2011-06-16 06:48:19
The moderator's opening remarks
The bursting of the dotcom bubble in 2000 saw the shares of a wide range of
technology companies plunge, Icarus-like, to earth. Now there is talk of a new
tech bubble inflating. The debate is being stoked by the fact that the market
for initial public offerings (IPOs) by internet companies has suddenly come
back to life after a long period in the doldrums. The recent IPO of LinkedIn,
whose share price more than doubled on the first day of trading, and a host of
listings by Chinese and Russian internet firms on American stock exchanges are
reviving memories of the exuberance that accompanied the web investing frenzy
of the late 1990s.
During that period, there was a headlong rush into internet stocks which ended
in tears for many investors. Now valuations of web firms are inflating rapidly
once again. Groupon, an online coupon business that recently filed for an IPO,
is rumoured to be worth around $15 billion, and trades in Facebook shares on
private markets have put a price tag on that giant social network of some $70
billion.
Sceptics say that, while a number of web companies may justify such heady
valuations, investors seem to be assuming that every big internet business that
goes public will be a winner. And they claim that the prices of relatively
mature internet firms are being driven up to excessive levels in private
markets by hedge funds, private-equity companies and other deep-pocketed
investors who are desperate to get their hands on the hottest stocks around.
But there is another view that is widespread in Silicon Valley and other tech
hubs. This holds that the world is a very different place to the one in which
the shares of such notorious firms as Pets.com and Boo.com boomed in the 1990s
and then went spectacularly bust. For one thing, many more people are now
plugged into the internet and the prospects for e-commerce are brighter than
ever. For another, many internet start-ups this time round have impressive
revenues and robust business models.
So, are the fears of irrational exuberance in technology investing justified?
Or is bubble talk completely overblown? To debate the issue I am delighted to
welcome two experts whose knowledge and experience make them eminently
qualified to tackle the subject before this House. Steve Blank, who will argue
in favour of the motion, is a former serial entrepreneur turned academic, and
Ben Horowitz, who will oppose the motion, is a highly experienced manager and
co-founder of Andreessen Horowitz, a venture-capital company that has invested
in a number of prominent web start-ups.
Both of our debaters have produced impressive opening statements in support of
their positions. I hope these will encourage you, our readers, to engage in the
discussion by contributing your own comments online. And I look forward to a
stimulating discussion bubbling up in our virtual debating chamber.
The proposer's opening remarks
We won't get fooled again
We don't get fooled again
Don't get fooled again
No, no!
The Who 1971
First, let us start with a definition of a tech bubble.
A tech bubble is the rapid inflation in the valuation of public and private
technology companies that exceeds their fundamental value by a large margin. It
is accompanied by the rationalisation of the new pricing, and then followed by
a spectacular crash in value. (It also has the "smart money" investing early
and taking profits before the crash.)
Bubbles are not new; we have had them for hundreds of years (the Tulip Mania,
South Sea Company, Mississippi Company, etc.). And in the last decade, we have
had the dot.com bust and the housing bubble. This tech bubble is unfolding just
like all the other bubbles before it.
Today, the signs of the new bubble are the Linked-In initial public offering
(IPO), Facebook's stratospheric valuation and the rapid rise of early-stage
startup valuation. Hiring technology talent in Silicon Valley is getting
difficult, and the time it takes to drive across Palo Alto has tripled all
signs of the impending apocalypse.
Dr Jean-Paul Rodrigue, in the Department of Global Studies & Geography at
Hofstra University, observed that bubbles have four phases; stealth, awareness,
mania and blow-off. I contend that we are approaching the early part of the
mania phase.
In the stealth phase, prescient angel investors and Venture Capitalists (VCs)
start investing in an industry or market segment that others have not yet
found. In the case of this bubble, it was social networks, consumer and mobile
applications, and the cloud. VCs who understood the ubiquity, pervasiveness and
ultimate profitability of these startups doubled-down on their investments.
Long before others, they saw that these applications could have hundreds of
millions of users with "off the chart" revenue and profits.
The awareness phase is where other later-stage investors start to notice the
momentum, bringing additional money in and pushing prices higher. The Russian
investment group, DST, is an example, with their $200 million investment in
Facebook, at a $10 billion valuation, in 2009. This was followed by another
$500 million investment (along with Goldman Sachs) in 2011, at a $50 billion
valuation. Meanwhile, the bubble for "seed stage" startups began when Ron
Conway's Silicon Valley Angels and DST guaranteed every startup out of a
YCombinator $150,000. And it was hammered home with Color a startup without a
product raising $40 million, at a reputed $100 million valuation, from brand
name VCs who should have known better. When they did launch their product, it
was compared to boo.com, and entered the dot.com bubble hall of infamy.
