Description
Dobson (DCEL) - $5.99 as of 11/20/03
I'd like to introduce a rural wireless carrier called Dobson Communications. The stock has taken a tumble lately after the 3rd quarter earnings. It has about $1.60/share of cash on hand today and is generating about $0.87/share of equity free cash flow a year (2004). At this rate, the company will have over $2.50/share of cash on hand by the end of 2004. Depending what the company decides to do with the cash, the company would be trading at 4x-6x P/FCF multiple by the end of 2004. The first obvious question to ask is, what causes the stock to trade at such multiple? Although there is not much Wall Street coverage on this stock, it has recently absorbed a pretty broad institutional investors base, among which includes sophisticated funds such as Ouch-Ziff. We should not simply assume that it is one of those misunderstood neglected stocks.
WHAT HAPPENED?
--------------
The rural wireless sector was a "well-misunderstood" sector in 2002. A majority of analysts were extremely bearish on the sector, citing overbuilding risk by national carriers, rapidly declining roaming revenue and liquidity problems at most of the rural carriers. As the sector evolved in 2003, much of the concerns have been proven unwarranted and most analysts have caught on and recognized the stable cash generation of these businesses. Companies such as Western Wireless, Centennial and Dobson Communications have all skyrocketed from its low, generally trading around 7x EBITDA today. More importantly, people have also recognized the sustainability of their capital structure (around 5x leverage) and started valuing the businesses based on their substantial and growing equity free cash flow.
Over the last few months, Dobson has acquired its formerly 50% owned subsidiary American Cellular at the brink of bankruptcy, refinanced its entire capital structure, bought back a large amount of its high coupon preferred securities and signed long-term GSM roaming agreements with Cingular, T-Mobile and AT&T Wireless, leaving most of the issues behind the cmopany. Just when everyone is comfortable with the company's prospects, it announced a quarter of weak earnings on 11/10/03. While the core subscriber business look very normal and healthy, the ultimate fear about declining roaming was reinvigorated. Roaming revenue saw a run-rate decline that exceeds 15% year-over-year in the quarter. Caught by surprises, a lot of people turned bearish on the stock and even expected the company to miss its 2004 guidance. As I looked into the results closely, it seems to me that there are a lot more to the seemingly bleak picture.
ROAMING REVENUE DECLINE
-----------------------
The recent drop in share prices is entirely about the risk in roaming revenue. Understanding that issue will get you 90% of the pieces in this puzzle.
To understand the picture here, you have to look at roaming revenue as a stream that consists of 2 components, TDMA roaming minutes and GSM roaming minutes. Roaming revenue is the product of roaming minutes of use (volume) and the roaming yield (pricing). Since Dobson is still deploying its GSM network, it has received no GSM roaming revenue thus far. TDMA roaming minutes of use (MOU) has historically grown at a rapid pace (well exceeding 30% y-o-y) due to widespread adoption of national plan by big carriers. But as most TDMA carriers migrate their subscriber base to the GSM technology, the growth in TDMA roaming MOU has quickly decelerated. The y-o-y growth of Dobson's TDMA roaming MOU in the first 3 quarters in 2003 are 31%, 21%, 12%. You get the picture. It is the alarming deceleration in growth that has caught most by surprises. Compounded with the already well-known decline in roaming yield (from 24 cents in 3Q02 to 18 cents in 3Q03), this has reinvigorated the long-suppressed fear about declining roaming revenue, as the growth in TDMA MOU fails to make up for the declining roaming yield.
The bears correctly noted and reacted to a 15% decline in roaming revenue. They're also probably right in predicting that roaming revenue will suffer from bad yoy comparison in the next few quarters. With some pretty detailed guidance that the company has for the first time provided on the roaming yield contractual stepdowns, I am projecting that roaming revenue in the next 4 quarters will see yoy decline of 18%, 18%, 16% and 1%. While all these look gloomy enough, once you look beyond the obvious and put things in perspectives, one could still take comfort in the long-term stability of the roaming revenue.
