1. Venture capital is not good for investors. Historically, the average of VC investments outperforms the NASDAQ, but the median does not. Post-bubble the record is even worse.
2. Venture capital returns depend very heavily on the IPO market. When the IPO market is hot, everybody looks good. When exit conditions are bad, median IRR is 9%. When exit conditions are good, median IRR is 76%.
3. Venture capital returns are highly skewed and correlated. The top 10% does a lot better than the bottom 90%. And, performance of funds from the same VCs are heavily correlated. Unlike fund managers, the top VCs stay in the top year after year.
4. The IPO market is broken. During the boom the average IPO offering went from $23M (1994) to $73M (2000). It hasn't gone back down. I would have like to learn more about this--- is it that the underwriters are living in fantasy-land and can't be bothered for less than $50M now? Or is it really that the market punishes smaller IPOs? Janeway points out that this limit seems to be throttling the IPO market, because an $80M offering means something like $100M in revenue (under a bunch of reasonable assumptions) which is a pretty large bar. The number of IPOs in recent years is just a trickle (even compared to pre-boom.) VCs are going to have to either commit to the long haul, or else aim at private sales as the default exit strategy.
5. VCs have too much money. The industry went from $20B under management to $200B in a decade. That is way too much compared to the needs of "distributed R&D for big companies" (or as he put it, "feeding Cisco") and lightweight web startups. As a result, money is going to be put into dumb things (and lost), or investors will go elsewhere.
On the more speculative side:
6. Venture capital has only been successful in a narrow range of industries, specifically "information and communication technologies" and biotech. It has been notably unsuccessful in materials science or other "high-tech" areas. (He mentioned that of 200-some laser startups, exactly one was a success...) Looking forward, VC investment in energy, nanotechnology, and green-tech promises to eat up a lot of money ($1B per "unit" is how he puts it) but is not a good fit.
7. Venture capital's successes are based on government spending or investor optimism. ICT successes may have their roots in the government's heavy research investment in these areas, with the resulting IP relatively unencumbered. More troubling, biotech startups have not been good investments post-IPO; they've underperformed the market. But the VCs profit by delivering these startups to the IPO stage and then getting out.
Janeway's suggested conclusions for the next "new economy" (green tech or whatever else is next) is that both factors are needed. First, a "base of science and technology, created for non-economic purposes, available for commercial exploitation" and second, "access to speculative, liquid markets--- so that VCs can win even when the venture fails, and to provide the capital need[ed] to fund deployment at scale."