Description

This report provides a credit analysis supporting a long position in Liberty Media's bonds & trust preferred stock, which I believe are severely mispriced and offer a massive margin of safety and substantial return potential - one of the best risk-reward opportunities I've ever seen. In addition to the bonds, there are four exchange-traded trust preferred issues (essentially repackaged bonds) outstanding. Due to their relative small size & liquidity, retail investor ignorance and general neglect, they trade at substantial discounts to the bonds, offering smaller investors access to Liberty's bonds at a better risk/return proposition.

While Liberty is leveraged with debt of $12.5BN, it is also liquid, with $2.5BN cash, in-the-money derivatives soon adding $1.45BN cash and an equities portfolio of $2.9BN. I believe Liberty's assets are worth over $20BN, making asset coverage 1.6x. Even in my conservative liquidation analysis, assets coverage is 1.1x. With no refinancing or liquidity issues until the QVC bank debt maturity in March 2011, a buyer of Liberty's 8.25% bonds due 2030 at 52% of par will receive two years of coupons before any risk of default, returning a third of their investment and leaving them with a net basis of ~35% of par in March 2011. If in the event of default recovery value is at or above this, investors get back their investment, ensuring preservation of capital. I believe the likelihood of a default arising from that March 2011 refinancing is extremely low, and even in the event of default, my fire-sale/liquidation asset valuation indicates Liberty's senior unsecured debt is fully covered by asset value (would recover 100% of par).

I believe the elimination of the QVC refinancing uncertainty (coupled with improvement in the economy and credit markets), will result in these bonds increasing in price from 52 to around 80, offering investors a 50% plus capital gain to go with the 16% current yield.

Better still for small investors, the third-party trust preferreds trade around 40% of par (an over 20% discount to their underlying bonds), offering a 22% current yield, a return of 44% of investment prior to any default risk, and capital appreciation potential of 100%.

These trust preferreds trade ~$110k per day, so institutional investors will want to focus on the bonds or CDS. While this analysis offers something for investors of all sizes, ironically it's the small retail investor who gets the superior risk-return opportunity.

Liberty Media overview

Liberty is a complex media conglomerate. For brevity's sake, I will only go into detail on individual assets to the extent necessary to make the case that the bonds are fully covered by a conservative liquidation analysis. I can value the operating subsidiaries at 4x recession-level 2009E EBITDA and take 50% haircuts to other assets' estimated values and still get to an asset value that fully covers the bonds. I have researched and analyzed these assets in much greater detail and I'm happy to provide additional details or answer questions. I would point those who'd like incremental background or details to the recent write-ups of Liberty's three tracking stocks (LCAPA, LINTA and LMDIA).

Liberty's equity is structured as three tracking stocks: Liberty Capital (LCAPA), Liberty Interactive (LINTA) and Liberty Entertainment (LMDIA). From a creditor's perspective, this isn't important as the assets and liabilities are legally all held by Liberty Media. Liberty has "attributed" assets and liabilities to each tracking stock but these can be moved around at fair value without a shareholder vote, so while this analysis groups assets by tracking stock to follow Liberty's disclosure, it otherwise considers the assets and liabilities as all being Liberty Media's.

Liberty is in the process of spinning off its DirecTV stock and several smaller LMDIA assets. Since the spin is likely to be completed in May/June 2009 and results in the separation of ~$12BN of assets but only $2BN of debt, it is a negative for creditors, so this analysis assumes the spin-off is completed as planned.

Liberty Media is controlled by Chairman John Malone. On the positive side, Malone has decades of experience operating leveraged companies and executing complex financial transactions. He also has a long track record of creating substantial shareholder value through ownership of media assets. On the negative side, he is aggressive in using leverage. While he has toned down the size of his equity repurchases and shifted to repurchasing debt, his investment in Sirius XM shows that management will attempt to increase shareholder value with an aggressiveness that can make creditors unhappy.

