Read on Twitter
Yesterday, I tweeted that fund managers should be not project returns in fund decks (particularly those that claim 5X+ return expectations). https://twitter.com/Samirkaji/status/1290655227461476354
There was great dialogue; Let's revisit the basic math mgrs must go think through. Pls read the below
/start
There was great dialogue; Let's revisit the basic math mgrs must go think through. Pls read the below
/start
1/ Some basics first:
Dilution to exit is generally higher than most people factor (Fred Wilson had a good model on this). Factor ~50% or so.
There is Massive skew in returns (“Babe Ruth effect). https://a16z.com/2015/06/08/performance-data-and-the-babe-ruth-effect-in-venture-capital/
Dilution to exit is generally higher than most people factor (Fred Wilson had a good model on this). Factor ~50% or so.
There is Massive skew in returns (“Babe Ruth effect). https://a16z.com/2015/06/08/performance-data-and-the-babe-ruth-effect-in-venture-capital/
2/ Let’s take a very modest $10MM fund and make the following assumptions (we can debate these assumptions later)
-40 companies
-$250K per (no follow-on reserves)
-Average initial ownership 3.33% ($7.5MM post at inv).
-20% carried interest
-50% dilution to exit
-40 companies
-$250K per (no follow-on reserves)
-Average initial ownership 3.33% ($7.5MM post at inv).
-20% carried interest
-50% dilution to exit
3/ a 5X net return to LPs on a $10MM fund is $60MM in total gross fund liquidity (50MM profit, with $10MM going to GPs for Carry)
4/ Per data from a Correlation Ventures study (and similar findings to a study by Horsley Bridge):
65% of portfolio returns 0-1X
25% of portfolio returns 1-5X
6% return 5-10X
2.5% return 10-20X
Only .4% return 50X.
In seeing 1000’s of funds myself, this rings true.
65% of portfolio returns 0-1X
25% of portfolio returns 1-5X
6% return 5-10X
2.5% return 10-20X
Only .4% return 50X.
In seeing 1000’s of funds myself, this rings true.
5/ Using that math, and being gratuitous on recycling/returns:
26 companies (65% of 40) return 1X = $6.5MM
10 companies (25% of 40) return 5X = 12.5MM
That’s $19MM, leaving $41MM necessary to hit the 5X net (4 remaining companies)
26 companies (65% of 40) return 1X = $6.5MM
10 companies (25% of 40) return 5X = 12.5MM
That’s $19MM, leaving $41MM necessary to hit the 5X net (4 remaining companies)
6/On the $1M invested in remaining 4 companies, that requires an AVERAGE 41X net return per company, or in other terms,with about 1.7% ownership at exit, $2.5B in enterprise value.
Of course, returns are decidedly non-linear and getting a 41X on each investment isn’t realistic
Of course, returns are decidedly non-linear and getting a 41X on each investment isn’t realistic
7/ Using the Correlation historical data:
2 companies return 10X each (6% of 40) = $5MM returned
1 company (2.5% of 40) returns 20X = $5MM returned.
This leaves $31MM to be returned by final investment
2 companies return 10X each (6% of 40) = $5MM returned
1 company (2.5% of 40) returns 20X = $5MM returned.
This leaves $31MM to be returned by final investment
8/ This means the final investment ($250K initial investment) would have to return 124X, at an implied exit value of just under $2B.
9/ Note the model above assumed:
Fund is recycled so that full $10MM would go into companies
The high end of the Correlation return multiple was used for each cohort of portfolio companies
Dilution was limited to only 50% for the enterprise value computed
Fund is recycled so that full $10MM would go into companies
The high end of the Correlation return multiple was used for each cohort of portfolio companies
Dilution was limited to only 50% for the enterprise value computed
10/ I covered the math in blog form a few years ago here as well. https://medium.com/vcdium/micro-vc-smaller-is-better-but-the-math-is-still-really-hard-c0cd4f62ccc2
11/ In summary, my post yesterday was to make sure everyone understands the difficulty of the venture model, and cavalier claims of 5X returns won’t help with the fundraising process with smart LPs (who know this math well).
12/Instead, show that you know the math/difficulty and focus on why your path to outsized returns is clearer than the others in your peer cohort (many inst. will only add 1-3 names each year, so you are competing for similar pools of capital).
13/ It CAN be done and that should be the orientation for investment decisions. One note- in a data set of 985 funds over a 15 year span, we found only 10 that provided a 5X+ DPI.
14/ Thankfully, a 2.5X+ is considered a very successful venture fund (a consistent 2X net will keep you in business a long time, especially if sprinkled with a few funds well above that)
