4 ms·

The problem is this: (1) Money invested provides returns because it's invested in some profit-generating process. Large corporations generate predictable but l

The problem is this: (1) Money invested provides returns because it's invested in some profit-generating process. Large corporations generate predictable but low returns. Small companies, if the right ones are picked, should be able to do better. (2) Putting money in the smartest people (technologists) should be a no-brainer winning proposition. But... (3) Venture capitalists don't just want to give people money. They want to manage, and they want to take what they see as "the fun work". This leads to adverse selection. The best entrepreneurs limit VC involvement (they'll raise money, but on their terms) so VC gets stuck with the middle.

graycat 13y ago
For why the low average returns, here's my guess: (1) The golden rule is, he who has the gold makes the rules. Well, the people with the gold are the limited partners, heavily pension funds and university endowments. At these limited partner organizations, the people making the decisions are in the tradition of banking and finance MBAs. They are not particularly technical or entrepreneurial. So, they like to look at usual accounting data and otherwise think much like bankers. For them, venture capital is a small 'asset class' for which the limited partners do not want to make exceptions. So, the limited partners have the venture partners on short leashes, that is, constrained. For the information technology venture partners (biotech may be quite different), apparently the main constraints are, the venture partners will invest only in the 'themes' they presented to the limited partners and then only based on near surrogates for accounting measures. The main surrogates are the parts of 'traction' -- for a Series A, usage, 'monetizable' in a large 'space', significant and growing rapidly. With this guessing about rules, one issue is 'bubbles': Without some constraints, the venture partners may be able to blow bubbles. Then we could hear again "Never be between a VC and the door when the lockup period is over.". Likely then the limited partners could be losers when a bubble bursts. So the limited partners want to hold down on bubbles. (2) The limited partners like to see venture partners of a certain kind. So, if look at the Web sites of the venture firms, apparently someone up there doesn't like venture partners to hold Ph.D. degrees, not even in STEM fields. Instead, the 'vanilla' background of a venture partner is an Ivy League non-technical Bachelor's, a Harvard or Stanford MBA, a few years in management consulting, and some time in a successful start up in a non-technical slot, say, sales, marketing, or business development. There is an assumption: The 'technology' is just routine practical computing. My view is, with this setup, no wonder on average venture capital is doing poorly on return on investment. I wouldn't know what the heck to do to make money in that situation. It appears that the smarter venture partners want (A) to get a lot of publicity so that they can see the best deals early on, (B) do a lot of pressing of the flesh, especially in Silicon Valley, Boston, and maybe NYC, to hear about the best deals early on, and then (C) get in the best deals. So, e.g., they want to recognize and get in on the next Facebook when it has 1 million users and is growing at 10% a week or some such. To me this approach throws out the baby and drinks the bathwater. Uh, the big advantage is the technology! We should push the technology forward and exploit the technology for more powerful, valuable solutions. But the venture partners do not want to have to try to evaluate technology and, really, want to assume that all the technology is just routine software. Or, usage significant and growing rapidly in a big space is terrific; if there is also some