In 2008, Liberty generated consolidated EBITDA of $1.58BN, with QVC contributing $1.5BN, Starz Entertainment $0.3BN, Starz Media losing $185MM & Corporate/Other of -$29MM. Liberty compares very favorably to most leveraged companies whose bonds trade at lower yields. Liberty is highly liquid, generating free cash flow and has financial flexibility. Over 38% of its debt is locked in until 2023. But because of Liberty's complexity, traditional credit metrics are not only unhelpful, they're misleading (Liberty's credit situation looks poor when measured on Debt / EBITDA). Although many of Liberty's assets don't generate cash, the two major cash-generative assets (QVC and Starz) should produce $1.6BN of EBITDA in 2009, which covers Liberty's interest expense of ~$490MM over 3x.

Liberty's 8.5% bonds due 2029 last traded at 60.25% for a 14.7% YTM, while the 8.25% bonds due 2030 last traded at 51.50% for a 16.6% YTM. Liberty's exchangeable bonds trade at even lower prices (20-30% of par), offering greater capital protection. Because they are lower coupon, they offer lower current yields than the 8.5s and 8.25s.

Third-party Trust Preferreds overview

A third-party trust preferred stock (TPTP) is a repackaging of a corporate bond, which is placed into a trust that issues trust certificates which trade on an exchange in small sizes for a retail investor.

There are four of these trusts containing the Liberty Media bonds. These are pass-through structures; for tax purposes the investor is treated as if they own the underlying bonds. The TPTPs are somewhat complicated, so I encourage a read of the prospectuses. Across all the TPTPs, daily trading volume has averaged 11,000 shares in 2009, with daily dollar volume traded around $110,000, making them suitable for small investors, or those willing to build a position over several days. The four trust preferreds trade on the NYSE under the symbols PYA, PYL, PIS and PKK. Their relevant features summarized in the table below.

Some features of the TPTPs are similar. They have $25 par value (each share is a 1/40th interest in a $1,000 bond), pass through coupons when received from the underlying bonds and mature when the bonds mature. One negative aspect of the TPTPs is that in the event of default, the trustee sells the underlying bonds 30 days after default at whatever price is available. This exposes investors to the risk that this price is below ultimate recovery.

PKK is the simplest TPTP structure. $31MM par value of Liberty's 8.5% bonds due 2030 were placed into the trust, which issued 1.24MM trust certificates of $25 par value ($31MM total). The PKK trust passes through the semi-annual 8.5% coupons and repays $25 at maturity. It is a pure pass-through of the bonds and should trade at parity to underlying. With the 8.5s trading at 60% of par, PKK should trade at $15, but is currently at $9.70, a 36% discount.

The PIS trust holds the 8.25% of 2030 and is over-collateralized. $133.5MM of bonds were placed into PIS, which issued 5.035MM trust certificates with stated value of $25 for a total of $125.875MM. So each trust certificate was over-collateralized to 106% of par to create a 8.75% yield out of a 8.25% bond. PIS is larger and more liquid than PKK, offers a higher coupon, 106% overcollateralization in the event of default and repayment at maturity of $26.51 instead of $25. So PIS is a slightly superior security to PKK, and it should trade at a small premium to PKK and to its underlying. With the 8.25s at 51.50%, PIS should trade at or above $12.50 vs. its current $10, a 22% discount.

PYA and PYL have complicated structures. The PYA trust purchased $30.55MM of the 8.25s and issued 1.22MM shares paying a 7% coupon. The PYL trust purchased $35MM of the 8.25s and issued 1.4MM shares paying a 6.7% coupon. Both represent "Class A" trust certificates; Class B trust certificates were also created. The Bs are interest-only securities and not publicly traded. They skim interest from the trust but receive no payment at maturity. In the event of default the prospectus gives them a share of the trust's assets based on a formula. I calculate PYA's claim is ~88% of the underlying bonds while PYL's is ~85% of its underlying bonds. So in addition to having lower yields than PIS and PKK, they have a smaller collateral claim. PYA and PYL are inferior securities and should trade at a discount to underlying and to PIS and PKK. A 20% discount is warranted to equalize the yields and account for the smaller collateral claim. PYA and PYL trade at far greater discounts - around 38% discounts to the underlying 8.25s.

