break down the financial results into the following components:
- Operating income (excluding stock-based compensation) as a proxy for business cash flow before interest, capital expenditures, acquisitions, and taxes (an adjusted EBITDA calculation)
- Non-acquisition capital expenditures
- Interest and A/R Factoring Expense
- Acquisition Costs (capex + assumed debt + issued shares)
- Stock compensation
I arrange the components above as follows:
On the basis above, here is how FGP's "Free" Cash Flow To Equity and Distributions appear over the past 19 years:
The above chart may seem alarming - that the company's ability to generate cash is nowhere near the amount it pays out in distributions and partially explains why both the share count and debt levels have risen so much. However, in my opinion things are far worse than even the data shown above.
In the above analysis I haven't even factored in what the company has been spending on a perpetual stream of acquisitions! Recall that the company has spent approximately $1.5 Billion (about the current market cap) buying companies since the 1994 IPO. When we include the acquisition costs and the stock compensation expense into our formula above, the funding gap goes from merely outrageous to downright ridiculous:
Using the above adjustments, under the (appropriate, in my opinion) assumption that acquisitions by FGP are required to maintain the business (a form of maintenance capital expenditure), the analysis shakes out as follows (warning: you may want to sit down before looking at this chart):
The scale on the above chart is important. Note the green line represents zero adjusted free cash flow. When the cost of acquisitions are factored in it becomes clear that FGP generates almost no free cash flow at all. I included stock compensation in the above calculation as it is a true economic transfer of wealth from shareholders to employees (it has averaged about $10M for each of the past 5 years).
Critics of the above analysis may (somewhat validly) point out that acquisitions are not equivalent to expenses in that they generate positive cash flow for a time post-acquisition. My response is that if one is to ignore the cost of acquisitions, then one must base an analysis on the true organic decline rate in the business. In that light, I thought it would be helpful to breakdown the above factors on a cumulative basis since 1994. Here is the output of that analysis:
It might be helpful to grab a coffee, sit back, and stare at the above analysis for a while. Where to begin? We see that over the entire period that FGP has been a public company, it has actually generated negative free cash flow when one factors in the cost of acquisitions (note the above table does not include over $100M of stock compensation). Despite this, the company has managed to pay out $1.8 Billion in distributions over this period. How has it done this? The answers are highlighted in yellow above. The gap has been filled primarily by issuing over $1B in equity to the public markets, taking on over $400 million debt relating to acquisitions, and issuing $235 million in stock to acquisition targets.
I don't want to use the P-word as I don't believe there is anything illegal, fraudulent, or shady here, but using the Wikipedia definition of a Ponzi scheme and crossing out the word "fraudulent" gives us:
A Ponzi scheme is a
fraudulentinvestment operation that pays returns to its investors from their own money or the money paid by subsequent investors, rather than from profit earned by the individual or organization running the operation. The Ponzi scheme usually entices new investors by offering higher returns than other investments, in the form of short-term returns that are either abnormally high or unusually consistent. Perpetuation of the high returns requires an ever-increasing flow of money from new investors to keep the scheme going.
One should ask the question: what would happen if FGP's access to the capital markets closed, and it were unable to make further acquisitions (e.g. buyers unwilling to accept stock and the company unable to take on more debt)?
So why is the company doing this? Why imperil the long-term health of the company for the sake of maintaining an unrealistic and unsustainable yield?
In my opinion the reasons are twofold:
- A high "yield" results in a high stock price as yield-hungry investors chase anything that generates high "income." This high share price and investor demand allows the company to keep accessing the equity markets or use shares as acquisition currency (over $200M in 2011 and 2012).
- 27.5% of the company's shares are owned by employees of Ferrellgas through the "Ferrell Companies Inc. Employee Stock Ownership Trust."
The Employee Stock Ownership Trust is an interesting construct - a pool of ownership that was created in 1998 to own shares for the benefit of employees. Initially the "ESOT" purchased a large block of shares from James Ferrell. Interestingly, FGP's filings refer to the ESOT as a "leveraged" employee stock ownership trust. I cannot find a financial statement for the trust itself, but note the original 1998 disclosure reads:
The ESOT purchased the stock of FCI using funds provided primarily by a private placement of $160 million of debt and $40 million of seller financed notes. -- 8-K Dated July 31, 1998
The 10-K still refers to the ESOT as "leveraged," so I assume there is still debt outstanding against this large pool of employee-owned shares.
A large block of employees that are currently receiving $44 million in annual distributions and have stock "value" of over $400 million could be considered an HR nightmare-in-waiting were the distribution reduced or eliminated. Leverage against that pool of shares (if still leveraged) would be potentially toxic.
From reading the filings it appears that FGP shares "allocated" to employees in the ESOT cannot be traded or sold by an employee - effectively ensuring that nearly 22 million of the company's 79 million shares can never hit the market, regardless of how the individual "owners" may feel.
FGP markets the ESOP as a core benefit to employees, and even positions it as "retirement savings":
I think I've given a decent overview of FGP and how it came to where it is today. So now let's have a look at the company's situation.
We can see from the table above that FGP's enterprise value (market cap + long-term debt - working capital) is approximately $2.7 Billion.
Given the torrent of acquisitions, coupled with weather volatility from year to year, it is difficult to triangulate normalized Operating Income/EBITDA figure on which to judge FGP's $2.7 Billion price tag.
