This opinion is based on our analysis of the value of the marginal customer Vonage adds to its network. Today, each customer Vonage adds to its network loses them money, and incrementally each new customer loses them more money than the last. We see nothing in the business plan or financials to support an argument that Vonage will be able to reverse this trend ever, let alone in time to arrest a coming cash crunch in 2008.
With negative cash flow, no change to positive cash flow in the foreseeable future, and a voracious appetite for dollars to support its marketing efforts, we believe that VG equity is a loser at any price. We are short the stock and long some puts and believe there is considerable upside to these positions from current prices.
Most recently we have seen the stock rally about 20% from a low of $6.40 to $8.10 on the back of a number of tepid sell-side reports that have established price targets ranging from $7.50 to $10.00 based on discounted cash flow [DCF] analyses. As we believe that the long term value of VG equity is $0 as a stand alone business and less than $3 on a take-out basis, we see this as an opportunity for investors who feel they may have missed the initial move in the stock down from its IPO price.
The Problem with Vonage – New Customers Destroy Value
The problem with Vonage is that its marginal customer creates negative value for the firm – i.e., the cost of customers acquired today is never recouped by the EBITDA that those customers generate before they churn off the network. This is exacerbated by the fact that this negative contribution is incrementally increasing with each new customer as the cost of reaching the marginal customer becomes more expensive over time (the easiest customers are found first). Finally, since VOIP has moved through the early adopter (meaning the easy customers have been found), there is no evidence that the trend that is driving the negative NPV will reverse itself at all, let alone move into positive territory before the company runs out of cash.
This means that each new customer VG adds never creates cash flow for the company, and even worse, costs more money incrementally because the cost of adding customers is growing faster than the EBITDA margin improvements coming from the “scale” in Vonage’s business. This is exactly the opposite of what you want in a business that is supposed to “scale” into its financial model to justify a DCF valuation.
Many believe the key to the Vonage story is that it can add subscribers and by burning IPO cash and out run its churn rate. But this dot.com-esque view of VG’s prospects misses the point. The real key to the Vonage story is if it can it show that its marginal customer can turn a profit before the Street decides to stop giving them money to fund that experiment – because by our estimates the company will need a cash infusion no matter what before the end of 2008 – but the Street is unlikely to pony up much more cash for Vonage in the wake of its disastrous IPO without some evidence of improved fundamentals. The clock is ticking and if the company cannot show that the business will scale over time as advertised the funding will dry up and the company will go into a cash death spiral starting in 2007.
Vonage’s problem isn’t that Vonage’s customer churn rate is 27.6% per annum – a number that is within reason for a telecom company and actually pretty good for new telecom company (and therefore unlikely to improve significantly). It’s that Vonage has small EBITDA margins, little in the way of levers to improve those margins, combined with a high cost of customer acquisition. With that combination, ANY churn rate of significance will be toxic – as we will show.
Why Vonage is Different from Ma Bell or CLECs
Unlike other providers of communications services that use their own facilities to service customers – where scale is met with drastic improvement in COGs and EBITDA as network utilization rates increases on a fixed operational cost, Vonage resells other companies’ facilities in a non-facilities business model that has relatively inelastic COGs margins – but a variable operational cost. Vonage’s network operations cost scales with each customer added. From a capital standpoint a non-facilities approach is attractive – especially when a business is small – as you don’t have large fixed up-front costs as a barrier to market entry. The downside to a non-facilities approach, however, is that margins don’t improve markedly with scale. This means that there is far less advantage from scale for Vonage than a more traditional provider – and far more risk from churn as the cost of adding the incremental customer increases but net EBITDA margins stay relatively flat – and in Vonage’s case, small.
The math for the customers acquired during the last quarter works out as the following:
- Number of Lines Acquired in Quarter: 377,005
- EBITDA per line per month: $5.21
- Cost of Acquisition Per Line: $239
- Percentage of Lines Lost Per Month: 2.3%
- Average Customer Life: 3.6 years
- Number of months it will take for those 377,005 lines to turn a profit after accounting for the cost of acquiring them: 66,755 (and that’s before taxes and depreciation)… just a wee bit more than the average customer lifespan.
Now admittedly this is just a snap shot in time… but it stands to note that the negative contribution of the marginal customer has deteriorated since the IPO as the company has scaled up. Bulls will argue that this doesn’t consider the benefit of improving business conditions looking forward – which we believe a) won’t improve as much as projected to support the DCF valuations and b) ignores the fact that the margin of error in our analysis of churn rates is so large that no amount of projected business improvement will stave off a liquidity crisis sometime in 2008.
