Introduction:

Level 3 Communications (LVLT) shares are an ideal short sale candidate, with large downside and almost no realistic upside scenario, for the following reasons.  It is overleveraged:  Level 3's total debt is equal to 7.1x its LTM EBITDA and this ratio is rising because EBITDA is declining.  Its EBITDA / interest coverage ratio is 1.7x and its (EBITDA- maintenance CAPEX)/interest coverage ratio is hovering around 1 (it can barely cover interest payments after CAPEX).  It does not (and has never) generated free cash flow.  Revenues, EBITDA and margins are all declining despite the improving economic environment and there are no catalysts for a turnaround.  Competition is intense - LVLT sells a commoditized service and its industry suffers from overcapacity, too many competitors and constantly declining prices.  LVLT's 2010E sales of $3.6BN are below 2003's $3.95BN despite over $4BN of acquisitions made.  Other than M&A fueled growth in 2007, revenues haven't grown in any year since 2003.  Projected 2010E EBITDA of $746MM is just a bit above 2003's $714MM (EBITDA peaked in 2008 at $901MM).

With no free cash flow generation, declining revenues and profit margins and a very poor outlook, Level 3 cannot get out from under its massive $6.4BN debt burden anytime soon in a way that doesn't impair its equity value.  It could take a while, but the only feasible solutions are debt-for-equity exchanges, convertible bond sales that dilute equity, and bankruptcy.  Share count has already increased from 400MM in 2002 to 1.66BN thanks to "liability management" that retired debt by diluted equity and M&A that only helped the company stand still instead of decline.  LVLT recently reported 1Q 2010 revenues down 7% YoY and EBITDA down 22% YoY - and 1Q 2009 wasn't exactly a banner quarter.  The consensus estimate for 2010E EBITDA was marked down from $850MM to $813MM and I'm at $750MM.  The company once again had to back away from its guidance for positive free cash flow "this year" - which they first projected for 2004 but have still failed to achieve.

Despite all of these issues, LVLT trades at an EV/2010E EBITDA multiple of 9.5x, well above every telecom company I track, including higher quality peers who are less leveraged, growing revenues and generating free cash flow - tw telecom at 7.5x and Global Crossing at 6.6x.  Even allowing for an imminent return to peak EBITDA of $900MM, the stock price implies a 7.9x EV/EBITDA multiple.  In my "optimistic" scenario I give LVLT credit for $900MM of peak EBITDA (note that even at that level they generate almost no free cash flow after $520MM interest expense & $350MM CAPEX) and value LVLT at 11.5x unlevered cash flow of $550MM (equal to 7x EBITDA) and derive an $0.81 / share value, down 35%.  Due to the company's tremendous debt level, my less optimistic valuation of 10x current $450MM of unlevered cash flow leaves no equity value.  I think even this is generous for an average quality at best company with an unsustainable capital structure that's barely keeping its head above water despite benefiting from low interest rates on its debt that are nearly certain to rise.  There are no imminent catalysts as most debt is locked in for several years and LVLT can avoid bankruptcy for a while.  The only positive outcomes - an acquisition of LVLT or value-creating M&A by LVLT - are highly unlikely due to the company's very high valuation (preventing an acquisition), lack of cash / debt capacity and overvalued shares (no one wants to sell assets and take back LVLT stock).  I don't know how this story ends but all realistic outcomes are very poor for equity investors at a share price of $1.25.

Company history:

Level 3 Communications, Inc. (LVLT) is a communications and information services company headquartered in Broomfield, Colorado. It has operating locations throughout the U.S. and Europe. Level 3's North American operations (mostly the U.S. but some assets in Canada) account for 92% of its revenues, with Europe contributing 8%.  The company operates one of the largest facilities-based (they own or lease most of it) communications and internet networks in the world.  Level 3's network is comprised of both a long-haul or backbone network component and metro fiber in 130 local markets.  LVLT provides data, video and voice services to corporate clients and telecom carriers, including telephone companies and cable companies.  LVLT estimates that it accounts for 20% of internet traffic in the U.S. and it is the largest provider of IP services in Europe.

