At the current price of $8.71, Earthlink offers a substantial margin of safety with a decent potential for mid-double digit annualized returns over a 1.5-2 year period (with minimal dependence on the broad market's movement). Other than cash generation, the primary catalyst for unlocking the underlying value should occur following the company's first option to redeem its 3.25% convertible notes during November 2011. At that point, I believe the company will fully redeem the notes and then issue a sizable cash dividend.
ELNK's major business segments include providing narrowband (dial-up) and broadband internet services. As a side note, they separate operations into consumer/business groups-with consumer accounting for 90%+ of operating income. Over the past 2+ years the company has been divesting from some terrible business ventures which came about through the previous CEO; please see pat110's write-up from 2/13/08 for further background on these investments. Since that point the company has also monetized their Covad bonds for $6.1M and exchanged their stake in Helio for Virgin Mobile shares (subsequently acquired by Sprint). As would be expected, there has been relatively high attrition in the company's dial-up customer base (which has shrunk from approximately 2.6M in 4Q07 to 1.2M in 4Q09). The company has also reduced marketing expenses, which has perhaps predictably resulted in declining broadband subs which were down 10% YOY. There is no doubt that the former core narrowband business is a "dying", and most traditional investors will typically avoid these types of companies.
Although the business characteristics are abysmal, there are a handful of factors which make ELNK a relatively compelling investment opportunity on a risk adjusted basis.
(1) Both the narrowband and broadband lines generate a substantial amount of free cash flow, and as a result of the new CEO, ELNK has stopped flushing money down the toilet. They've typically been conservative on estimates and are currently guiding for $180-190M in EBITDA and $160-180M in FCF for 2010. These are similar to strong cash flow characteristics which have also been present in other run-off situations like the paging industry.
(2) Very liquid balance sheet including cash at year end of $611M, along with govt agency notes at $42.1M plus auction rate securities of $48.1M including a put-right (which allows them to sell the ARS back to the broker at par + accrued interest on 6/30/10).
(3) As subscribers become more seasoned, their churn rates drop significantly. This dynamic ends up creating a longer tail than would be expected by using historical averages, even for the higher churn narrowband business, and allows ELNK to harvest cash flows for a greater period of time. For instance, first year broadband subs have a churn rate of 6% versus 1% for similar third year cohorts.
(4) Within the consumer segment, which accounts for approximately 80% of revenue, broadband subs already account for 60% of revenue. Because the churn with narrowband runs much higher, ELNK will eventually be even more skewed toward broadband. This may gradually have an impact on EBITDA margins since broadband has a bit lower margins, but I believe it will make the company a cleaner possible acquisition candidate-there's probably less interest from other players in possibly taking over the narrowband subs.
(5) Management has already started returning capital to shareholders by buying back stock and initiating a dividend (6.4% yield currently). They've also been fairly cautious about possible acquisitions. I believe they looked at potentially acquiring or merging with AOL's dial up business, but were perhaps too low or realistic on pricing.
(6) More recently, management has suggested that they'll either look to scale up the Web Hosting business in order to make it more cost affective or just sell the business. This should end up being a minor positive under either scenario because it's not contributing much in terms of EBITDA or FCF.
(7) Due to the losses incurred on discontinued ops, ELNK still has $350M in federal and $178M in state NOL's. As a result, they should pay minimal taxes through 2011.
With slightly over $700M in cash + marketables on their balance sheet, and the possibility of generating an additional $160-180M during 2010, the main question is: "what are they're going to do with all the cash?" At this point, they've managed to stay away from paying up for subs despite some potential opportunities. This seems to correspond with management's indications that they're not interested in paying high prices for acquiring other businesses, when they can just buy their own stock for a better value. In addition, share buybacks and the implementation of a sizable dividend starting in 3Q09 also suggest they're interested in maximizing shareholder value. However, as referenced previously, the primary catalyst for highlighting the company's value should occur after the convertible bonds are eliminated from the capital structure.
In Nov06, ELNK issued $258.8M of convertible bonds which now have an exercise price of $8.82 (subject to adjustment for dividends). These can be redeemed on Nov 15, 2011 for cash, at the company's option. Assuming continuation of the current dividend policy, the bonds will be convertible at $7.98. Due to the relative size of this issue, it actually creates the incentive for the company to keep its stock price low until after its converts are redeemed-for each additional $1.00 the stock appreciates, it will end up costing ELNK an incremental $32.4M. So, I believe the company intends to hold off on issuing a large special dividend until after they redeem all the convertible bonds. This would also suggest that they'll hold off on additional share repurchases unless the stock starts trading down a fair amount.
For the purpose of this analysis I'm going to assume that ELNK will close on the initial redemption date at $9.50 and s/o are consistent with today's 107M, although the pricing will actually look more attractive the closer it is to $7.98 and lower the share count.
So using what I feel as conservative projections, ELNK would then have $5.64 in net cash. They could pay out a special dividend of $5, leaving an adjusted price of $4.50 (enterprise value of $413M including $68M in cash). At that point, the company would be mostly a broadband ISP that generated around $140M in EBITDA over the trailing year; churn would get very low as subscribers became more seasoned and skewed toward broadband. At that point, ELNK's stub would likely be a decent acquisition target for the broadband subs. If we then apply a conservative multiple of 4.0x EV/EBITDA, then this stub would be worth $5.87.
There are certainly a fair amount of assumptions involved with this valuation, but hopefully these inputs are conservative. The cash generative abilities of this business combined with substantial cash on the balance sheet should help to provide a large margin of safety, while the possibility for lower churn (or a sale of the Web Hosting business) would create some upside. Also, the company could have some other options like slightly levering the capital structure. But based on a total return over a 2 year period these base assumptions could provide $0.84 (regular dividends) + $5.00 (special dividend) + $5.87 (residual value) = $11.71 or a 16% annualized return.
Another possible question could be, "why not wait until much closer to the actual event?" since much of the return would be driven from re-pricing of the stub. The stock provides a 6.4% dividend yield plus potential price appreciation prior to convert redemption, while the business itself should be expected to have downside protection in a double-dip (through slower narrowband churn), I feel it's a moderately attractive position to hold in the interim. However, if ELNK's subs and EBITDA continue to stay predictable while the share price stagnates or drops over the next 1+ year, then it could easily turn into a more attractive IRR.
Sale of Web Host business
Redemption of convertible bonds, in conjunction with special dividend or dutch auction
Recapitalization or acquisition of "stub" company after converts are redeemed