really advanced technology, then maybe that is not too bad! I don't see it; I don't see any hope. To me the key is to concentrate on more and more powerful technology, in a military aviation analogy, to "push the envelope" on altitude, speed, rate of climb, turning rate, etc. So, in that famous one word answer, what we want is "more". My guess is that what we need to do for a 'project' is to pick a 'big' problem, that is, one where the first good or a much better solution will seen as a 'must have' by, say, 1+ billion people (or a few people each willing to spend a lot of money). Then use the most powerful technology we can to get the most valuable solution we can for this big problem. But that's not what venture partners want to do if only because they want nothing to do with trying to evaluate advanced technology. Yes, projects of the kind I believe are needed arrive in the e-mail in boxes of the venture partners far too rarely. But, when they do arrive, the venture partners do not see another Facebook and ignore the contact. The solution is (A) more projects of the kind that are needed and (B) venture partners paying careful attention to such projects. To get more projects of the kind I believe are needed will require work all the way back in the 'high technology supply chain'. How to do that? The US did that in a big hurry once Sputnik went up, on TV, in middle schools, in high schools, in summer programs, in colleges, in the best research universities, with lots of grant money in STEM fields, and, during the Cold War and the Space Race, a product of that supply chain in two weeks going on seven interviews and getting five offers, each for an annual salary six times what a new, high end Camaro cost. It can be done. At the very least, venture capital will have to reach out and encourage much more in advanced technology. Then venture capital will have to learn to evaluate such projects. In the meanwhile, venture capital will look for new ways for teenage girls to gossip and look for boys!
- hga 13y ago
  Your mention of technology and pushing the envelope makes me think of one of my current laments, "Who's going to finance the development of the next FPGA?", where FPGAs are an example of a neat and important general purpose technology. The '80s were very exciting, all sorts of hardware ventures got funding, and we're largely living on the fruits of that period ... including the FPGA (http://en.wikipedia.org/wiki/Field-programmable_gate_array http://en.wikipedia.org/wiki/Field-programmable_gate_array). What besides, say, GPUs can anyone think of that's getting attention right now?
  - oijaf888 13y ago
    SSD technology? Companies like FusionIO and such seem to have been/are VC backed.
    - hga 13y ago
      Interesting thought. SSDs are in part a classic disruptive technology, a repackaging of something new and at least originally not quite ready for prime time that's eating away at various disk drive niches, some lower end. On the other hand they're at least a bit of a sustaining technology, just another mass storage technology (well, at least until people start using them as other than drop in replacements for disk drives and use them with their advantages and limitations). Funding them as a class of technology is not a hard decision for VCs, they're not as much a new thing as FPGAs, and there's e.g. Intel validating the concept.
  - rdl 13y ago
    Lots of awesome radio stuff happened in the 1990s. (UWB, RFID/NFC, etc.), and optics/optoelectronics. Low power high performance seems to be the focus of mainstream chip design right now. Display tech is finally getting awesome again (IZGO, IPS-on-reasonable-devices, etc.) CCD/CMOS, very proper. I mean, cheap 30MP cameras? Lots of great stuff in virtualization (sw and hw features), cloud, and SDN. Arista alone makes up for approximtely 2000-2008. HSMs, smartcards, tamper resistance, etc. have kind of stagnated :(