LVLT was founded in 1985 as Diversified Group, a subsidiary of Kiewet Corporation, a large privately held construction company based in Omaha, Nebraska.  Diversified Group was created to invest the cash generated by Kiewit's construction business in other industries; it entered the telecommunications business in 1988.  Largely through the guidance of current LVLT Chairman Walter Scott and current CEO Jim Crowe, in the late 1980s and 1990s Kiewit built a large alternative telecom provider, MFS/UUnet.  This company was acquired by Worldcom in 1996.  Following the acquisition, in 1997 Scott and Crowe re-hired several former executives and set out to start again, building a next-generation communications company from scratch using advanced fiber-optic technology.  LVLT separated from Kiewit in 1998 and was listed on the Nasdaq in an IPO at $54 per share.  LVLT raised $14BN of debt and equity capital to build its network.

Level 3 utilized the capital raised, as well as Kiewit's start-up capital and construction resources, to build a nationwide intercity network (also called "long haul" or "backbone").  It took nearly three years to complete the project, which was finished just in time for the 2000-2002 telecom bust.  LVLT shares peaked above $120 in early 2000 and fell to $5 by mid-2001.  Due to the telecom bust, numerous telecom companies' bankruptcies and substantial excess network capacity, LVLT's revenues and cash flows have grown much more slowly than expected.  This has made it difficult for the company to support its debt load, which since 2000 has ranged from $5BN to $7.5BN and currently stands at $6.37BN.

But Level 3 managed to survive the telecom bust, partly thanks to Walter Scott's connections to financial lenders and investors.  Recognizing the excess network capacity and low asset valuations in the 2002-2005 period, Level 3 set out to consolidate its industry.  In 2002, Scott arranged for Berkshire Hathaway, Legg Mason (Bill Miller) and Southeastern Asset Management to invest $500MM in Level 3 through 9% convertible notes maturing in 2012.  Level 3 has used acquisitions to add valuable metro fiber to its long haul network.  It's important to realize that Level 3's original business plan - long-haul fiber - has failed to generate the expected profits and the company has only managed to retain its 2003 level of sales and revenues by making $4.5BN of acquisitions since then.

The endorsement from Buffett attracted numerous value investors to Level 3's equity and debt securities, despite the fact that Buffett dumped his $100MM investment a year later.  A core group comprised of Southeastern/Longleaf, Fairfax Financial and Loomis Sayles owns nearly 50% of Level 3's equity (part of this equity was received in equity-for-debt conversions).  These investors have repeatedly stepped up to provide financing to Level 3 when public markets were unwilling or unavailable to finance Level 3 on attractive terms.  Unfortunately this has left them trapped with large illiquid equity positions in a company with little to no equity value.  I don't know how long they will be willing to throw good money after bad.

Industry economics can be difficult to ascertain since no companies publicly disclose detailed pricing and volume data.  But industry consultants estimate (and management teams confirm) that in recent years, bandwidth demand has ranged from 30-60% annual growth while annual pricing declines have ranged from 10-30%.  It is important to note that this refers to providing basic high-speed IP access and not more advanced services.  This basic service accounts for a minority of LVLT's revenues and pricing in its other businesses is much more stable.  However, in general, technological advances in optical communications equipment installed by the network operators in recent years have increased the capacity of existing fiber optic networks.  This has once again led to overcapacity and price cutting.

Network overview:

                Since 1999, LVLT has invested over $14BN in capital expenditures and since 2002 it has made $4.5BN of acquisitions.  LVLT estimates that including acquired assets, the original construction cost of its network was $25BN and its estimated replacement cost today is $20BN.  LVLT's network utilizes advanced optical and internet protocol technologies.  It is comprised of 54,000 intercity (long haul) miles and 27,000 metro miles of fiber optic cable in North America and Europe, of which 10,000 intercity and 500 metro miles are in Europe.  LVLT's metro fiber gives it a presence in 130 metro markets.  LVLT also owns capacity on a trans-Atlantic undersea cable system connecting the U.S. and Europe and has purchased or leased capacity on numerous trans-national cable systems to round out its network.  The extensiveness of this network and its end-to-end nature allows LVLT to capture most of the profitability from data traffic without sharing with other network providers.

LVLT provides a range of services:

(1) Internet Protocol and data transport services-including both internet access and IP and Ethernet Virtual Private Networks and Broadband Transport Services such as wavelengths, dark fiber and private line services (transoceanic, backhaul, intercity, metro and unprotected private line services).  These basically amount to customers transporting data across LVLT's high-speed network.