Evaluating the four TPTPs based on current yield is an effective shortcut. For example, a buyer requiring a 20% current yield would purchase PKK at or below $10.63, PIS at or below $10.94, PYA at or below $8.75 and PYL at or below $8.38.

Capital structure & liquidity

The following table displays Liberty's capital structure and the trading levels of its bonds.

A few observations: (1) Besides the $117MM of 2009 notes and the Sprint collar borrowings (which are self-funded), the only near-term refinancing is the $5.2BN of QVC bank debt; (2) the long-dated bonds trade below 60% of par and 14% yields, very attractive if fully covered by asset value; and (3) Liberty's $5.7BN of bonds have a market value of just $2.7BN, opening the door for Liberty to repurchase debt at discounted prices, a big positive for creditors.

I've ignored a large liability on Liberty's balance sheet: at 12/31/08 Liberty had a deferred tax liability of $5.8BN. Without getting into too much complexity, I'll explain why I've ignored it. This liability comes from three main sources: unrealized gains on DTV stock, on the other stocks in Liberty's portfolio and on "phantom interest expense" and unrealized gains on the exchangeable bonds. The liability related to DirecTV stock will be transferred to the spun off Liberty Entertainment entity. The unrealized gains on the LCAPA & LINTA equity portfolio have been accounted for through my 50% haircut to their current value. The phantom interest expense related tax is not due until maturity in 20+ years and the unrealized gains on the carrying value of the exchangeable bonds may not ever be paid, and if they are it'll likely be at maturity in 20+ years.

Asset valuation

This valuation assumes a fire sale or liquidation of Liberty's assets. The valuation methodology is to value operating businesses at 4x 2009E EBITDA and haircut the estimated values of non-operating assets by 50%. I believe this represents a rock-bottom valuation under current market conditions (which could worsen). This yields a current asset value of $13.7BN against total debt of $12.5BN, suggesting Liberty's bonds are worth close to 100% of par even in a forced liquidation or default scenario (with over $1BN to spare), much higher than the 40% - 60% range implied by the bonds' and trust preferreds' deeply discounted trading prices.

The next table shows Liberty's cash and "near-cash" (the Sprint puts).

Liberty's publicly traded equities portfolio is shown below (some amounts are net of shares sold short for tax purposes):

Liberty Capital's privately held assets are detailed below. I've included some notes on the valuation. I will haircut these values by 50% in the liquidation analysis.

The final table presents my fire sale / liquidation analysis. The cash and "near cash" from tables above are unadjusted, but the private assets are haircut 50%.

I have not provided my undiscounted, intrinsic valuation of Liberty's assets, which I've is ~$20BN. While it is based on a more detailed analysis, reducing the haircuts to 10-20% and using 6x EBITDA multiples gets you pretty close to $20BN.

I will first comment on the smaller assets, then discuss larger ones below.

In 2007, Liberty set up a structured vehicle (LCAP Investments), a leveraged media debt investment fund. This vehicle borrowed $750MM from a financial institution at LIBOR+25bps and these borrowings are available to purchase bonds Liberty finds attractive if Liberty contributes 20% cash collateral. As of 12/31/08, Liberty had invested $293MM and $518MM in restricted cash remained. I haircut the bonds' value by 50% and add that to the cash to reach my $665MM valuation.

Liberty purchased collars on its Sprint position to hedge it. Liberty holds ~50MM puts with strikes around $48 that mature in late 2009 and late 2010. These puts are currently worth $2.2BN. I assume that Liberty pays a 35% tax when the puts mature. However, Liberty is creative in its tax deferral techniques and has said they're "working on" minimizing the immediate cash taxes paid. Liberty has a credit facility secured by the puts and borrowed $625MM against it. Recently Liberty disclosed that it's borrowed the full amounts available under the facility to eliminate the counterparty risk of these in-the-money derivatives.

The Other LINTA businesses are a collection of e-commerce websites.  They are growing rapidly and generated $50MM of EBITDA in 2008, so I've valued them at $200MM.

