What are the factors driving performance of venture capital firms? Does their size, or the size of their network, matter?
These are questions people have been asking for a long time. If there were simple answers, institutional investors — the universities, pension funds, banks and others who are pumping billions of dollars into this risky industry — would have a much easier job.
Two recent articles try once again to suggest there’s correlation between good performance and (1) size of a VC and (2) the size of a VC’s network. While they’re worthy of discussion, they both have limitations.
In a first article, Techcrunch refers to a study done several years ago that suggested a venture firm’s network size is correlated to its success, and then tries to build on that with analysis of the present-day VC industry. The second, at peHub, claims that firms which moved to reduce their fund size during the last bubble bursting in 2001 actually did no better than the industry average, and in fact may have done worse.
First, the TechCrunch piece. It reasserts that a firm’s network size is related to performance, pointing to a study done years ago by Yael Hochberg, Alexander Ljungqvist, and Yang Lu. The TC article then points to the top 100 most networked venture firms in its database, and suggests this is a list of firms most likely to succeed (or at least it asserts that there’s a correlation with success for these firms). There are several comments on that post which point out the limitations of this, though the study may not be as flawed as you think. At its base, the methodology is actually quite sound: It controls for size, for example. A firm’s network tends to get larger the more active a firm is, and a firm is more active the more money it has raised. A prolific angel like Ron Conway, even though he is incredibly networked, would be overshadowed by a much larger firm with more partners, who each do less networking individually but do more overall as a firm. So the study does well on that front.
However, here’s where the study hits a problem: success breeds followers, and it’s not clear how this is factored out from influencing the rankings. The leading firms, such as Sequoia or Kleiner Perkins, naturally draw lots of co-investors eager to partake of investments alongside them, in order to share in success. And so their networks get more dense as a result. Question is, what came first, the success or the network? For example, at the very top of TC’s list of top 100-networked firms is DAG Ventures, which has an explicit policy of partnering in deals solely with top-tier venture firms such as Sequoia, Kleiner and Benchmark. But is DAG itself successful? Well, we actually don’t know how it has performed (TC provides no numbers), so it’s possible it represents the exception. It may reflect one of those cases that does not correlate — because network size by itself may not be the cause of success. We just don’t know.
Taking this point further, we also see a firm like Crescendo on the top-100 list. But we know that firm is on its last legs (it’s been struggling to raise funds for some time and hasn’t done very well at all in its recent funds; it had one big recent success in Pure Digital, but even there it did so by teaming up with Sequoia on the deal, and it wasn’t enough to make its fund profitable). Crescendo raised a lot of money, so it’s done a lot of deals and generated quite a network. But how exactly has it used its network? The Pure Digital case suggests that at least it did well by networking by Sequoia. In fact, this may show just why the networking is so powerful. If a firm does everything else wrong, at least if it networks, it may be able to pull something off despite itself (by drafting on someone else’s success!). It’s network may help it, on the margins, do better than it would have with no network, but you could argue there are other more significant factors that predict success than network size.
Other firms on the list look shaky. I haven’t done a definitive study, but WaldenVC is another one on the top 100, and it’s also struggled for some time to raise a fund (indeed, one of its partners, Alex Gove, recently left the firm). Final example: Ron Conway, an individual investor high up on the list, sprayed investments everywhere during the last boom, and has a huge network. However, he admits he didn’t do that great on those funds. He did better than some others, but when the whole industry is in the dumps, it’s kind of pointless to point to anyone beyond the top 10 or 20 performing funds — because everyone else basically loses money. Bottom line: I’d caution against seeing a firm’s network size by itself as a strong reflection — or predictor — of success.
The second article, at peHUB, shows that even as large venture firms moved to reduce their fund sizes during the last crash — in 2001 — their actions did little to boost their performance compared to the industry average. Does that mean the move to reduce fund size was misguided? No. Had they not reduced their size, they may have performed even worse.
The funds that reduced their size were clearly the ones who believed they were the most vulnerable to bad performance (why else would they reduce funds?). First, they were the most bloated: Most of these firms had grown to a fund size of $1 billion-plus, from a fund size of only around $200 million a few years before. They’d hired too many young, inexperienced partners (fresh-from-Harvard MBAs and wide-eyed investment bankers), and were clearly too big to invest properly at a time when the economy had shrunk so rapidly. Second, they were typically the most hasty firms — those who had been caught raising a new fund quickly (one or two years, tops) after they’d raised their previous fund, and so they were caught with more money to dispense in less time than other firms. Again, had they not cut back, their performance may have been much worse. But they did cut back, in part also because their limited partners put pressure on them to do so. It was the most prudent decision to make at the time. Bottom line: These firms had been caught with their pants down, and they were going to do horribly regardless. Cutting back did not lead to the bad performance. It was the other way around.
In conclusion, while it’s true that a venture capital firm’s size is correlated with success, and that a large network may be behind that success, there are many factors that suggest there’s not a direct causal link in either case.
VentureBeat is studying mobile marketing automation
, and we’ll share the data.