Quarterly Letters


Quarterly Update

October 25, 2006

Third Quarter Summary

Our composite Equity account is up 8.3% year-to-date, in line with comparable indices.  Our composite Balanced account is up 7.1% year-to-date, modestly ahead of relevant benchmarks.  For the third quarter, our composite Equity account return was between comparable indices and our composite Balanced account return underperformed relevant benchmarks.  Of note, our year-to-date returns have been accomplished with about 15-20% cash balances throughout the year.  Longer term outperformance remains persistent as shown in the table below.

Performance Summary

  Annualized as of 9/30/06
  3Q 2006 YTD 2005 1 Year 3 Year 5 Year Since
Roumell Equity 2.69% 8.30% 12.53% 16.99% 15.23% 15.90%
S&P 500 5.65% 8.51% 10.77% 12.28% 6.96% 2.67%
Russell 2000 0.43% 8.73% 9.98% 15.53% 13.81% 8.64%
Roumell Balanced 2.59% 7.05% 9.76% 13.81% 12.21% 11.47%
Vanguard Bal. Index Fund 4.23% 5.97% 7.58% 9.19% 7.05% 4.62%
Thomson US Balanced Index 3.74% 5.45% 6.81% 8.33% 5.84% 3.62%
Please refer to performance disclosures at end of document.

Anatomy of a Mistake

In our first quarter report this year we highlighted Dell Computer (DELL) as one of our top purchases in the quarter.  A mere six months later we sold the position having determined we made a mistake.  What happened?

We’ve stated many times we typically use one of three types of analyses to quantify value: net asset value, private market value, and going-concern value.  A going-concern value situation is one where we cannot reasonably expect a company to be acquired by a peer (private market value), nor can we easily quantify the value of its assets where the sum of these assets (net asset value) is greater than the company’s current public price.  Rather, in a going-concern analysis, we estimate the cash flows the business generates and thus view the situation as a bond alternative, i.e., income generating investment (hopefully one in which the income is growing, which doesn’t occur in a straight bond investment).  To wit, we stated the following in this year’s first quarter report: “In the case of Dell, we discovered a company generating roughly a 6.3% free cash flow yield based on its prior three years average free cash flow…This 6.3% yield compared favorably with the 4.75% yield on the 10 year Treasury bond at the time of purchase.  In our view, Dell’s free cash flow is likely to grow.”

The consequence of investing in a pure going-concern situation where the cash flows do not materialize as expected can be severe due to the absence of an obvious asset or private market fall-back valuation.  Dell’s net margins have dropped by nearly 40% (dropping from 6.8% over the past two years to an estimated 4.2% for the next two years) as a result of execution missteps and unfavorable industry trends.  In other words, the cash flows so instrumental to a going-concern analysis dropped materially and, naturally, so did the company’s share price.  Dell will most assuredly remain a powerful distributor of computers and related gear.  The investment question, however, turns on what the free cash flow generating abilities will be going forward, and what yield should be demanded given what is clearly a riskier, less predictable business model than its past results indicate.

We sold Dell because we are not in the business of making investments based on business model changes - turnarounds - that cannot be underscored by either a net asset value or private market value analysis.  When we purchased Dell, we did not anticipate it being a turnaround; rather, we viewed it as a mis-priced going-concern situation.  We were wrong.

Our Three Top Purchases in the 3rd Quarter

Pioneer Drilling Co., PDC.  Pioneer Drilling is a land drilling contractor that operates 63 rigs (as of 10/06) primarily in Texas, Oklahoma and the Rocky Mountains.  PDC is the sixth-largest U.S. based land driller based on footage drilled, according to The Land Rig Newsletter.  Pioneer has been in operation for almost 40 years; the company was founded in 1968 under the name South Texas Drilling & Exploration, Inc.  Pioneer Drilling is led by an experienced management team comprised of Stacey Locke, CEO, and Franklin C. “Red” West, COO, who respectively have 27 and 40 years of service in the oil and gas contract drilling industry.

The contract drilling industry has historically been characterized by “boom/bust” periods with natural gas price swings serving as the tail that wags the dog.  As commodity prices rose and demand for drilling increased, drilling dayrates improved and subsequently additional rigs would come online to share in what was temporarily a highly profitable business.  When prices fell and demand subsided, dayrates would decrease and the additional supply created an intensely competitive pricing environment.  Rigs that were only marginally profitable during “boom” periods were often taken out of service during the “bust”.  A shift, however, is taking place in the domestic natural gas industry.  Production has decreased from approximately 1,400 bcf (billions of cubic feet) in 2001 to 450 bcf in 2005.  During this five year period, however, rig count increased from about 1,100 to 1,400.  This bifurcation in what historically have been two congruent markets has been caused by the depletion of easy-to-reach gas deposits and a subsequent search for “tight” shorter life gas formations to replace eroding supply.  While some believe that additional capacity and the pending entry of liquefied natural gas will cause another bust, we feel that Pioneer Drilling is extremely well positioned to take advantage of this shift in the contract natural gas drilling market and can be profitable in both boom and slow periods.

