Stian Westlake dispels the myth that the technology industry is in meltdown.
It seems that economists have it in for venture capital. Last week, economics professor and blogging grandee Tyler Cowen posted the chart below, which seems to show that the returns to VC are collapsing.
Cowen took a gloomy view of the data; he suggests that the chart is an omen of a new age of technological stagnation, an idea he’s been promoting for some time. He was backed up by other economics bloggers, including Tim Worstall and Noah Smith.
The chart shows the average return of all VC funds raised, year by year. Each line represents a different source of data on VC returns. Returns are scaled to the stock-market: “1.0” means that VC funds launched in that year performed exactly as well as the market, “4.0” means they did four times better.
Tyler is an provocative and insightful economist. But when it comes to VC, he’s got it wrong. This is a shame, because fretting about a technological Gotterdämmerung is a distraction from the more important issues the VC industry faces.
Is tech slowing down?
Let’s look at the idea that Cowen and Smith are advancing in a little more detail. Smith wonders whether “something broke the VC business model after the dot-com crash”, and both of them wonder whether the problem is deeper: the deals aren’t there to do because the Great Technology God has stopped smiling down on us.
The idea of a technological slowdown is a fascinating one; it’s been argued in recent months by sci-fi author Neal Stephenson, and by as savvy an investor as Peter Thiel. It’s very hard to prove or disprove, and economic historians will no doubt argue the case for years to come.
People like Bryan Arthur of the Santa Fe Institute or MIT’s Erik Brynjolfsson argue exactly the opposite. If Cowen were right, it would have implications stretching far beyond tech investing: he has speculated, for example, that the tech slow-down caused the global financial crisis, by driving investors into ever more reckless speculation to make up for slowing performance in the “real” economy.
Interesting though the idea is, the chart of VC returns data doesn’t give any real support for the great stagnation idea.
The first problem is where the chart begins. By picking the 1993 vintage as its first year, the chart starts us right in the middle of the dot-com bubble (funds raised in 1993 were typically making investments in the mid-1990s, and exits in the mid-to-late 1990s).
If we take a longer run of data (below), the last few years of performance look less like a new dark age and more like a reversion to the norm.
If we’re worried about a long-term tech slowdown, the numbers to compare are not the early 2000s with the mid-1990s, but the 2000s with the 1980s and early 1990s. If recent VC performance compares well with the age of Cisco, Apple, Genentech and Geron, things could be much worse.
The second problem is that the chart ignores the volume of capital being invested. If anything, today’s figures look better than those of the pre-bubble period. Remember that the industry today is investing much more than it was before the dot-com bubble.
Before 1995, US VC funds never raised more than $10 billion in a year (typically they raised in the low single figures of billions); since 1995, they’ve raised between $10 billion and $30 billion every year, even in the aftermath of the dot-com bubble. So today’s average returns are being realised over a much larger volume of investment than in the 1980s or early 1990s. Again, this isn’t what we’d expect to see if innovation was somehow slowing down.
Finally, VC funds’ performance is likely to improve for the next few years of data that become available, as the benefits of social media investments come out in the data, even as the performance of equity markets has tanked, meaning that the excess returns of VC will be even higher for the late 2000s. And the data will never include non-VC tech investors either: so the returns of angels, which are likely to be more important in the age of super-angels and AngelList than before, won’t be reflected.
Some bigger challenges
All this shows that, even though VC is not doing as well as it was in 1999, there’s little evidence of a terminal tech meltdown. It’s less a case of unprecedented decline, more a return to business as usual. Talk of a great stagnation is a distraction from the more interesting issues the industry faces.
The VC industry has been pretty staid in its operating model for most of the last three decades. The deluge of money that flooded into the sector since 1995 has mostly been deployed in funds with standard structure. In some cases, it’s worse than this. When public money is involved, it’s often invested through funds that are subscale, poorly managed or otherwise unable to achieve high performance, as Nesta’s research on the UK market showed.
The last few years have seen an increase in variety, from new funds like Andreessen Horowitz and DST to the growing role of angels and crowdfunding. It’s too early to say whether any of these individual developments will disrupt the market, but there is plenty of room for fresh thinking.
This is particularly important if we care about the prospects of the wider tech sector. The standard VC fund is adapted to a particular technological landscape: the world of IT, and in particular internet businesses. The underlying technologies drive the economics of businesses and of deals – how much capital is required, how long businesses take to develop, what multiples can be achieved. When it comes to internet businesses, this has been changing rapidly in the last five years: cloud computing and the lean start-up movement have drastically changed the economics of many start-ups.
When it comes to non-internet tech businesses, the economics vary even more. If we think it’s a good thing for finance to available for high-growth businesses in semiconductors, life sciences, clean tech, space, or wherever else the next big breakthrough will come, we have to ask whether the economics of today’s venture funds will deliver it.
Rather than worrying about the decline of innovation in the world at large, maybe we should be more worried about the state of innovation in the VC industry itself.