Let’s start with churn. There are really only two numbers that we feel can move enough to make the Vonage business model profitable from current levels – lower churn rates or lower per line customer acquisition costs [CAC]. For churn rates, the margin of error in our analysis is large, e.g., if the churn rate were only 1.5% (which is lower than the best in the industry and meaning that we were off by about 35%) the payback on the 2nd quarter customers given current EBITDA rates would still be 78 months (6 years 6 months – still well in excess of the average customer life span). Even using the longer term projected “mature” churn rate published by Deutsche Bank (an underwriter of the IPO and therefore likely an optimistic estimate) of 2.18% we still get a negative NPV on customers at current EBITDA margins... Ultimately all of the math on churn nets a negative NPV per customer on their current subscriber additions.
Furthermore for any improvement in churn to improve the marginal NPV, it would have to be done without a corresponding increase in operational cost or lower average price per unit [ARPU] – either of which would partially offset the benefit of lower churn by compressing the EBITDA margin. Lowering churn without increasing service costs or lowering prices relative to the competition is something that is difficult for even the best run companies.
In the second quarter of 2006 Vonage’s acquisition cost per line was $239 – the highest to date. Long term estimates for customer acquisition costs are in the $218 range. We think this is optimistic because the marginal customer tends to be more expensive to find and not less (you pick the low-hanging fruit first) and Vonage is already in the range of industry norms for this line item. Vonage and other VOIP players have tapped the early adopter market and now are in the same highly competitive boat for customers as all retail telephony suppliers. Assuming that the company is only able to keep acquisition costs near where they are today as they grow rather than shrink them, the company will not reach the necessary margin improvement to make new customers profitable.
Retail telecom operations are like sharks, they must keep moving forward and adding new customers, or the churn will overcome them and they will sink to the bottom. Part of the reason for this is that churn is not a straight line event. It is industry standard for churn to be heavily weighted to the early portion of the customer lifecycle. This means that the most expensive customers in Vonage’s network also tend to be the most likely to leave. Cuts in marketing costs that lead to lower subscriber additions result in a rapidly shrinking subscriber pool – not one that shrinks at the 2.3% per month churn rate but at a much higher rate. Perhaps the company can squeeze significant efficiencies from its marketing efforts without impacting growth projections – but we doubt it. Even if they can, we doubt they will be significant enough to offset a more realistic forward churn rate of 2.3% per month. In order for that occur they would need to cut customer acquisition costs by 10% per customer addition (on top of the already expected decrease in customer acquisition costs factored into analyst models) without any impact on projected customer growth rates. Not likely.
Based on the above, we believe that the odds of Vonage ever reaching EBITDA cash flow as an entity are slim. Holding all other factors equal, as each incremental customer becomes more expensive to find it is reasonable to expect that the yield per marketing dollar spent will deteriorate and the rate of cash attrition will only increase from current levels either through higher marketing costs or through lower subscriber adds.
In our estimation it is only a matter of time (7 to 9 quarters) before the company runs out of cash – and a few less before it become obvious to all that it will… and this is before we include the potentially devastating cash impact of Vonage’s ongoing lawsuits regarding its IPO and patent infringement, ESF/UCC contributions, or the impact of net neutrality on Vonage.
The Bull story or “We All Wanted to Go Back to the 90’s Anyway”
Despite what we have outlined above, the valuations supported by the banks still hover in the $7-$10 per share range. How do these analyses net a positive valuation when our analysis of customers added is always negative? Let’s take a look.
In order to reach positive NPV in their DCF calculations, the sell-side analysts need to show cash flow on the marginal customer. They do this through a process that we call “putting the company on a knife’s edge.” In other words, there are small (and some in their own right reasonable) improvements to all aspects of the business to eke a small positive EBITDA margin in the out years of their models (where its hard to challenge the underlying assumptions) – in this case starting in late 2008/2009. That margin can then be extrapolated into future cash flows to discount back to present value in the DCF model.
* Avg. of Bear Stearns and Deutsche Bank Estimates
By putting the company on the knife’s edge, the analysts have set a very tough bar for Vonage, one where any single failure to meet assumptions and expectations will destroy the DCF models’ assumptions and expose the company’s cash flow weakness.
