Let me preface my takedown of NFLX “the stock” by first stating that I am a big fan of NFLX “the company”. It has all the qualities an investor should desire: a product with solid long-term demand prospects, high ROE and profit margin, a leg-up on its competitors, and impressive opportunities for growth. As for management, it's hard not to fall for CEO Reed Hastings, a charismatic leader if there ever was one. You don’t here much stale, obfuscated MBA jargon from him. Just reality.
NFLX “the stock” is a different story altogether. Current investors in the stock have driven its price to a trailing P/E of 80x. Homo economicus they are not.
Stockholders often defend the valuation with reasoning similar to the following: “80x seems reasonable for a company whose earnings will grow by 45% this year.” This argument is about as poorly thought out as the Iraq war, and, unfortunately, is more common than caffeine addiction. Let me show you why.
This year’s earnings likely contribute somewhere between 1% to 5% of NFLX’s total value and, yet, most market participants have concentrated 100% of their analysis on it. If we were to consider the presumed infinite number of years to follow, what might we make of NFLX’s stock price then?
A fun game to play when analyzing an investment is to jump straight to the Absolute Best Case scenario and see how that valuation compares to the current market cap. If the stock is fully priced versus that scenario, then you’ve probably got a losing investment.
I put together a fairly complex Discounted Cash Flow model to determine NFLX’s value, based on a number of factors, NFLX’s subscriber growth being paramount. At present, it appears the world is NFLX’s oyster. It ‘only’ has some 25 million American subscribers and 1 million international subscribers … and there are billions of people in the world. What is the possible upper limit for NFLX’s growth? How many worldwide subscribers could NFLX possibly have?
First, subscribers are not individuals. Movie watching, more often than not, is a social experience and it makes sense for people to share profiles whenever possible. So the potential subscriber base is a factor of the number of households in the world.
Now consider this. Facebook has become a common tool for basically anyone with an internet connection and a democratic government, but an interesting effect has started to occur. Once Facebook membership reaches ~50% of a country’s population, membership appears to flatline. At this point, the remaining population is either too old, too young, or too busy doing more important things. Facebook may be one of the most ubiquitous experiences in human history, right up there with death and taxes, so I think this 50% figure seems to be a good, if aggressive, estimate to use with respect to NFLX for the Absolute Best Case scenario.
And since this is the Absolute Best Case scenario, I am going to assume Netflix captures 100% of the potential market, which is 50% of all households (with a downward adjustment for international markets as per below).
Beginning with the Domestic Segment, the US Census Bureau counted ~309 million citizens in 2010 and project ~346 million by 2025. The Census Bureau also tells me there are 2.6 Americans per every separate household in the United States. My model assumes that this number declines continuously over the next 15 years. By my math, Netflix currently counts 19% of all US households as clients. Assuming it takes Netflix just 5 years to capture the remainder, Netflix will have 64 mil US subscribers by 2016. From there, subscriber growth matches population growth.
As for the International Segment, the UN’s base case projection for total world population in 2025 is 8.01 billion, plus or minus a billion. So the potential international market for Netflix in 2026 is 8.01 billion – 0.34 million ~ 7.67 billion.
I struggled to find a figure for people per household for the entire world. This statistic is readily available for individual countries but none of these powerhouse public research shoppes has taken the time to do a weighted average, and since I’m not being paid for this article, I’m not taking the time either. In much of Europe it is close to 2, in China it’s ~3.5, in India it’s ~5, and in Iran they cram 7 people into a home. So I think a good ballpark estimate for the world average is 4 people per household, declining over time as well.
What percentage of those international households can NFLX possibly capture? Honestly, it’s anyone’s guess. As of 2005, the World Bank estimates a quarter of the world’s population lived on an income of less than $1.25 per day. Even if all 4 residents of a household had this income, I doubt they’d spend $7 of their $150/month on a movie subscription service, given that it would cost monthly, say, $25 for the internet connection, $50 for the computer, and $500 for a home containing a living room. Another quarter of the population lives the high life on $2 per day. I don’t see many movies in their future either. By comparison, the US department of Health and Human Services set the official poverty line for 2011 at $22,350 annual income for a family of four and ‘only’ ~14% of Americans fell below that threshold.
So I somewhat arbitrarily chose 25%, though frankly I think this is absurdly high. 50% of the world’s population can’t afford NFLX’s service, a good chunk of the others won’t have access to it for regulatory, cultural, and political reasons, and still more just plain won’t want it. But an aggressive 25% will only serve to further my argument that NFLX is too expensive.
SO if Netflix captures 25% of all households in the world by 2025, it will have 7.7 billion x 3 x 25% ~ 639 million international subscribers - growing from just short of 1 million today. That’s good for an annual CAGR of ~50%.
My model makes the following assumptions:
- Average monthly revenues per subscriber grow at 5% per year, reaching $24/month/subscriber by 2026.
- Net income is assumed to be 14% of revenues. This is Netflix’s stated target.
- Acquisition costs are assumed to increase 1:1 with revenues. I believe content providers will always negotiate to maintain a consistent piece of the NFLX pie.
- Amortization of content costs is assumed to be 2x net income, NFLX’s historical norm.
- Current acquisition costs are assumed to always be greater than the amortization costs. As it expands into international markets, NFLX will always be required to shell out large dough to fill its back catalogue. Also, contract renewals will likely always come at a higher cost than existing contracts.
FCF grows at 3% per year ad infinitum after 2026.
Here are some interesting details from the output of the model:
- Netflix must grow its international subscriber base ~50% per year for 15 years to reach the target of 25% of all international households.
- Netflix would add more than 200 million subscriber in years 2025 and 2026.
- Netflix would add almost 400 million subscribers during years 12 to 15.
- In 2026, Netflix would have income of ~$15.5 billion, more than AAPL has today.
- In 2026, Netflix would amortize over $31 billion in content costs.
What intrinsic value results from the model? About $12 billion versus NFLX’s current market cap of ~$14 billion. So even in the most favorable scenario imaginable, Netflix is overvalued.
Disclosure: I have no positions in any stocks mentioned, but may initiate a short position in NFLX over the next 72 hours.