What's missing here is the GSM MOU. Once it starts kicking in in the 4Q03, the picture isn't nearly as bleak any more. Although GSM MOU yields significantly below the current TDMA roaming yield (low teens versus 18 cents), GSM could account for 274 million MOU in 2004 and growing at an explosive pace. About 15% of AT&T Wireless' subscribers and 35% of Cingular subscribers are already GSM-based, and T-Mobile has nothing but GSM subscribers. Although only GSM subscribers with dual-band phones can roam on Dobson's 850MHz network, given the trend of dual-band phones adoption GSM roaming MOUS should still account for about 14% of 2004's total roaming MOU. That is how I arrived at 274M minutes in 2004.
Of course the real reason for the $16M decline is not just the loss of MOU to GSM platform. The main reason is actually the 23% contractual stepdowns in existing roaming contracts that are not offset by TDMA MOU growth. Harm done is already done. Fortunately there are only two stepdowns left in the contracts, one in December 2003 and one in July 2004, stabilizing roaming yields around mid teens thereafter. Therefore, although the huge July 2003 stepdown would cause the yoy comparison of the next 4 quarters look bad, roaming yield outlook are quite stable beyond 2004 (note that there is no stepdown for GSM MOU).
With the new insight about future roaming yields that the management unusually disclosed recently, we can actually make pretty good forecasts of roaming revenue. Following are the assumptions that I made in forecasting roaming revenue:
- TDMA roaming yield would stabilize at 15 cents after 3Q04 (although management guided closer to 16 cents)
- GSM roaming yield would be 12 cents without stepdowns (management guided that it is 30% below current TDMA rate of 18 cents)
- Total combined romaing MOU in 2004 would exceed 1.9 billion and the overall pie would grow by at least 5% a year (which is very conservative given that national subscriber growth alone would be 5%, and that minutes growth usually exceed subscriber growth by an order of magnitudes)
If these assumptions are valid, which I think they are, I believe there wouldn't be a scenario in which the roaming revenue declines after 2004. It doesn't matter how you divide the MOU pie between TDMA and GSM, as long as the total pie grows more than 5% a year, the roaming revenue is not going to go down. In fact, under what I think are conservatively realistic assumptions, I'm projecting $286M of roaming revenue in 2004 growing to $312M of roaming revenue in 2006.
Once you have the conviction that roaming revenue would have a stable outlook beyond 2004, the thesis of the idea would be very obvious. Just like Western Wireless, the non-roaming core business itself is also a very stable business even with some growth potential. The bottom line is, if it is a stable business and the EBITDA and hence free cash flow would unlikely decline over the long-term, there is no reason a stock should trade at 5x P/FCF. Below we will spend a little time on the core business and then go through the thesis.
SUBSCRIBER BUSINESS
-------------------
There isn't much disagreement over the rest of Dobson's business. Like Western Wireless, Dobson competes in rural America. Dobson operates in the more sub-urban end of the rural spectrum, thus facing slightly more competitors than Western (most of the time Dobson sees 2-3 other competitors). The sub-urban areas have higher population density and more desirable demographics than the rural states such as Montana. Sub-urban areas also bring more substantial roaming revenue. The flip side is that the proximity to national carrier's core footprint might invite overbuilding easily. But the exclusivity agreement with AT&T Wireless and various long-term roaming contracts with other carriers have mitigated this risk in the years to come.