Even with such optimistic views on Vonage’s ability to improve its business, according to Bear Stearns (an underwriter), the company will reach its low point in cash in late 2008 at about $13MM before beginning to grow cash. This is a razor thin margin for error. A mere 3% miss in the number of customer net adds for 2008 (or 30,000 out of 1.265MM) would cost the company more than that amount. Pushing back EBITDA breakeven for even one quarter would bankrupt the company. This sets the stage for what we feel will be a host of potential negative catalysts to drive VG stock lower from current levels.
While we are skeptical of use of the DCF valuation that sell-side firms have put on Vonage (but not surprised as any other form of valuation would net a big fat zero for the foreseeable future), we are even more skeptical of the series of dubious assumptions that must be made to get the highly compliant DCF to generate a terminal value significant enough to discount back to current share prices. In order for the stars to line up for VG ALL of the following must occur:
1) ARPU Cannot Fall – Let Alone Collapse: Raise your hand if you think you will be paying the same for basic telephony in 2008 as today… I didn’t think so. But that’s what analysts are expecting in order to justify the DCF valuations for VG. This is by far the most difficult of the “knife’s edge” assumptions to swallow. A host of established players have already matched Vonage’s $24.99 price, making customer wins away from their existing provider more difficult. Furthermore, a handful of independents, such as Sun Rocket, which recently raised $33MM, are rolling out even less expensive services. Sun Rocket provides a year’s service for $199 without any additional fees and taxes – that’s $16.58 per month compared to over $30 when taxes and fees are included in a Vonage bill. Telephone bills don’t need to go to $10 for it to be a disaster for Vonage – a modest 10% decline of ARPU by 2008 to $23.50 would result in a liquidity crunch late in 2008 based on analyst’s models. If ARPU’s fall to $20 the liquidity crunch would come mid-2008.
2) Good Churn Rates Must Decline to Industry Best Numbers without Higher Costs: The reality is that Vonage already has churn rates that are pretty good. The best wireless churn rates are between 1.6% and 2.0% and they have the advantage of equipment-tied contracts to lock up customers through the early months of service (when customers are most likely to churn). Long distance providers like Vonage generally have churn rates of 2.3% and above. An estimated churn rate of 2.18% would be an industry best – and highly unlikely as Vonage has the disadvantage of “piggybacking” on occasionally unreliable DSL or Cable Broadband services. To lower their current churn without increased costs or lower ARPUs strikes us as highly unlikely. If churn rates stay steady instead of declining as expected the company will never reach the EBITDA inflection point necessary to avoid running out of cash.
3) SG&A Expense ex-Marketing Must Improve by Nearly 20% per Line: This is one that may be achievable, but we aren’t certain. As SG&A lines tend to be the “Watergate slush fund” of corporate income statements we don’t have a good view on whether this number can be achieved or not. There is some speculation that a portion of the company’s customer acquisition costs are included in SG&A line to mask the true cost of CAC. If this were true then we would not expect SG&A to be as fixed as one would expect – making a 20% decline less achievable.
4) Equipment Subsidies Must Decline by nearly 28%: No one who signs up for Vonage service is going to pay for the box to hook it to the Cable or DSL service. Assuming that it costs only $3 per box to ship a customer install kit (aggressive in today’s high fuel cost world) this means that by 2008 Vonage will need to source its modems for about $15 per unit assuming a free modem for new customers (they are paying $22 in 2006 based on crude analysis of the equipment subsidy) – a possibility but still a stretch in our mind. Each dollar more than $15 per unit is at a minimum $1.47MM in additional costs per year in 2008 based on current estimates of customer additions (assuming that Vonage only needs to provide modems for their net added customers and somewhat higher due to non-returned or outdated returned modems.)
5) Customer Acquisition Costs Must Decline by 12%: This is the book end to our ARPU discussion as to a degree these items are linked, e.g., lower customer prices (on a relative basis) make marketing more effective but result in a lower ARPU. Analysts believe that CAC will improve as the Vonage brand grows and expectations are for it to level out where they are for other telecom providers – 20-30% of revenues. This ignores some important points – 1) Most of the brands that Vonage competes with are far more established brands and therefore have efficiencies of reach and imprimatur of quality that Vonage lacks; 2) Those brands have multiple product lines which help drive cross-product efficiencies that Vonage does not have with its single product line; 3) Those operations are much larger so they are spreading their mass marketing expense across more customers and revenues; 4) The cost of the marginal customer addition tends to be more difficult (and more expensive) as a market is penetrated. If Vonage is unable to cut CAC costs – and they instead merely hold steady at current levels this will cost Vonage $33MM in additional costs in 2008 – again creating a cash liquidity crunch in late 2008.