The most flattering approach would be to ignore the dilution and increased debt load, and the fact that EBITDA has been propped up by acquisitions and look at the average over the past few years as a starting point:
Of course we know this number is inflated by the $86 Million in acquisitions FGP has completed during this period, but I am in a generous mood and will ignore that. I will also be generous and not penalize the valuation for the warm 2011/2012 winter (using the full 5-year average)
On a "normalized" EBITDA figure of $218.5 million, FGP is trading at a multiple of 12.3X. Of this $218 million "normalized" number, about $100 million is consumed by interest and AR factoring expenses, and another $40M+ is consumed by non-acquisition capital expenditure. This would leave approximately $78 million left over to cover a current distribution payment level of over $156 million.
How reasonable is that 12.3X multiple on an artificially normalized EBITDA figure? If we assume no erosion whatsoever (0% growth) in the business and no additional cost or dilution from future acquisitions to plug the hole, and a very conservative cost of capital of 7.7% (I note FGP's bonds have a YTW of 7.7% - equity cost would surely be higher but we are in a generous mood, no?) we get:
Using a similar approach but taking into account the company's current capital structure and interest costs, we can take a similar approach on the above assumptions (bumping up the cost of equity to a modest 8% - just over the bond YTW of 7.7%):
Even under this rosy scenario (stability with no further dilution or acquisition costs), FGP is overvalued by 27-59%.
We have a pretty good sense from the analysis above that both the industry and the company are in a state of long-term secular decline. For simplicity, let's use a very mild organic decline rate of cash flows of 3% annually. That does not seem to be much of a stretch given the industry decline rate of around 2% and the operating leverage (opex and G&A running at about 70% of variable COGS). Doing the same math on this decline rate we get:
Under these more realistic scenarios, we get a fair value per share in the $7.50 to $9.00 range, meaning the shares are overvalued by approximately 120% to 170%.
FGP's debt load is very high. The long-term debt alone at $1.1 Billion is about 5X even our generous estimate of normalized EBITDA. This fact has not been lost on Moody's, which downgraded FGP's junk-rated debt last year to B1 with a negative outlook.
I actually believe that businesses in a modest but continual secular decline would do well to trade at an EV/EBITDA multiple of 6x (Intel trades at 4.6X !). Confirming this, there is evidence in the propane industry that deals in the 5-6X EBITDA is not unreasonable - please see here and here. If FGP were to trade at 6X EV/EBITDA, the equity would be worth approximately $2.60 per share.
As stated in my opening paragraphs, I cannot come up with any set of reasonable assumptions that values FGP anywhere close to where it is currently trading. I am of the view that investors simply do not understand this complicated situation. A casual reading of stock message boards sums up the long/hold thesis as this:
"They've always paid me my $2.00 per share."
Little regard is given to the company's ability to continue to pay the distribution, its secular challengers, the debt load, or the acquisition treadmill that sucks up cash, adds debt, and dilutes shareholders.
In my opinion a "normal" company would have taken its lumps long ago and cut the distribution to a more sustainable level (e.g. a 60-75% cut from today's level). However, given the business model's dependence on acquisitions and continued access to capital markets, and the high ownership of captive employees and retirees in a potentially leveraged trust (receiving distributions of $44 million annually), it is clear why they haven't done so yet.
A review of the company's bond covenants shows that they have little wiggle room to add further debt. It's possible that the current cooler (vs. last year) winter will help them, but in my opinion the longer they delay the inevitable the worse the ultimate reckoning could/will be. An increase in U.S. propane export capacity could also add margin pressure over the medium term.
Cracks are beginning to show. Last year the company saw the company begin laying off people in its head office. The company also just paid out a large class action settlement in response to allegations that the company began underfilling propane tanks and fixing prices with a competitor:
The 20-pound tanks, sold under the Blue Rhino brand, normally hold 17 to 18 pounds of propane to guard against expanding gas. With propane prices rising in 2008, the suit alleged that Ferrellgas and AmeriGas conspired to reduce the amount to 15 pounds and charge the same price without telling customers.
The companies' executives allegedly discussed pricing plans in the spring and summer of 2008. The suit said the two companies control 75 to 90 percent of the U.S. propane tank market.
In settling, the companies denied wrongdoing but agreed to more accurately label each tank's weight.
More recently, the company has disclosed that:
The Offices of the District Attorneys of several counties in California and the Federal Trade Commission are investigating cylinder labeling and filling practices and any anti-trust issues relating to the amount of propane contained in propane tanks.
There is also an outstanding class action lawsuit filed in the U.S. District Court in Kansas alleging that Ferrellgas violates consumer protection laws in the manner it sets prices and fees for customers.
An analogue exists in Superior Plus - a Canadian propane distributor. Superior's stock lost approximately 60% of its value following a dividend reduction.
Buyers and holders of FGP may want to revisit their "they've always paid me $2.00" investment thesis before their portfolios risk being left out in the cold.
Additional disclosure: This article reflects my personal views only. I have a short position in FGP stock. All data and calculations presented are accurate to the best of my knowledge but have not been vetted, checked, proofread, or independently verified. This article should not be relied upon for any purpose other than for entertainment. I welcome comments and or corrections.