In arriving at the projections for the subscriber business, I have made the following assumptions:
- churn would go up 25 bp from the current 1.8% due to wireless number portability
- virtually no subscriber growth, but 1% annual increase in ARPU
- I'm not factoring in any cost savings, particularly the stepdown in incollect roaming rates (the rate that Dobson pays roaming partners), which we know will go down with the outcollect rate
- I'm not giving them credit for any USF funding, which they will probably get in 2004 (about $10M subsidy from the government per year, give or take)
FINANCIAL PROJECTIONS
---------------------
2004 2005 2006
Subscriber Revenue $ 764 $ 774 $ 782
Roaming Revenue $ 286 $ 295 $ 312
Other Revenue $ 52 $ 52 $ 52
Total Revenue $1102 $1120 $1146
EBITDA $ 488 $ 500 $ 516
FCF $ 125 $ 137 $ 153
FCF/Share $0.87 $0.95 $1.07
In calculating FCF / share, I'm not giving any credit to the $230M cash on hand and the cash that the company will be generating in the next few years. For sake of simplicity, I also assume the company will not redeem the $284M of high coupon preferred that will be callable into 2004, which is highly unlikely. I'm also not factoring in the $25M ($0.17/share in FCF) GSM upgrade capex in 2004 that will disappear in 2005. All these will produce additional snowball effect to the company's FCF growth.
THESIS
------
Stretching projections into 2006 begs the obvious question to their accuracy. But what's more relevant here is the long-term trend of EBITDA and FCF, which in this case is reflected by the growth in 2005 and 2006. I feel quite comfortable with the assumptions that I made in arriving at the upward long-term trend of EBITDA/FCF. This conclusion is extremely important because a stock that trades at 4-6x P/FCF is cheap only if FCF is not going downhill.
Between the cash on hand and cash generated in 2004, the company should be able to redeem all the straight preferreds, bringing another $0.25/share of FCF accretion, leaving the stock at ~5x P/FCF. Chances are the company would not use the entire cash balance for it and would rely on financings from a stock/convertible offering or the $150M undrawn revolver, which would slightly inhibit the accretion from redemption. But whether the company redeem preferred, buy back stock or just pile up the cash, a stable business that provides FCF yield of 20% is still a very attractive investment with little downside. Any upside from my conservative assumptions, USF funding, refinancings and GSM growth would provide the gravy.
RISKS/ISSUES
------------
Disappointing quarters to come - Until the stepdowns are all behind us, we would continue to see weak roaming revenue yoy comparison the the next few quarters. Therefore the stock might have room to go down further. But the stock has gone down to a level where I think the value/upside outweighs the downside. I would start accumulating and leave some room to add to the position.
Overbuild - No bear would finish his argument about this industry without talking about the risk of being overbuilt by national carriers. When I invested in Western Wireless, I have done a cost/benefit analysis that concludes that it does not make economic sense for national carriers to overbuild in rural areas. The subsequent roaming agreements signed between various rural carriers and national carriers confirm this. Nowadays most people agree that the overbuild risks are alleviated at least in the medium term. Why would national carriers sign 7-year contracts if they intend to overbuild shortly?
Number Portability - Number portability will be enforced in ubran markets starting 11/24/03 and in rural markets starting 5/24/03. The effect of this is clearly detrimental to the whole industry. However, rural markets are just not as competitive as the cutthroat urban markets and hence the negative impacts shouldn't be as profound. Moreover, we should not lose sight of the simultaneous enforcement of landline-wireless portability, which would likely benefit rural carriers in a significant way. My expectation is that a 25bp deterioration of churn would adequately reflect the risk.
Taxes - Due to the NOL on hand, the company won't be a material taxpayer until the end of 2005. Since the depreciation is much bigger than capex, taxable income will remain significantly lower than FCF. Even discounting for future taxes, the company still looks cheap.
Liquidity - Besides $230M of cash on hand, Dobson has an undrawn $150M revolver. For a company with ample free cash flow and no maturity before 2008, this is more than adequate liquidity.
Catalyst
- As GSM roaming minutes kick in in the 4Q03, the picture of roaming revenue will drammatically improve
- I believe a lot of the investors base have lost perspectives of the value this stock offers and the trend of FCF growth. You never know when people stop being overly pessimistic but the pessimism provides a good potential catalyst when the worst case scenario is priced in.