(2) Content distribution services, including caching and downloading, streaming as well as video broadcast services.  This involves transmitting and distributing live content, such as sports events, and also involves finding ways to more efficiently distribute large, frequently-downloaded data such as video files and streaming video feeds.

 (3) Colocation services.  Providing colocation services in facilities directly connected to LVLT's network attracts communications-intensive customers by allowing LVLT to offer reduced bandwidth costs, rapid provisioning of additional bandwidth, interconnection with other third party networks and improved network performance.

(4) Softswitch and voice services-including wholesale VoIP component services, enterprise or business voice services, wholesale voice origination and termination services and managed modem for the dial-up access business.  LVLT provides both long distance transport of traditional telephone calls and VoIP services.

Most of LVLT's customers are wholesale capacity purchasers or large corporations.  A large percentage of revenue derived from these customers is tied to relatively commoditized services, such as basic high-speed internet access or data transport.  LVLT competes based on the extensiveness of its network, its reliability and its customer service.  LVLT's customer churn is around 1% per month, suggesting the average customer tenure is 8-8.5 years.  Importantly, this churn measure does not include customers renegotiating contracts at a lower price, which the company doesn't like to talk about but is widely known to be happening.

The integration of LVLT's eight acquisitions completed in 2005-2007 created significant operational problems in 2007.  LVLT underestimated the difficulty of integrating these companies' network management technology, services and processes.  Additionally, LVLT's layoffs of experienced employees in connection with its acquisitions were too severe.  It was left without sufficient employees to support its acquired systems and could not provision sales (enable new business activation) quickly enough.  LVLT's inability to quickly integrate the companies' systems made it difficult to provide services to current clients and forced the new customer addition process to take longer than expected.  LVLT went from an industry-leading provisioning time (the period from a customer order to bringing them operational) to well below the industry average.  These problems first surfaced in 2007 and continued into early 2008.  LVLT initiated a program named "Unity" to simplify and integrate its many acquired systems and processes.  The Unity processes were largely in place as of year-end 2008 and LVLT had brought its customer service and new business provisioning time back to desired levels.  Still, this incident had a meaningful impact on the company's reputation.

Perhaps LVLT's largest growth potential is its opportunity to add new business clients.  LVLT's network currently connects to 8,100 buildings but their metro fiber network reaches to within 500 feet of over 100,000 buildings.  LVLT expects to add "a few hundred" buildings each year.  Unfortunately this new business opportunity requires substantial capital expenditures.

Competition:

While increasing demand for network services has slowly absorbed the excess capacity created from 1997-2002, the industry remains highly competitive.  Across its portfolio of businesses, Level 3 competes with both traditional telecom providers, including long distance carriers, local carriers (ILECs and CLECs), and also newer alternative networks and service providers, including Global Crossing, Cogent, tw telecom, XO Holdings, Akamai and Savvis.  Within its long-haul businesses, LVLT often competes for business with 5-6 other networks.  AT&T and Verizon are the strongest competitors, but Global Crossing, tw telecom, Quest and Sprint compete as well.  The large number of competitors makes the long-haul business particularly poor.  Level 3 also operates on the more "commoditized" end of the network transmission and services spectrum, while Global Crossing has a higher proportion of value-added services.  In metro fiber markets, LVLT is more likely to be among just 2-3 other providers, of which one is often the legacy telco with an inferior copper-based network.  As a result, metro fiber tends to be more profitable and more valuable since it benefits from a more favorable competitive environment.

Pricing pressure has been a continued problem for the industry over the past several years.  Neither Level 3 nor its competitors disclose estimates of supply, demand, volumes sold or pricing.  However, it is widely believed that the industry continues to suffer from overcapacity and price competition.  While bandwidth demand growth has been slowly reducing overcapacity, it will likely be several years (at least) before capacity becomes constrained and suppliers gain pricing power.  Until then, it is unlikely that Level 3's revenues will grow dramatically and unlikely that its profit margin will increase much.

Financial overview:

                Level 3 has failed to grow its revenues over the past decade.  The large jump in 2007 revenues was thanks to 7 acquisitions completed for $3.4BN in 2006 & 2007.