First, the company is extremely well capitalized with cash of $90 million, zero debt and 85% Net Equity/Total Assets.  In layman’s terms, the company has financed 85% of its asset base with shareholder equity through the use of both retained operating earnings and capital raised through equity issuance.  Pioneer’s strong balance sheet provides financial flexibility to either construct its own rigs during periods of high demand or, conversely, purchase assets from competitors at attractive prices in the private market when industry-wide rig utilization drops.  Second, at our average purchase price, we paid roughly $8 million per rig for PDC’s portfolio of operating assets.  This compares favorably with the estimated $9 million the company currently pays to construct a new rig and does not take into consideration the value of Pioneer’s operations, including its management team, its skilled engineers, and its many customer relationships.  Lastly, we believe that Pioneer Drilling strategically manages its rig portfolio in a manner that maximizes return and minimizes spot market dayrate risk.  The company accomplishes this favored risk/reward balance by locking up roughly 75% of its rigs to long-term contracts priced at highly profitable dayrates.  In the contract drilling market, a long-term contract is generally considered to be at least six months.  Pioneer, however, has secured unusually long contracts of two to three years on its recently built rigs and maintains contracts in excess of one year on 12 of its mature rigs.  A strong balance sheet, experienced management team, favorable replacement cost dynamics and a 7x current P/E multiple make PDC an attractive investment in our opinion.

Tecumseh Products Co., TECUA. Tecumseh Products is a struggling United States manufacturer that goes back seventy years and is run today by one of the founder’s grandsons, Todd Herrick.  The founding family maintains a significant investment in the company, with the Herrick Foundation owning about 27% of outstanding shares.  Tecumseh currently operates in three segments: electric motors sold under the FASCO brand, small engines principally used in lawnmowers and snow-blowers, and compressors used in refrigeration and air conditioning.  In our view, Tecumseh has identifiable pieces that, when added together, result in a value much greater than the one we paid.

Tecumseh’s enterprise value (market capitalization plus net debt) was about $525 million at the time of our purchase.  The company has an over-funded pension plan of about $200 million, a rare instance among old manufacturing companies.  If the company were to reclaim this money and pay the penalties associated with moving from a defined-benefit to a 401(k) style plan, it would net somewhere between $80 to $100 million.  If, on the other hand, Tecumseh were acquired by another company that had an under-funded pension liability, that company would be able to utilize dollar for dollar Tecumseh’s excess, thereby giving the acquirer $200 million of value such that they would effectively be paying $325 million for the company.

Tecumseh’s purchase of FASCO in 2002 provides another important component in quantifying its value.  FASCO was purchased for about $405 million in cash.  The company recently wrote-off $100 million in goodwill associated with this acquisition that consequently values FASCO at about $305 million.  To be conservative, we marked it down further and attached a private market value of $250 million to FASCO.  At this point, we can subtract out a total of roughly $350 million from Tecumseh’s enterprise value ($100m for the over-funded pension plan plus $250m for FASCO) leaving $175 million for “everything else.”  The last element of this sum-of-the-parts analysis is based on the company’s revenue.  Tecumseh’s annual revenue (less FASCO’s $400 million contribution) is about $1.4 billion, for which we are paying about twelve cents on the dollar ($175 million divided by $1.4 billion).  This is indeed a very cheap valuation even for a struggling manufacturer.  Earlier this year, Tecumseh sold its Little Giant pump business for $121 million cash, or roughly 1x sales.  Little Giant was a higher margin business, but its valuation is still instructive in our opinion.

Even though manufacturing can hardly be called a sweet spot in investing given the dramatic rise in Asian capacity, there are some encouraging signs at Tecumseh.   The company was awarded a Toro contract for its new electric-start mower to be sold at Home Depot, at the expense of a larger competitor, Briggs and Stratton.  Additionally, FASCO recently won back some business it lost earlier to Asian competitors because of poor quality.  In fact, though margins have declined, Tecumseh’s overall revenue has held fairly steady over the past few years. Finally, Tecumseh has real estate holdings both in the U.S. and abroad that, although not quantified with precision, provides a further margin of safety.  In short, the identifiable pieces add up to much more than the whole as valued by the public market at the time we made our purchase. 

