"Far more money has been lost by investors preparing for corrections, or trying to anticipate corrections, than has been lost in corrections themselves." Peter Lynch
"The only value of stock forecasters is to make fortune-tellers look good." Warren Buffet
Lately there have been a lot of articles claiming that we are not in a tech valuation bubble, and that this time "it is different". I will go through the arguments people are making, and build a case that we are potentially in the inflationary phases of an internet private company investment upswing equivalent to roughly ~1997 in timing (i.e. the bubble is in its early inflationary stage, things are still reasonably rational for later stage companies, and the cycle has not peaked). Throughout this article I will use "tech" to refer to "Internet" companies (consumer, ecommerce, ads tech etc), which is the specific subsector where the bubble is happening. "Investing" or "valuations" refers to private equities (angel and VC funded companies, again where the valuation inflation is occurring), not the stock market (which is still fairly valued). Finally, "bubble" means "cyclical upswing in the venture financing market". Just as there is a business cycle, I believe there is a venture cycle that goes through peaks (frothy/crazy valuations) and troughs (ridiculous bargains for investors).
Signs we are in the inflationary period of an early stage, internet company valuation bubble:
- Round size and valuations have increased dramatically in the last 12 months.
- Series A companies are raising $7-15 million instead of $3-5 million, and many of their valuations have doubled or tripled to where they would have been a short time ago. The same holds true for seed companies, that are raising more money on better terms then just 1 year before.
- Similarly, some companies who were unable to raise money a year ago have now gone back and found raising money easy - but without substantial additional traction or progress as a company. This is an especially telling sign.
- If this increase in valuations pushes up into later financing series B, C, D etc. a real bubble could inflate across the broader Internet sector
- This is part of a normal venture cycle. Just as there is a business cycle, there is also a longer venture cycle. There have been prior boom/bust venture cycles that seem to exist on a 10-14 year cycle including semiconductors, microcomputers, the internet, and now today. Arthur Patterson of Accel Partners has a very thoughtful way of phrasing this broader set of venture cycles and the history of the tech boom/bust value acceleration cycles over time for specific sub-industries (semis (67-69), microcomputers(81-83), internet (96-99), etc.) This view suggests that this upswing in valuations is business as usual, which also means we are likely to hit a "free fall" period in which valuations come back down very quickly. (Thanks to Kevin Efrusy at Accel for conversations on Arthur Patterson's venture cycle.)
- More and more people are justifying the fact that there is *not* a bubble (see below for arguments against a bubble and why I think these arguments are wrong). This reminds me of the Economist's "recession index", where increase use of the word "recession" in media correlates with a recession actually occurring, even if it is too early to call with harder economic data (which often takes time to collect and analyze).
- MBAs and BD people are starting lots of companies again.
- The Harvard MBA class has a crapload of people starting companies instead of going into banking. To me, swings in MBA aspirations seem to track to a perception of "where the easy money is" or, in bad times, "where the safe jobs/career paths are".
- Too many hot people at tech events.
- All these really good looking, non-technical people are suddenly interested in tech, or in dating tech founders (anyone attend the Crunchies after party?). Really, you want to work for, or date, nerds like us??? This is, of course, meant tongue-in-cheek. But the mix at tech events has definitely changed. Certain types of people are attracted by money vs impact and the pendulum is unfortunately swinging towards more people who are attracted by money.
3 Arguments Being Made Against The Inflation of an Internet Valuation Bubble
1. "Industry leading companies are not overvalued."
The primary argument against a valuation bubble is to point at the better known companies in tech that have raised at multi billion such as Facebook, LinkedIn, GroupOn etc. and say they are fairly valued. This argument states that since there is no bubble for the late stage, well known companies, there therefore must not be a an unjustified upswing in the valuations of earlier stage companies.
- I 100% agree that these companies may be (more or less) fairly valued These are global franchises that are making real revenues/profits and will continue to thrive over the long term
- Unfortunately, too many of the startups being funded today are getting valuations that suggest all of them will be future Facebooks, LinkedIns etc., which is obviously incorrect. Just because the market leaders are fairly valued, does not mean every other startup is a market leader.
- Similarly, during the last bubble, a number of companies went public (and were considered the category leaders) and ended up being long term sustainable global franchies including eBay, Priceline, Amazon, and Yahoo!. These companies are still worth tens of billions of dollars.
- However, a bunch of less sustainable companies rode their valuation coat tails (CMGi anyone? VerticalNet?). In the last bubble, this happened in the public markets. In 2011, it is a private market phenomenon where early stage investing is being positioned as "it doesn't matter what you pay as long as you get into the next Facebook", even though 99.99% of companies are not going to be the next Facebook.
Every time there is a valuation bubble, people say "this time it is different". And to be honest, each time things really are different (see below), but that does not mean valuations are rational.
- In 1998-2000, people pointed to the cheap distribution of the internet, and the disintermediation of middle men, as the reasons the Internet bubble was not a bubble. "It has never been so cheap to reach so many people with so little friction" was the mantra. And it was true. But the valuations went to irrational levels for crappy companies.
- In 2010/2011, people are saying two things have changed: (a) It has never been less expensive to start a company (true), and (b) It has never been so cheap to reach such a large global audience of users before (also true).