Meanwhile, smart VCs continue to invest in this segment and increase their
ownership of existing companies. The technology blogs (TechCrunch, et al.)
start cheerleading, and the general business press/blogs start paying
attention. And all of the investors trot out explanations of "why this time
everything is different".
Four stages
We have just entered the mania phase. The Linked-in IPO valued the company at
$8.9 billion at the end of the first day of trading. It sent a signal that
there is an irrational demand for tech IPOs. Silicon Valley startups are
falling over each other to file their S-1 documents to go public.
Some precursors to the bubble happened when Chinese Internet companies listed
on United States stock exchanges. In December 2010, Youku the YouTube of China
went public, with a valuation of $4.4 billion at the end of the first day (on
$58.9 million in 2010 sales). In May 2011, RenRen the Facebook of China had a
first day valuation of $7.4 billion (on $76.5 million in 2010 sales).
Dr Rodrigue's description of what happens next sounds familiar: "the public
jumps in for this 'investment opportunity of a lifetime'. The expectation of
future appreciation becomes a 'no brainer' Floods of money come in creating
even greater expectations and pushing prices to stratospheric levels. The
higher the price, the more investments pour in. Unnoticed from the general
public, the smart money as well as many institutional investors are quietly
pulling out and selling their assets Unbiased opinion about the fundamentals
becomes increasingly difficult to find as many players are heavily invested and
have every interest to keep asset inflation going."
"The market gradually becomes more exuberant as 'paper fortunes' are made and
greed sets in. Everyone tries to jump in and new investors have absolutely no
understanding of the market, its dynamic and fundamentals statements are made
about entirely new fundamentals implying that a 'permanent high plateau' has
been reached to justify future price increases."
We are seeing this bubble unfold by the book.
No one doubts that social networks and web and mobile applications are
reinventing commerce. Obviously, some of these companies will have hundreds of
millions of customers, unprecedented revenue growth and great profits. Yet none
of these companies have earned the valuations that they are receiving.
For all of these reasons, I believe this House should vote in favor of the
motion before it.
The opposition's opening remarks
We are not in a technology bubble. We have not even taken a major step towards
a technology bubble. Predicting such things is a bit like predicting the end of
the world; the prediction will eventually come true, but almost everyone who
listens to you in the meanwhile will regret having done so.
Let us start by understanding the nature of bubbles. Warren Buffet recently
made the following remarks about the housing bubble:
"The only way you get a bubble is when a very high percentage of the population
buys into some originally sound premise that becomes distorted as time passes
and people forget the original sound premise and start focusing solely on the
price action People overwhelmingly came to believe that house prices could not
fall significantly. And since [property] was the biggest asset class in the
country and it was the easiest class to borrow against it created, you know,
probably the biggest bubble in our history."
So let us first ask if "a very high percentage of the population" has bought
into a distorted premise about the future growth prospects for technology. If
they have, then we should be able to see some evidence that the dominant public
technology companies are moving towards bubble valuations. Here is a telling
statement from an analyst on Apple's most recent quarter:
"Apple's stock price to earnings ratio has dropped to 16.72. Ex-cash it's 13.5.
On a forward basis (my estimates) it's 8.3. Apple's valuation is now a case for
business historians to discuss because I don't think there are modern
precedents."
-Horace Dediu, Asymco
If we are in a bubble, that is a bit of an odd commentary for a company that
grew revenues 83% year-over-year and grew earnings 93% year-over-year.
Similarly, Google, well on its way to owning the dominant smart phone operating
system and which maintains a near monopoly position in search, trades at a
price/earnings (P/E) ratio (ex-cash) of around 13.7.
Amazon trades better than both with a price of 24 times last year's cashflow.
But this is still hardly a bubble multiple. For comparison, the average P/E for
the Standard & Poor's (S&P) 500 over the last 100-plus years is approximately
16x. In the last bubble, the S&P hit 44x in January 2000. Currently, the S&P is
trading at 22x.
So it looks like the market leaders are trading closer to recession multiples
than bubble multiples. But what about newer companies like Netflix,
Salesforce.com, and LinkedIn? All of those companies trade at high multiples.
Do they signal that we are moving towards a technology bubble, or are those
multiples high for excellent underlying reasons (Buffet's "sound premise")?
Before we can accurately consider the prices of these firms, we need to note
the four main reasons why technology companies should be trading at higher
multiples than they traded at in the past.
Generational adoption
"Science advances one funeral at a time."