Failure to meet any one of the above expectations (except for the equipment subsidy) will likely result in a liquidity crunch rather than future cash flows to discount in a DCF. If the two more likely scenarios play out – lower ARPUs and no reduction in CAC efficiency – that liquidity crunch could come as early as the 1st quarter of 2008. If, in addition to the two points just listed, churn rates do not improve as modeled, VG could run out of cash before the end of 2007.
The Other Problems with Vonage
Now usually the above should be enough illustrate the lack of long term prospects for Vonage – but as they say on late night TV, “wait there’s more”…
Not factored into the above are another 3 critical challenges that Vonage faces – none of these are factored into current expectations, but all stand to create a negative impact for VG shareholders. If you needed more reasons to believe that VG stock is in trouble, here they are:
1) Net Neutrality: Vonage could become a bystander casualty in the battle between the telco and cable providers and content/commerce internet sites. The infrastructure providers want a piece of the content/commerce pie generated by Google, eBay or YouTube and are pushing congress for the ability to “tier services” providing differentiated performance (bandwidth throughput and latency) to different sites and customers across the internet. This would be devastating to Vonage because, while the amount of bandwidth used to move the voice packets that make up a telephone call are small, they are very sensitive to latency. Too much latency and your phone call ends up sounding like a call to the space shuttle – with long pauses. Should the telcos win this battle (and I would rate it a toss up right now), the telcos could discriminate based on customer tiering and packet destinations. It’s not hard to see where Vonage would either suffer serious quality issues for its traffic, or have to pay more for “higher tier” bandwidth to insure the highest quality routing of customer calls across the internet – either of which would negatively impact expectations and create a liquidity crisis.
2) Lawsuits: Vonage right now faces two groups of lawsuits. The first is from patent holders for VOIP services and technology. Sprint and Verizon have filed infringement lawsuits seeking injunctions, compensation and treble damages for how Vonage uses VOIP to provide services, and Klauser Technologies has sued Vonage over the technology used to provide Vonage voicemail services – seeking $180MM in damages. Additionally, there are a number of IPO-related lawsuits now outstanding against the company for which the company has made no allowances in their projected financials. While the outcome of any of these suits is undetermined it should be noted that merely defending these suits may be enough to eliminate the $13MM cash “cushion” that the company currently enjoys before it starts generating cash in a best case scenario. An adverse judement in any of these suits, even where Vonage need only post an appeal bond, could be a death sentence for VG – accelerating a liquidity crunch by several quarters or worse.
3) USF/ICC Compliance: The final “wild cards” for Vonage are the current proposed rulemaking by the FCC that would require VOIP providers to pay into the universal service fund and potential changes to inter-carrier compensation agreement. The former would require Vonage to contribute 10% of interstate and international revenues (determined by a formula) to the USF – this would likely be passed through to the customer as a surcharge – but that eats into Vonage’s price advantage over competitors – eroding marketing advantage based on price. The latter, the ICC issue, is more troublesome as it would likely further erode Vonage’s EBITDA margin. The ICC is a proposal to provide ILECs that terminate VOIP calls (currently free of charge) compensation for that termination. This could be as much as $0.01 per call – and would hit Vonage hard on the cost side of their EBITDA margin.
While it is entirely possible than none these additional “woes” could impact VG, any one could be the straw that breaks the camel’s back with regard to the fine balancing act that is currently being used to defend VG’s stock valuation.
The Trouble with the “Street’s” Valuation
That the banks would defend VG with a DCF is not surprising – after all, they only made $37MM in underwriting fees from Vonage’s IPO. Traditionally DCF valuations have been used for communications service providers. In fact, for companies with recurring revenue models and high initial capital costs, a DCF may be the only way to properly assess the value of the assets being built by the company during the early phases of business deployment. In the case of telecom service providers these assets are network facilities – facilities that have a useful life of in excess of 15-20 years.
Additionally, DCFs are most useful when long term business growth rates, pricing and margins are somewhat steady – providing a measure of stability in the out year projections of cash flow and terminal value that are then discounted back to find present value. The problem is that Vonage’s business lacks any of the characteristics that lend themselves to reasonable DCF analysis. The company’s only true asset is its customer base (which is temporal at best) and the dynamics of pricing, margins and growth rates are far too volatile and subject to overly optimistic assumptions by the analysts to make accurate predictions of the state of business 10 years into the future.