Premiere Global Services Inc., PGI.  Premiere Global Services is a global outsource provider of business process solutions.  The company operates in two segments:  Data Communications and Conferencing.  The Data Communications segment delivers blast faxes and, more recently, blast emails on behalf of its Fortune 500 customers.  This segment represents about 45% of total sales and 35% of EBITDA (Earnings before interest, taxes, depreciation and amortization).  Data Communications revenue has been declining for the past several years yet remains profitable.  The Conferencing segment is a growing and profitable business that provides a suite of traditional and VOIP-based solutions.  This segment represents about 55% of total sales and 65% of EBITDA.  PGI was founded in 1991 by its Chairman and CEO, Boland Jones. 

PGI was purchased for our portfolio using a private market analysis.  Recent merger and acquisition transactions involving targets similar to Premiere Global suggested that PGI shares were undervalued at the time of our purchase.  In early 2006, West Corp. bought Raindance Communications (with about $75mm in sales and net margins of 5.5%) for roughly 1.5x EV/Sales and 9.5x EV/EBITDA.  Subsequent to this transaction, an investor group announced a deal to take West Corp. (with about $1.6b in sales and net margins of 9.5%) private for roughly 2.0x EV/Sales and 8.5x EV/EBITDA.  We bought PGI, whose net margin is comparable to West’s, for approximately 5.5x EV/EBITDA – a 35% discount to West’s private market value.  To view it another way, if one assumes PGI’s Conferencing business is worth simply 8x EV/EBITDA, we purchased its Data Communications business for only 1x EV/EBITDA.  Although the company has net debt on its balance sheet, PGI is a good free cash flow generator with a free cash flow yield of about 7% (not to be confused with a dividend yield).  In addition, PGI has been repurchasing shares at a premium to our purchase price.  Finally, we are aligned with the management team as Boland Jones and other company insiders own about 8.5% of PGI.  We believe we have a compelling investment in this consolidating industry.

Once again, we want to thank you for the trust and confidence you have placed in us.  We will continue to watch what the market appears to make cheap, diligently research those ideas and, in the end, select our investments cautiously.  If you are aware of individuals who would benefit from our approach, we would welcome the opportunity to talk with them.  You can always direct such individuals to our website, www.roumellasset.com.  Additionally, we have lots of terrific restaurants in our Chevy Chase neighborhood and would be happy to take you and a friend(s) to lunch to discuss our investment approach.  As always, please feel free to contact us directly about your account.

Performance Disclosures:

Investment Strategy:  Roumell Asset Management, LLC (“Roumell”) employs a value investment strategy in managing client portfolios.  Roumell Equity accounts can have up to 100% of assets invested in stocks; Roumell Balanced accounts typically have 65% of their assets allocated to stocks (though the figure can range from 50% to 80% depending upon the needs of the client).  

Calculation of Rates of Return:  First and foremost, readers of this letter should recognize past performance is no guarantee of future results.  Returns are reported net of all management fees and applicable trading costs; annualized returns are the result of linking quarterly returns (only accounts present for the entire quarter are included in a given quarter’s performance composite); returns are time-weighted; returns reflect reinvestment of dividends and other earnings.  Returns include an insignificant number of accounts that utilize margin.  These returns are based on a composite of Roumell’s accounts and therefore were not necessarily duplicated in any specific account.  These consolidated performance numbers include all of Roumell’s fully discretionary accounts within each category.  Discretion is defined as the ability of the firm to implement its intended investment strategy without restriction.

Inclusion of Accounts:  Currently, in the performance calculations for Roumell Equity, there are 423 equity portfolios totaling $147.6 million.  This represents 65.6% of total portfolios and 58.3% of total dollars under management.  The standard deviation for Roumell Equity in this quarter is 1.0%.  Currently, in the performance calculations for Roumell Balanced, there are 157 balanced portfolios totaling $81.5 million.  This represents 24.3% of total portfolios and 32.2% of total dollars under management. The standard deviation for Roumell Balanced in this quarter is 1.7%.  A complete list and description of the firm’s composites is available upon request.  Additional information regarding policies for calculating and reporting returns is available upon request.

Comparative Indices:  Because Roumell utilizes an all-cap (large, medium and small companies) investment strategy, there is no perfect index for comparison purposes.  We have included two equity indices and two balanced benchmarks to allow readers to judge our performance against benchmarks that collectively offer good comparative illustrations.

Additional Disclosures:

The specific securities identified and described do not represent all of the securities purchased, sold, or recommended for advisory clients and the reader should not assume that investments in the securities identified and discussed were or will be profitable.

Back to Quarterly Letters

2 Wisconsin Circle, Suite 640 | Chevy Chase, MD 20815 | Phone: (301) 656-8500 | Fax: (301) 656-8501 | Map and Directions