                Level 3 has not managed to generate free cash flow (positive $44MM in 2009 was thanks to the working capital release from a shrinking business), largely due to its large interest expense and capital expenditure requirements.  This will not make much of an impact on the company's $6.37BN of debt.

 One of LVLT's most valuable assets is its net operating loss carryforwards, which amounted to a tax shield of about $2BN at 12/31/09.  These NOLs don't begin expiring until 2024.  I don't expect LVLT to generate positive taxable income anytime soon and be able to use these.  But they do have value and could retain value in a restructuring.  I've penciled in $1BN of NPV for the NOLs, which is generous considering there's no chance for taxable income for many years.

Level 3 generates a very low return on capital employed.  The capital investment in its network is $14BN and it has made an additional $4.5BN of acquisitions.  The actual cost of building the entire network was $25BN and Level 3 estimates that the current replacement cost is $20BN.  Total assets at 3/31/10 were $8.7BN, if I'm generous and let this proxy for capital employed (despite a $3.3BN goodwill write-off that knocked asset value down) and give the company credit for peak EBITDA-maintenance CAPEX of $550MM, return on capital employed is 6.3%.

Level 3 is a heavily leveraged company.  Total debt of $6.37BN is essentially unchanged from 2002-2003 levels, although EBITDA has increased since then.  Cash on hand amounts to $590MM.  Total Debt / LTM EBITDA is 7.1x and Net Debt / EBITDA is 6.4x.  LVLT has for the past several years stated that its target leverage ratio is 3-5x EBITDA.  LTM EBITDA is equal to 1.7x interest expense, partly because the company benefits from low interest convertible debt and floating rate debt.  LVLT's floating rate debt amounts to $2BN and its average rate is 4%.  Roughly half of LVLT's floating rate debt has been hedged with a $1BN interest rate swap, partially protecting the company from rising rates.  LVLT also has outstanding $1.7BN of convertible debt of which $1.5BN has coupons below 7%.  The interest rates are well below the current yield to maturity on LVLT's debt of over 10%.  Should LVLT's cost of debt rise from its current 8.1% to 10%, its annual interest expense would rise from $520MM to $640MM, reducing its EBITDA / Interest coverage from 1.7x to 1.4x and reducing (EBITDA-CAPEX) / Interest to 1x.

Over the past few years, LVLT has used every balance sheet management technique available to reduce its debt load and refinance near-term maturities.  These have included open market bond repurchases at a discount to par, tender offers at discounted prices, debt exchanges that push out maturities and debt-for-equity exchanges with holders of its convertible notes.  In 2007-2008, LVLT completed debt-for-equity exchanges totaling $713MM of convertible debt.  Out of LVLT's $1.7BN of convertible debt, only $875MM has a low enough strike price ($1.80 strike vs. $1.40 current stock price) that it will possibly be "naturally" converted into equity over the next few years (absent a debt-for-equity exchange).  The remainder will likely have to be repaid in cash or refinanced.

As shown below, Level 3's fixed rate non-convertible bonds currently have yields-to-worst of 10 % - 11%.  Maturities are $38MM in 2010, $196MM in 2011 and $249MM in 2012.  That makes $528MM of maturities over the next 3 years, just under current cash of $590.  Then there's $695MM maturing in 2013 and $2.9BN in 2014.

LVLT's current share price is $1.25 and there are 1.68BN diluted shares outstanding so its market capitalization is $2.10BN.  Total debt is $6.37BN and cash on hand is $590MM, resulting in net debt of $5.78BN.  I value the NOLs at $1.00BN and there are $209MM of other liabilities (asset retirement and lease losses).  The resulting enterprise value is $7.09BN.  Using peak EBITDA of $900MM and maintenance CAPEX of $350MM, unleveraged free cash flow is $550MM; valuing this with a generous 11.5x multiple (7x EBITDA, midpoint of higher quality peers TWTC & GLBC) leads to enterprise value of $6.35BN and equity value of $1.36 BN, equal to a share price of $0.81, down 35%.  I think a more reasonable valuation would use current EBITDA around $800MM and 10x EBITDA-CAPEX of $450MM.  This leads to an enterprise value of $4.50BN against net debt of $5.78BN, leaving no value for the equity.