Most venture capital funds lose money(finance.fortune.cnn.com)
finance.fortune.cnn.com
Most venture capital funds lose money
http://finance.fortune.cnn.com/2011/02/16/venture-capital-returns-more-in-short-term-less-in-long-term/
4 comments
In general I am shocked whenever a single anecdote like this is reported as the truth. They try to add some cachet by saying they looked at 1,300 firms, but picking only one start date is awful.
I'd be really interested in average 10 year returns on VC from the past 10 years. I.e funds raised from 1990 to 2000. Obviously even more interesting will be how the current funds do.
I'd be really interested in average 10 year returns on VC from the past 10 years. I.e funds raised from 1990 to 2000. Obviously even more interesting will be how the current funds do.
6th paragraph
"The median net return to VC fund investors has not been positive for any vintage year since 1998."
I'm curious, did you read the article and miss all the talk about vintage years? Or did you only skim it? Or did you only read the comment? Or did I misinterpret your comment?
"The median net return to VC fund investors has not been positive for any vintage year since 1998."
I'm curious, did you read the article and miss all the talk about vintage years? Or did you only skim it? Or did you only read the comment? Or did I misinterpret your comment?
I may not have known what the terms meant, but I took that to mean 10 year returns on funds. So, I thought it was only referring to funds started in 1998, 1999, and 2000. Still definitely not a overwhelming amount of data. I forgot to mention the other years in my comment. However, the general idea of the comment still holds- they took an exact bit of data, like 10 year returns for the worst years of funds to determine that VCs aren't valuable. As has already been commented, if you looked at the 15 or 20 year returns you will be amazed in the other direction.
The 15 and 20 year returns are stunning.
The 15 year ago vintage class does have amazing returns.
However, assume for a second we can take different vintage years and say that return(t-20 to t-0) = return(t-20 to t-15)*return(t-15 to t-0).
Then all the money was made from t-15 to t-10 since the t-10 vintage class has basically zero returns and the t-20 returns are approximately the t-15 returns annualized over 20 years instead of 15
However, assume for a second we can take different vintage years and say that return(t-20 to t-0) = return(t-20 to t-15)*return(t-15 to t-0).
Then all the money was made from t-15 to t-10 since the t-10 vintage class has basically zero returns and the t-20 returns are approximately the t-15 returns annualized over 20 years instead of 15
there are 15 and 20 year columns in the table under that post
Yet another example of getting some things wrong in the article, and then some things wrong in the title, to end up with a basically dishonest piece of journalism.
The article is based off of the "Cambridge Associates LLC U.S. Venture Capital Index." I see no reason to believe that measures the median venture fund--that would be a dumb way to construct an index--the responsible way to create an index is to weight by the size of the fund.
Second, as noted by dgabriel, the story draws its conclusion from the 10-year return--if you look at the quarterly returns on page 6 of the report (http://www.cambridgeassociates.com/pdf/Venture%20Capital%20I...) you'll see the real story is "Venture Firms Lose Money from 2000-2002." Which everyone already knows.
Really, really terrible work.
The article is based off of the "Cambridge Associates LLC U.S. Venture Capital Index." I see no reason to believe that measures the median venture fund--that would be a dumb way to construct an index--the responsible way to create an index is to weight by the size of the fund.
Second, as noted by dgabriel, the story draws its conclusion from the 10-year return--if you look at the quarterly returns on page 6 of the report (http://www.cambridgeassociates.com/pdf/Venture%20Capital%20I...) you'll see the real story is "Venture Firms Lose Money from 2000-2002." Which everyone already knows.
Really, really terrible work.
A) The quarterly pooled MEAN metric does not necessarily tell you how the majority of venture firms perform over any sustained duration. Looking at the MEDIAN vintage year performance on page 7 does (at least a better job of it). From that, it appears most vintage years haven't made money since 1998.
B) We have data for vintage years, but not necessarily firms. I don't think its a terrible generalization to say that the last 10 or 15 vintage years represents VC firms now in business, else there are funds that have been open for 15 years and I believe they usually have a fixed lifetime of less than that.
C) Given points A and B, do you still believe they are drawing their conclusion (and poorly so) only from 10 year returns?
B) We have data for vintage years, but not necessarily firms. I don't think its a terrible generalization to say that the last 10 or 15 vintage years represents VC firms now in business, else there are funds that have been open for 15 years and I believe they usually have a fixed lifetime of less than that.
C) Given points A and B, do you still believe they are drawing their conclusion (and poorly so) only from 10 year returns?
Thanks for pointing out that interesting, and relevant, data (upvoted).
The problem with looking at negative median returns since 1999 is that from 1999-2002 this is just picking up the bubble, and as you point out many funds from vintage years 2003-present are very much still in business, so a current negative return for these vintages is not particularly meaningful. Unless the authors have some methodology I'm not aware of to deal with the fact that venture funds often don't carry companies at anywhere close to their current value.
Finally, even if (B) were true, there's a huge difference between "Most venture funds lose money" and "Most venture funds currently open have lost money." While the first would make me worry about the business model, the second wouldn't.
The problem with looking at negative median returns since 1999 is that from 1999-2002 this is just picking up the bubble, and as you point out many funds from vintage years 2003-present are very much still in business, so a current negative return for these vintages is not particularly meaningful. Unless the authors have some methodology I'm not aware of to deal with the fact that venture funds often don't carry companies at anywhere close to their current value.
Finally, even if (B) were true, there's a huge difference between "Most venture funds lose money" and "Most venture funds currently open have lost money." While the first would make me worry about the business model, the second wouldn't.
Its not clear to me whether those returns are total or annualized. If the latter, the 15 and 20 year returns are eye-popping.
Annualized. If they were total, you'd be way better off with practically anything else.
10 years ago the 5 year return for VC funds would have been absurd. Would it be close to 500%?
There are good and bad VC's, and with funds lasting 10 years it takes a long time to sort them out. It'll be interesting to see if the institutional money really pulls out of the industry or if they start driving up the price of the good VC's. Folks with solid track records might use this data to justify a better deal than 2 and 20 with their LPs.
edit: pets.com evaporated 300 mil in less than a year in 2000, and that's just one example.