-Max Planck
History shows that major technology cycles tend to be around 25 years long with
the bulk of the purchases occurring in the last five-to-ten years. This has to
do with adoption rates; this period seems about right for the oldest cohorts
(less likely to adopt new technologies) to die off and for younger cohorts
(quickest to use new technologies) to enter the market.
Let us look at examples of the last two major computing cycles (prior to the
Internet).
Now let us look at where we are with respect to the Internet adoption cycle.
As you can see, we are poised to hit the major adoption wave for the Internet
technology platform over the next 8 years.
The internet is working
A lot has changed since the internet bubble eleven years ago. Firstly, the cost
of running an internet application has fallen 100-fold. In 2000, I was CEO of
the first cloud computing company, Loudcloud, where the price for a customer
running a redundant version of a basic internet application was approximately
$150,000 per month. The cost of running that same application today in Amazon's
cloud costs about $1,500 per month.
Secondly, developers are more productive. In 2001, Stewart Butterfield
abandoned plans to build a massively multiplayer online game (MMOG) after costs
became too great; he built photo-sharing service Flickr instead. Now Stewart's
new company, Tiny Speck, is again building that MMOG, but today it is working
brilliantly. Why? Because Stewart's programmers are ten times more productive
than they were in 2001 due to massive advances in programming language
technology.
Thirdly, the market is far bigger. In 1998, I was working at Netscape, which
owned well over half of the browser market. We had about 50 million users, more
than half of them on dialup connections which could not run many interesting
applications. Today, there are over 2.1 billion people on the internet, most of
them using broadband connections. The true market for internet businesses is
about 50 times larger than during the actual technology bubble.
With costs 100 times lower, programmer productivity ten times higher, and the
market 50 times larger, it stands to reason that many more internet businesses
will work today than did the last time around.
The markets for internet businesses will double in size again over the next
five years
International Data Corporation (IDC) estimates that there will be 1 billion
mobile internet users by 2013. That estimate will prove to be low. There are
currently 4.5 billion mobile phones worldwide; within five years almost all of
them will be more fully featured "smart" phones offering better access to the
web.
As smart phones become the volume leaders, the component costs for smart phones
will fall below the corresponding component costs for low spec "feature" phones
a trend that will eventually render feature phones obsolete. As a result of
smart phones replacing feature phones, the internet will double in size over
the next five years.
Software is eating the world
Back in 1994, very few people would have predicted that the largest bookseller
in the world would be a software company. Today, not only is it a software
company, but all of Amazon's most important competitors are also software
companies.
Books were just the first of many industries to be eaten by software. Some
other examples:
company, Apple; its largest potential threat, Spotify, is a software company.
company.
industry, they had to buy Pixar, a software company.
software company, Google.
What is next?
software platform.
new companies like Groupon, Foursquare and Square.
As software eats one industry after another, the market for technology business
expands, rendering previous market size estimates obsolete. That is not to say
that no price is too high for a technology company, but there is a fine case
that the old prices are too low.
Have we taken a big step towards a bubble?
So now let us look at those high multiple companies (Netflix, Salesforce and
LinkedIn) in the context of these underlying trends.
First, Netflix software is eating the world and Netflix is eating the cable
industry. Much like Yahoo once lifted the valuable bits out of America Online
(AOL; the US internet company), made them global, and became far more valuable
than AOL, Netflix is lifting the valuable bits out of the cable business and
making them global. Today, Comcast is worth $66 billion and Netflix is worth
$13 billion.
Second, Salesforce.com the internet is now working and customers like it way
better than the old software model. We have over 60 companies in our portfolio
and none of them use any products from the current software leader, Oracle.
Nearly all of them use Salesforce.com. The company is worth $18 billion and
Oracle is worth $158 billion.
Third, LinkedIn there are now 2.1 billion people on the internet and over 100
million maintain their resumes in LinkedIn. LinkedIn is already quickly
becoming the world's recruiting database, and it seems quite logical that there
will be only one (who wants to maintain their resume in more than one place?).
LinkedIn has already built a series of popular recruiting products that
produced $93m in revenue last quarter. The internet is growing, and LinkedIn
looks like it may become a gigantic company.
I am not arguing that the above companies are not overvalued; I am simply
arguing that their valuations have not become completely divorced from any
rational thought. If they have not, we have not taken a major step towards a
bubble.
The problem with predictions
One may still argue that a bubble is coming. A bubble will almost certainly
come eventually that is the nature of human psychology and of markets.
But what is the value of predicting a bubble with no time frame? What does that
even mean? If we are approaching a boom and huge growth in technology over the
next several years, do you want to miss it due to the eventual bubble? If the
true goal of the bubble promoters is simply to encourage caution in investing,
when does that advice not apply?
For all of the reasons that I have laid out here, I urge this House to reject
the motion before it.