- In 1996, Yahoo! raised $1 million on a $3 million valuation from Sequoia. They had bootstrapped the startup until then, and then only needed to raise an initial $1 million in capital. It definitely cost a lot more to start a company in the 1990s. However, a number of companies got going on little initial capital (see e.g. Amazon and its friends and family round) and than raised money to accelerate what they were doing. The bubble happened when a lot of companies got way over capitalized and burned money needlessly.
- The argument that it is cheaper then ever to start a company = higher valuations would make more sense if people are raising seed rounds later in the game or with more traction. But it feels to me more and more companies are able to raise money on powerpoint decks rather then prototypes (or traction). E.g. the low cost of starting something doesn't matter if you are raising money on nothing anyhow.
- GroupOn proves how quickly a massive global franchise can be built these days. Other companies will undoubtedly follow, and massive businesses will be built faster today than anytime before in history. The question is, what is the ratio of success to number of startups that matter (see below) - not just whether a startup can be successful faster.
- The math is interesting. It is not how cheap it is to start a company that solely matters. It is important to ask if the expected outcomes of all these additional cheap companies now differs. E.g. you could have 2X more companies, but if you have less then 2X more successes, valuations on average for a seed deal should go down rather then up (since the expected value of each company is lower). I wish I had hard data on this to be able to look at success rate versus startup number to see if this is a constant. I think this is the single point of data that would allow us to know if we are on our way up to a bubble.
- Where this argument fails is that it is not the *amount* of capital invested that matters, it is the marginal price that capital is willing to pay that matters. For example, if the aggregate value of the stock market goes up by $1 trillion, it is not like $1 trillion in additional capital has flooded into the market. It means that there are some tens of billions of $ who are willing to pay more for the same asset. Small amounts of capital can cause big price fluctuations when prices are set on the marginal cost someone is willing to pay (argument stolen/paraphrased from Naval). This dynamic is exploited heavily in party rounds, where you bring in a small amount of price-insensitive "dumb money" to set the valuation for the entire round.
- Furthermore, while *overall* seed stage investing is down, anecdotally it feels to me it is way way up for Internet companies. If a bubble exists (like most past tech bubbles prior to 1997-2000 e.g. microcomputers in the early 1980s) is sector specific.
- It is limited to the Internet sector, versus all sectors/industries. I am trying to find out more about industry-specific venture bubbles and will share information on this blog as I dig it up.
- Public markets are not serving as a venture capital vehicle. In 1997-2000, companies went public before they were de-risked. This is not happening and will help damper the heights which this inflationary cycle. It is unclear to me if this private-market-only phenomenon will extend or decrease the time it takes for the valuations to come back down.
3 Scenarios For How This Plays Out:
1. Hit the wall early. The inflation of company values continues to work its way up to in private markets, but never gets reflected in acquisition prices or public companies, and the bubble rapidly deflates. In this scenario, a number of startups/investors will get hit badly, due to a lack of good outcomes and hyperinflated valuation expectations they can not meet. In this both public markets (ie IPOs) and acquirers have a strong filter on quality/valuation and the inflationary cycle hits a valuation wall.
2. Get bought for a ton. Acquirers such as Google and Facebook are willing to buy companies at these higher valuations, helping to keep fueling valuation inflation for some period. This will happen iff the acquirers either: (a) really want to buy assets who have as an alternative to raise a big venture round (e.g. "Google, we can raise $20 million at an $80 million valuation, so pay up at $80 million or you never get us") or if the acquirers valuations go up commiseratively so the acquisitions are ultimately at a fair valuation relative to the acquirers stock price. In this case the public markets still only let strong IPOs out, which constrains the outcomes and keeps things semi-sane.
3. Public market mania. The public markets start to slip through lower quality IPOs (despite Sox), and the public markets start to become more of a financing vehicle for immature companies. This is the least likely scenario and it what happened in the broader financial asset bubble which culminated in 2000. However, public market focused money can still flood into the private Internet markets via hedge funds. Hedge funds have seen great success investing in secondary markets for Facebook et al, and so are starting to get tempted to move down market and invest in series B and C companies. Remember, one hedge fund may have $10 billion under management, which is probably within an order of magnitude of *all* the capital earmarked for Internet companies by early stage VCs. And that is just one hedge fund.
BTW, if 2011=1997, this give entrepreneurs and venture investors about 18-36 months to make a lot of money if they can get liquidity, at which point we will have a tech valuation correction (assuming scenario 2 or 3 above happen). This has significant implications for entrepreneurs and how you should think about timing fundraising and potentially an exit.
- It is possible we are in the inflation phase of an Internet company valuation bubble (~1997 equivalent). There will be ~24 or 36 months (could be off by +/-N months) of continuing valuation craziness before the Internet private funding market crashes.
- As an entrepreneur, you should time your financing events or exit strategy with this in mind
- As an investor, you should think through how to continue to invest in the upswing, but make sure to not loose discipline on company/team quality as bad companies have more great exits
- People who say "this time is different" always say that during bubbles. I really do believe things have fundamentally changed. However, I am guessing if you run through the math you will find that valuations have outpaced expected values. The math that can prove this is to look at the ratio of expected outcome of startups (e.g. (# of new startups) / (average outcome per startup). If this ratio is >1, average valuations should go up by that %age, but not more then that.
- Market timing is hard to predict. Markets are notoriously hard to predict. I may be completely off or wrong in my arguments, or I may see evidence that changes my mind that this inflation is happening at all. But for now, this is my take on what is happening.
Thoughts? Let me know what you think in the comments.
You can follow me on Twitter here.