graycat 13y ago

For why the low average returns, here's my guess: (1) The golden rule is, he who has the gold makes the rules. Well, the people with the gold are the limited partners, heavily pension funds and university endowments. At these limited partner organizations, the people making the decisions are in the tradition of banking and finance MBAs. They are not particularly technical or entrepreneurial. So, they like to look at usual accounting data and otherwise think much like bankers. For them, venture capital is a small 'asset class' for which the limited partners do not want to make exceptions. So, the limited partners have the venture partners on short leashes, that is, constrained. For the information technology venture partners (biotech may be quite different), apparently the main constraints are, the venture partners will invest only in the 'themes' they presented to the limited partners and then only based on near surrogates for accounting measures. The main surrogates are the parts of 'traction' -- for a Series A, usage, 'monetizable' in a large 'space', significant and growing rapidly. With this guessing about rules, one issue is 'bubbles': Without some constraints, the venture partners may be able to blow bubbles. Then we could hear again "Never be between a VC and the door when the lockup period is over.". Likely then the limited partners could be losers when a bubble bursts. So the limited partners want to hold down on bubbles. (2) The limited partners like to see venture partners of a certain kind. So, if look at the Web sites of the venture firms, apparently someone up there doesn't like venture partners to hold Ph.D. degrees, not even in STEM fields. Instead, the 'vanilla' background of a venture partner is an Ivy League non-technical Bachelor's, a Harvard or Stanford MBA, a few years in management consulting, and some time in a successful start up in a non-technical slot, say, sales, marketing, or business development. There is an assumption: The 'technology' is just routine practical computing. My view is, with this setup, no wonder on average venture capital is doing poorly on return on investment. I wouldn't know what the heck to do to make money in that situation. It appears that the smarter venture partners want (A) to get a lot of publicity so that they can see the best deals early on, (B) do a lot of pressing of the flesh, especially in Silicon Valley, Boston, and maybe NYC, to hear about the best deals early on, and then (C) get in the best deals. So, e.g., they want to recognize and get in on the next Facebook when it has 1 million users and is growing at 10% a week or some such. To me this approach throws out the baby and drinks the bathwater. Uh, the big advantage is the technology! We should push the technology forward and exploit the technology for more powerful, valuable solutions. But the venture partners do not want to have to try to evaluate technology and, really, want to assume that all the technology is just routine software. Or, usage significant and growing rapidly in a big space is terrific; if there is also some really advanced technology, then maybe that is not too bad! I don't see it; I don't see any hope. To me the key is to concentrate on more and more powerful technology, in a military aviation analogy, to "push the envelope" on altitude, speed, rate of climb, turning rate, etc. So, in that famous one word answer, what we want is "more". My guess is that what we need to do for a 'project' is to pick a 'big' problem, that is, one where the first good or a much better solution will seen as a 'must have' by, say, 1+ billion people (or a few people each willing to spend a lot of money). Then use the most powerful technology we can to get the most valuable solution we can for this big problem. But that's not what venture partners want to do if only because they want nothing to do with trying to evaluate advanced technology. Yes, projects of the kind I believe are needed arrive in the e-mail in boxes of the venture partners far too rarely. But, when they do arrive, the venture partners do not see another Facebook and ignore the contact. The solution is (A) more projects of the kind that are needed and (B) venture partners paying careful attention to such projects. To get more projects of the kind I believe are needed will require work all the way back in the 'high technology supply chain'. How to do that? The US did that in a big hurry once Sputnik went up, on TV, in middle schools, in high schools, in summer programs, in colleges, in the best research universities, with lots of grant money in STEM fields, and, during the Cold War and the Space Race, a product of that supply chain in two weeks going on seven interviews and getting five offers, each for an annual salary six times what a new, high end Camaro cost. It can be done. At the very least, venture capital will have to reach out and encourage much more in advanced technology. Then venture capital will have to learn to evaluate such projects. In the meanwhile, venture capital will look for new ways for teenage girls to gossip and look for boys!

hga 13y ago
Your mention of technology and pushing the envelope makes me think of one of my current laments, "Who's going to finance the development of the next FPGA?", where FPGAs are an example of a neat and important general purpose technology. The '80s were very exciting, all sorts of hardware ventures got funding, and we're largely living on the fruits of that period ... including the FPGA (http://en.wikipedia.org/wiki/Field-programmable_gate_array http://en.wikipedia.org/wiki/Field-programmable_gate_array). What besides, say, GPUs can anyone think of that's getting attention right now?
- oijaf888 13y ago
  SSD technology? Companies like FusionIO and such seem to have been/are VC backed.
  - hga 13y ago
    Interesting thought. SSDs are in part a classic disruptive technology, a repackaging of something new and at least originally not quite ready for prime time that's eating away at various disk drive niches, some lower end. On the other hand they're at least a bit of a sustaining technology, just another mass storage technology (well, at least until people start using them as other than drop in replacements for disk drives and use them with their advantages and limitations). Funding them as a class of technology is not a hard decision for VCs, they're not as much a new thing as FPGAs, and there's e.g. Intel validating the concept.
- rdl 13y ago
  Lots of awesome radio stuff happened in the 1990s. (UWB, RFID/NFC, etc.), and optics/optoelectronics. Low power high performance seems to be the focus of mainstream chip design right now. Display tech is finally getting awesome again (IZGO, IPS-on-reasonable-devices, etc.) CCD/CMOS, very proper. I mean, cheap 30MP cameras? Lots of great stuff in virtualization (sw and hw features), cloud, and SDN. Arista alone makes up for approximtely 2000-2008. HSMs, smartcards, tamper resistance, etc. have kind of stagnated :(