Trends in Initial Public Offerings

One of the sub-themes of this blog is the interaction between innovation and economics.  There are several reasons for this:

• Innovation is the major driver of economic growth and technological change, and as such affects all of our lives.  It also means that the innovation process shares a lot of responsibility for the negative externalities of economic growth, such as environmental and social stresses.
• The organized institutions of innovation that characterize our society are a fundamental difference between us and pre-industrial civilizations, and are one potential cause for optimism that we might avoid the fate of almost all pre-industrial civilizations (collapse) in so far as these institutions allow us to adapt better.
• I make my living as an inventor, and thus have a self-interest in understanding the ups and downs of the innovation process.

This morning, I have been taking a look at some statistics on Initial Public Offerings (IPOs).  Most new technologies in the modern era come to market via venture-capital (VC) funded startup companies.  There are two main ways such companies "exit" - meaning provide a cash return to the investors and employees of the startup.  The first is via being acquired by a larger company.  The second is going directly to the public stock markets to sell shares - the Initial Public Offering of shares.  This places a market value on the shares, and (following a delay) allows employees and early investors in the company to trade their stocks.  Within the technology industry, IPO's have always been the more prestigious kind of exit - building an actual standalone company that goes on to be a famous brand is the ultimate vindication for entrepreneurs and VCs alike (though both, especially the latter, are always going to be willing to look at a too-good-to-refuse offer from a larger company keen to acquire the technology).

In any case, I found some statistics at IPO Vital Signs on the number of IPOs per year since 1970, along with the total value of the share offerings in the new companies.  (My efforts to get the Internet to give me earlier statistics than 1970 have so far failed - if anyone has a source, let me know).

Here is the first cut at the data - the blue bars (left scale) are the number of companies making an IPO by year, and the red line (right scale) is the average size of an offering in millions of $2009 (the original data seem to be nominal, and I adjusted with the CPI-U)

Several things stand out - IPOs are much more sensitive to economic ups and downs than other measures  of innovation.  In bad times, the stock market goes down and it's very difficult for new companies to raise funds there.  Eg look at the almost complete collapse of this measure in 1974-1975, following the Arab Embargo oil shock, and the very poor conditions through the rest of the 1970s.  Similarly, following the 2000 tech crunch, there were several years of IPO drought, as again in 2008 and 2009 (the last year of the data).

Also, note how the average size of the IPOs goes up in bad times - I interpret this as meaning that only the strongest startups can IPO in a difficult market (think Google in 2004), and that such companies on average need to get to a larger size before they can successfully exit.  In better times, companies race out of the gate at smaller sizes, and in much larger numbers.

Also interesting is the overall upward trend in the red line - the IPO process, taken as a whole, seems to be becoming more valuable over time.  One possible influence here is just that the economy is getting bigger in real terms.  To factor that out, I constructed a ratio of the fraction of annual GDP that is associated with IPO offers.  That ratio looks as follows:

This measure peaked in 2000 at just over 1%.  The general impression of a rising trend continues here: even in the worst moments of the aughties, the ratio never fell to anything like the lows of the late 1970s, and the 1980s and 1990s were all way above the early 1970s peak.

There are several possible explanations - one is that society really is expending more effort on the innovation process, or that innovators (and their financiers, especially the latter) are on average capturing more of the value they create.  But another possibility is the decline in the large corporate research division.  The likes of AT&T, GE, Xerox, and IBM ran huge and highly productive research laboratories which probably had their heyday in the 1950s and 1960s.  While those labs still exist, they don't have anything like the prominence they used to, and a lot more innovation happens in start-ups.  It's also possible the apparent trend in the above data is really a cycle, and that would be clear if we had earlier data.  I'm just guessing really.

At any rate, with a major deleveraging cycle upon us, I imagine IPOs will continue to be a bit tricky in the coming decade, and only the best and strongest startups will achieve it.