Article Info
Akamai Alpha No More
Akamai (AKAM) was an exception in my portfolio. I typically avoid high P/E high market cap companies, but Akamai had a unique and dominant position in the marketplace that I felt people overlooked. But I’ve synthetically hedged out Akamai holdings since September 22.
I’m writing this post today because I haven’t written anything substantial for a week and a half and readers deserve better. Brightcove, a company I have written about, released their video content distribution platform today (WSJ Link, free article here). It seemed necessary to footnote what I have written before and summarize why I’ve closed the door on Akamai.
First and foremost: Everyone is talking about Akamai and how great they are (This is a great example, and is what catalyzed me to action). I rarely read articles or speak with anyone who has anything bad to say about Akamai. This indicates the market has built in perfect expectations to the price.
Of greater concern is whether Akamai is a high margin service provider, or a low margin hosting service.
Today, Akamai exists to outsource hosting for companies using their large, distributed datacenter. Whether it is Salesforce.com, NBC, Apple iTunes, etc. – Akamai sells a distributed hosting infrastructure that ultimately is application agnostic.
Building and operating large data centers is a depreciating barrier to entry. Google (GOOG) has them. Yahoo (YHOO) has them. Microsoft (MSFT) is building them. This blog is hosted on one. They are not the unique resource they used to be. Therefore, Akamai’s hardware infrastructure is no longer a unique advantage.
They do have a compelling advantage on the sales side, with a large direct Salesforce and great connections into the big media companies. This advantage is eroding as well. The issue is Akamai offers nothing but hosting. Companies like Microsoft, Google, and Yahoo can add value beyond hosting. These companies all have the infrastructure to not just host content, but also pair it with the appropriate advertising to extract revenue.
From ‘The Inevitable Competition of Akamai and Google‘:
Our assumption appears to be breaking down, as Google deftly repositions itself a more of a mediator of video content rather than an author/owner. The WSJ article captures the leading edge of this trend very well. Google appears to be convincing major networks like CBS to allow them to host and monetize high value content. This is very negative for Akamai.
This was before Google bought YouTube. Google bought YouTube because they wanted to monetize the video content. Akamai provides an agnostic solution that requires the content owners to monetize the content. Extracting value from content is inherently a higher margin business than hosting it on a server.
I believe that Google, Yahoo, Microsoft can better pair content and advertising for each individual viewer. NBC, CBS, ABC, etc. evolved in a world where 50mm people all view the same content and advertising (broadcast TV). It is hard to imagine they will evolve better advertising systems than the ones already evolving within the search behemoths.
The cost of hosting quality content that can generate revenue will be zero. Google, Yahoo, Microsoft, and others will line up for the chance to cache content for free in order to pair advertising with it (extracting a success based fee). Akamai has no means and has shown no interest beyond operating a very sophisticated hosting company (see Brightcove vs. Akamai from Feb ’06). This business model does not deserve a triple digit P/E.
Hosting is a commodity. Akamai’s service is no different. The market will eventually price this in; next week, next month, or next year. I don’t want to be a stockholder when it does.
I can calculate the movement of the stars, but not the madness of men. – Isaac Newton
There’s no doubt that AKAM has been fully discovered by the investment community. One curious thing about AKAM is that there is some special sauce behind their offering which might defend the company from outright commodity competition. The problem of distributing n pieces of content across n servers without knowing demand apriori is not all that easy. Adding servers can actually make the problem harder.
The two founders had some very strong mathematical algorithms behind the service when we looked at it years ago. I’d expect the company has been pretty aggressive on securing patents and all that.
I’m not fresh on the space anymore and while I might not go long AKAM right here right now I’d be too afraid to short it given the knee in curve of video content adoption right now…
Your concerns are unfounded, and your analysis on Akamai’s business model is flawed. This is because you’ve overlooked or got wrong *several* important facts about Akamai, Microsoft, Yahoo, and Google.
And to try and boil this collection of companies down to simply “infrastructure” or “advertising” is a gross oversimplification.
>I rarely read articles or speak with anyone who has anything bad to
>say about Akamai. This indicates the market has built in perfect
>expectations to the price.
This is just not sound logic. A little homework and you would’ve found different, but still interesting, points to discuss.
Please do share.
Hi Andrew,
Great blog. Happy to have found it.
I agree with your top line thesis about AKAM in that it is priced to perfection (which is always a fleeting achievement on the street).
However, just from a technical standpoint I think it’s worth pointing out that a true Content Delivery Network is actually quite different from a commodity hosting provider which is just offering cage space. If provoked (which I don’t suggest), CDN architects can bore anyone for hours with intricate descriptions of their distribution and caching algorithms and while I don’t understand most of them, it does appear to take more than a few PhDs to put a decent one together. (I am actually surprised not see comments from them on this post.) The same cannot be said for a regular hosting operation which can pretty much be put together with a Hosting for Dummies book and a few million.
That doesn’t mean that margins won’t come down in the CDN space over time as you suggest, but it does mean that it is a bit more complicated/value added than you give them credit for and it is possible that they will be able to sustain some premium pricing relative to commodity hosting based on actual or perceived superiority of their CDN design and operation.
My big question is the following:
Does massive, parallelized, centralized host plus unlimited bandwidth equal a distributed CDN? Where exactly is the value add? Can it be broken using brute force?
I’ve been contacted by the Akamai CTO. This is the question I have. I’m looking to be educated.
So then are you going to short it?
No. Given the sentiment surrounding this company, that would be crazy. The reality is that in the near term they are likely to be tremendously successful.
Likewise, I believe Cisco has a gaping hole in it’s fundamental ability to generate returns on R&D, but this is something that will not make itself evident immediately.
With a little bit more intelligence on the client side, the value proposition of reliabile service is eroded. With a little bit more intelligence, the value proposition of high speed can erode as well. Agreed that this will become a commodity market, hopefully sooner than later.
Akamai makes its money, for the most part, on $/bandwidth consumed. If you look at the innovations behind p2p broadcasting for on-demand and live streaming, you will see that CDN is not a long term viable solution for delivery of video. I dont think Akamai will go away and replaced by GYM, but I dont expect it to be a primary beneficiary of the upcoming video trend.
Trackbacks / Pingbacks