So why do the banks use a DCF? Simple, it’s about the only “traditional” valuation technique that would net a valuation for Vonage anywhere near current share prices. The DCF is built on a host of assumptions by analysts - assumptions that can be tweaked and tortured to output a valuation that supports current share prices. Because of the long term nature of the DCF even small changes to the inputs, a little stretch on the gross margin here, a little trim of SG&A expenses there, can have large impacts on the terminal value and net present value [NPV] calculation. A couple of changes, and voila… all of a sudden a company with a negative present value is worth $1 Billion dollars.
The underwriters of this company, having collected their underwriting fees, and now obligated to “cover” VG, have backed into a DCF-based analysis that justifies current valuations using all kinds of caveat emptor language and warnings – sell side code for “stay away.” But they have had to because it was the only way they could defend taking the company public in the first place.
The biggest problem with using a DCF for a company like VG, (as anybody that covered emerging internet infrastructure companies in the late 90’s will tell you… trust me, I was once on the other side), is that almost the entirety of the value of the company computed by the DCF is contained in the terminal value – 10 years into the future. This not only makes the DCF highly sensitive to terminal growth estimates (an estimate based on where the analyst thinks the company’s growth rate will be 10 years from now – so that’s really a guess not an estimate), but it also says that you are betting that despite creating negative value over the “short term” (10-years in this case) that the long term prospects for this company, based on the long term nature of assets being developed during that “trough” of net negative cash flow, will have some measurable value based on current assumptions for the business 10 years out. This is a bet we believe that no investor should make as the long term prospects for VG – as outlined above – are almost certainly not as rosy as the assumptions the analysts are inputting into their models to get their DCF to show a positive value; and the long term viability of its core asset, its customer-base, is questionable at best.
What about a Takeover?
The last argument for value in Vonage’s stock is that the company is “building a customer base” that someone will want to buy. Really? We really don’t believe that the VG customer base is an attractive acquisition for the traditional service providers, and even if it were, it certainly would not be anywhere near where the stock trades today.
The biggest hurdle for likely suitors of Vonage is the earnings dilution that the acquisition would have on the suitor’s stock. ILECs and Cable companies are very sensitive to earnings. If you don’t believe me, ask Northpoint shareholders what happened to their deal and stock after Verizon shareholders found out about the Northpoint/Verizon merger deal. After Verizon broke their deal, Northpoint ran out of money and declared bankruptcy – there were no takers for the business as a whole – ultimately selling the network for $135MM – no one wanted the customers. In the end, equity holders received less than $0.10 for equity that had been worth more than $14.00 when the deal was struck.
The only way that we think Vonage is an attractive acquisition would be at share prices far below current levels. Right now VG trades on an enterprise value of about $480 per subscriber line – considering that potential suitors can acquire customers for about $200 per sub, that customers are a “temporal asset” due to high churn rates, and that there are very low barriers to entry for VOIP service provision - with its attendant risks on pricing assumptions, we view that as a reasonable proxy for the maximum take out value for VG. At $175 per sub on an EV basis the implied value of VG would be $3.20 per share.
Having covered communications services for over 17 years and lived through the internet infrastructure bubble and burst as a sell-side analyst – VG is, to quote Yogi Berra, “like déjà vu all over again.” It is as if Vonage investors and management think that someone has turned back the clock to 1997. We expect that the banks will continue to value VG on a DCF basis for the foreseeable future because they have to, and enough of the Street will ignore the warning signs about VG’s impending cash crunch to make VG a good candidate for short selling and put buying in the near term.
The difference between now and 1997, though, is that VG doesn’t have the luxury of a technology stock boom as a tailwind to put fear in short sellers based on irrational stock performance, breathless sell-side analyst support or i-banks clamoring to provide easy cash to refill the company’s depleted coffers regardless of profit performance. Today’s market is far more skeptical and will be watching the metrics at VG closely – with the company already on a knife’s edge in order to support current DCF valuations – we believe any ONE slip, e.g., higher than expected customer acquisition costs, lower than expected margins, missed sales targets, higher equipment costs/subsidies, lower than expected ARPUs due to competition, unexpected costs due to USF, or the negative impact of lawsuits, will push the company off the knife’s edge and into a cash death spiral that will ultimately result in VG equity being worth significantly less than it is today – if its worth anything at all.
VG since IPO:
Full Disclosure: The author is short Vonage at the time of writing.
Update, 9/13/06: Following publication of this article, Seeking Alpha received an email from Gerald Weinberger, President of Rates Technology Inc. ('RTI'), stating that the first patent suit filed against Vonage was that of RTI, asking for $3.3 billion in damages, a permanent injunction against further infringement, and the recall of all examples of the infringing Vonage products and services.