Forget the Earthlink report, buy for the 'free options'

Earthlink's core business stands to post between $190 million and $200 million in EBITDA this year. After balancing out losses from losing operations, such as the company's Helio project, the company should do $105 million to $120 million in EBITDA for the year.

To truly see the value proposition, you must consider the company's balance sheet and comparable companies. Earthlink's balance sheet has almost $100 million in net cash, good for an enterprise value of roughly of about $850 million. Let's assume the company does $113 million in EBITDA -- the mid-point of my estimates -- then Earthlink currently fetches 7.5x this year's EBITDA. More importantly, though, I expect the company to do at least $130 million in EBITDA next year. This puts the company at 6.5x EBITDA. Considering the company's comparables are trading at an average of more than 12x EV/EBITDA, the stock looks like it has very limited downside here.

The way I look at it, the margin of safety constructed in the stock essentially creates a "free option" on any upside in the Helio and municipal wireless businesses in the future. Considering the fact that Helios should have 200-250k subscribers and $140-170 million in revenues for 2007, according to the company, this option is very interesting. The company's municipal wireless business is developing in Philadelphia, Pennsylvania, and Anaheim. On top of that, the company was selected by Alexandria, San Francisco, Atlanta, Houston, and St. Petersburg to build their municipal wireless networks. Considering all this, there are no options to scoff at.

Another way to consider the valuation on this stock, you could value the "core" businesses and the Helio business separately. Helio is essentially a mobile phone network operator that rents cellular towers and high-speed networks from existing providers. The segment has been showing solid growth as a result of continued product innovations and subscriber growth. Although the segment is losing money now, once the infrastructure is fully developed and the cost structure is cleaned up, I expect it to turn profitable. I've seen mobile network operators trading at anything from .8-3x sales depending on profitability. Being conservative, let's assume the business is worth 1x this year's sales or about $155 million. This is conservative because Helio is growing its top-line and subscriber count faster than any company I saw below 1.5x sales.

So you have $100 million in net cash and a $155 million valuation on Helio, thus putting a valuation of $680 million on the core business. Mind you, the core business is set to do about $200 million in EBITDA this year. Therefore, you're getting the core business for less than 3.5 EBITDA. I think this is completely inefficient and I estimate the core business's valuation to be in the 5-6x EBITDA range. Even assuming the low end, 5x, the stock is worth $2.40 more per share.

In this valuation, like the first scenario, the municipal wireless business remains a free option and any upside is not being considered in the stock price, in my opinion.

If this stock is so cheap, you might ask, then how come it hasn't moved in the last year? To explain this I'll reference a stock I worked on last year.

Last year I worked for James Altucher and helped him with his Internet Review newsletter. Altucher's basic philosophy in running this newsletter was finding value in unexpected places, either through dirt cheap stocks or stock priced way below their future growth. One company both of us loved was Keynote Systems, Inc. (NASDAQ: KEYN) -- a cash-rich provider on internet metrics which was spitting off tons of operating cash flow. As you can see from the chart on the right, the stock was stuck in a range for nearly a year but, sure enough, value prevailed and buyers at the $11 per share level now hold 50%+ gains in the stock.

Earthlink, like Keynote Systems, is a perfect example of "time arbitrage." Bill Miller described this concept in Value Investor Insight:

Time arbitrage just means exploiting the fact that most investors – institutional, individual, mutual funds or hedge funds – tend to have very short-term time horizons, have rapid turnover or are trying to exploit very short-term anomalies in the market. So the market looks extremely efficient in the short run. In an environment with massive short-term data overload and with people concerned about minute-to-minute performance, the inefficiencies are likely to be looking out beyond, say, 12 months. That's a pure time arbitrage. You can see the hedge funds saying "It's not going up now. It might go up in December, so I'd better start buying in October and November." Well, that's fine, but I'd rather have it right now than I would when everybody else is trying to buy it then.

Miller is absolutely right -- I'll often talk to investors at hedge funds who say something along the lines of "the stock is cheap but I wouldn't buy until the third quarter when the inflection point hits." While this isn't necessarily the incorrect way of doing things, for true investments I don't believe in trying to time the market that precisely.

Here's the point I'm trying to establish: Earthlink's stock is cheap because investors aren't sure about the future of its dial-up business and don't think now is the optimal time to buy. However, the market is only ascribing a 3.5x EBITDA value to the core business, therefore all the downside is priced in unless the business goes bankrupt tomorrow. Any upside in the core EBITDA, Helio growth, voice subscriber growth, the municipal wireless business, and so on are all potentially "free options."