QVC

QVC sells jewelry, apparel, home goods & electronics in the U.S. and internationally through cable networks and websites. QVC has been a steady cash machine for the past several years, but results deteriorated in the second half of 2008. From 2006-2008, QVC had sales of $7.1BN, $7.4BN and $7.3BN and EBITDA of $1.66BN, $1.65BN and $1.5BN. A detailed discussion of the sustainability of QVC's earnings power is outside the scope of this analysis, I'll make a few points. Relative to traditional retailers, QVC's business model features higher margins, lower fixed costs, lower inventory risk and enhanced merchandising flexibility. Nonetheless, it is not immune to weakening consumer spending and both sales and margins declined in late 2008 & will continue to decline in 2009. In 3Q08, EBITDA declined 14.3% year-over-year and in 4Q08 it fell 21.6%.

I am projecting 2009 EBITDA of $1.3BN. At this level, I estimate QVC will generate ~$640MM in free cash flow (after paying ~$220MM of interest expense on its bank debt). QVC should support its bank debt load and remain solidly profitable and free-cash-flow generating under current poor retail conditions. However, its declining profitability raises concerns related to its bank debt. For the liquidation analysis, I am valuing QVC at $5.2BN, just 4x 2009E EBITDA.

QVC has $5.25BN of bank debt; $5.23BN was drawn at 12/31/08. This bank debt is extremely low cost: the interest rate is currently LIBOR+87.5 bps or ~2.1%. There are two major concerns on the bank debt. It contains a covenant that QVC's debt/ trailing EBITDA ratio must stay below 4.0x. Also, it must be refinanced in March 2011. QVC will breach the covenant if EBITDA falls below $1.3BN. If EBITDA decreases by more than 20% year-over-year in the first half of 2009, QVC will brush up against the covenant at June 30th. Liberty's CEO recently stated he doesn't think they will breach the covenant (QVC's EBITDA won't decline below $1.3BN). Several remedies are available if they do. The simplest is using cash to pay down bank debt to get under 4x. Liberty could also negotiate an amendment with the banks to increase the interest rate, or offer incremental security from other Liberty assets.

The maturity of QVC's $5.2BN of debt in March 2011 is the first major refinancing Liberty faces. Assuming repayment of the '09 notes, no free cash flow generation over the next two years and that cash increases from the maturity of the Sprint puts (net of the $625MM borrowed), Liberty would have $3.2BN of cash in March 2011. Even if the credit markets and QVC's business do not improve, I strongly believe Liberty will be able to repay the QVC debt through a combination of paydown with cash on hand and refinancing (probably a new, smaller bank deal with tougher covenants and a higher interest rate).

Starz Entertainment / Starz Media

Liberty's Starz Entertainment business is comprised of the Starz and Encore cable networks. An indepth discussion of the sustainability of its profitability is outside my scope.  I will note that as long as Liberty/Malone maintain interests in DirecTV and several other cable networks, Starz's market power should remain at current levels. Starz has started to focus on original content a la HBO, including series and movies produced by Starz Media. Liberty mentioned on its 4Q conference call that Starz finished the year tied with HBO as the ninth most viewed cable network among households with premium channels.

Starz has grown revenues 3-4% for the past few years by adding subscribers. EBITDA has grown by 9% in '06, 42% in '07 and 27% in '08. This has been due to old expensive movie distribution agreements rolling off & being replaced by lower-cost content without losing viewers. Liberty has guided to 20% EBITDA growth in 2009 but I'll stick with 2008's lower $300MM and use a 4x mutiple to value Starz Entertainment at $1.2BN.

Starz Media is Liberty's movie and TV studio. It produces original content and distributes DVDs. As Liberty has been ramping up its operations in recent years, it has been burning cash. Starz Media generated EBITDA of -$189MM in 2008. Liberty has said it expects to spend $130MM cash on Starz Media's budget in 2009, so its operating loss will likely be less negative in 2009.

Sirius XM

Liberty structured its recent investment in Sirius XM as a two stage transaction. Liberty will invest a net $500MM cash for $530MM of senior secured loans (a $30MM "structuring fee" was paid to Liberty). $100MM will purchase existing loans from banks; $400MM will be in the form of new senior secured loans paying 15% interest and preferred stock convertible into 40% of Sirius XM's equity. While the second phase hasn't been completed, I have reduced Liberty's cash balance by the entire $500MM and included the $530MM loan investment in LCAPA's assets. I haven't analyzed Sirius XM in depth, but Liberty's investment appears reasonably secure, and I've haircut it by 50% in the valuation. I'm also ignoring the value of the preferred stock, which based on Sirius XM's current market cap of $500MM, the market is valuing Liberty's preferred stock at $330MM.

Recent Events

Liberty conducted two debt repurchases in late 2008. The first retired $900MM of 2009 bonds at par, avoiding paying high coupons. The second tender offer was targeted at the 8.25 and 8.5 bonds: Liberty retired $176MM of the 8.5s and $309MM of the 8.25s at a price of 58.75% of par.

The fiscal stimulus law contains an advantageous tax provision for companies buying back debt at a discount. The issuer owes taxes on the gain between the purchase price and par (called cancellation of debt or COD income). Under the new law, the company pays no tax for 5 years, then pays tax ratably over the next 5 years. So the tax payment is deferred by an average 7.5 years; using a 10% discount rate, the present value of the tax payment is around $0.49 per dollar, so the PV of the taxes has been halved. This is advantageous to Liberty Media; it doesn't have to come up with cash right away, and with the effective tax cost halved, the economics of a debt repurchase have improved.

Liberty Media's management stated during their 4Q 2008 conference call that they are looking hard at buying back debt, mentioning the new tax provision & how it makes this more attractive. While another tender offer targeted at the '29s and '30s would be a strong positive for holders of those bonds or the trust preferreds, with the exchangeables trading at lower prices, Liberty could retire large amounts of exchangeable debt with a smaller cash investment.

At an investor conference in January, someone asked Liberty's CEO whether they would ever try to repurchase the bonds underlying the trust preferred issues. He replied that they were probably too small & too difficult to unwind. He also mentioned that he was a little miffed that officers of Liberty were purchasing the trust preferreds in late 2008 and pushing up their price at a time when they were trying to execute a tender offer on the underlying bonds. With the trust preferreds now trading back at levels in November, I feel very happy that I'm in the company of senior Liberty management in thinking they're a great buy.

Conclusion

Liberty Media's bonds and trust preferreds offer the opportunity to generate 15% current yields and 50% capital appreciation over several years. More importantly, the combination of Liberty Media's large cash reserves and the bonds' low dollar prices relative to asset values that fully cover the bonds all provide a tremendous margin of safety. The four third party trust preferred stocks offer even better risk/return opportunities as the trade at over 20% discounts to the recent prices of the bonds they hold. Institutional investors can consider pairing the 8.5s or 8.25s with a CDS hedge or investing in Liberty's exchangeable bonds. While it's difficult to handicap the likelihood of a debt tender (or which bonds might be targeted), Liberty repurchased $1.3BN face value of its debt in late 2008 and has repeatedly stated it's considering debt repurchases, especially in light of the new tax deferral. Even without a tender offer, over the next few years the prices of Liberty's bonds should increase as the economy and credit markets recover and Liberty handles the QVC bank debt maturity.

Risks

Liberty uses cash aggressively, reducing liquidity and/or credit support (stock buyback, acquire assets as in Sirius XM)

Deteriorating economic conditions or credit markets between now and 2011 damages ability to refinance QVC bank debt

Catalyst

Credit-friendly actions, including tender offers for low-priced debt (which Liberty management has recently said they're considering)

Sprint derivatives mature (or borrowed cash is utilized earlier), adding to cash/liquidity

QVC's EBITDA doesn't breach the bank covenant (or Liberty deals with any breach) and ultimately successfully refinances QVC's debt in March 2011

Recovery in credit markets and credit asset pricing