Venture capital (VC) has undergone a seismic shift over the past two decades, evolving from a niche cottage industry into a complex, fragmented asset class with multiple sub-strategies (seed, early-stage, growth, multi-stage, etc.).
A really well-written piece. I don't think most LPs even realize how much the median/top quartile FoF outperform. Those side-by-side charts are definitely worth calling more attention to
The data shared feels a bit unbelievable, I'd fully expect the floor to be higher for FoFs, but to have both the floor and ceiling be higher than direct VC in several years feels unlikely.
Could be driven by sporadic reporting by FoFs (where only the best share their data), would also need to look at the relative sizes of the universes. Perhaps if there are only 50 FoFs, you should actually just sample 50 VCs randomly, not use the huge universe which will have the effect of moderating the quartile bands.
For the heck of it, I went and downloaded the same data from PB, and I got wildly divergent numbers (happy to share my raw data)
- Direct VCs had higher top quartile figures in 10 out of 15 Vintage years (expected result)
- Direct VCs had higher median TVPI figures in 7 out of 15 vintage years (a bit lower than expected but right around 50%)
- Direct VCs had higher bottom quartile TVPI figures in 3 out of 15 vintage years (right in line with expectations since FoFs should have a higher floor)
Also, my top quartile numbers are much lower than yours (typically in the high 1.8x TVPI in most years, breaching 2.5x in exceptional years), it seems you may have chosen "Top Decile" accidentally.
There is no way to diversify away risk and consistently generate higher raw returns, otherwise all assets on earth would flow into FoFs because they have found shang-gri-la. Sharpe ratio remains undefeated at the asset class level.
I agree the data is eye-opening, and it's quite possible you are looking at a different data set, as this was a custom project that I worked with PB on. Of course any single venture fund has the possibility of outperforming a FoF, but the VC asset class as a whole has such high return dispersion with pretty lousy median returns and very bad lower quartile returns. FoFs simply tend to have better access (because it's all they do) and do a better job of finding top quartile managers and avoiding bottom quartile managers.
You're also correct that the universe of VC FoFs is much smaller than the total VC universe. But this only reinforces my point. If the universe of FoFs is much smaller, and, on average, performs better than the significantly larger broader VC market that has more dispersion, and, on average, worse returns, doesn't that make VC FoFs even more attractive? (probability of outperforming is higher)
really great article... only criticism is that i think you buried the lede here.
your title / headline for this post should be something like "VC FoF outperforms traditional VC fund investing, EVEN WITH A DOUBLE LAYER OF FEES" ;)
A really well-written piece. I don't think most LPs even realize how much the median/top quartile FoF outperform. Those side-by-side charts are definitely worth calling more attention to
Great post. From 2010-2015, the *median* FoF outperformed the *top-quartile* Direct 4 outta 5 years.
The data shared feels a bit unbelievable, I'd fully expect the floor to be higher for FoFs, but to have both the floor and ceiling be higher than direct VC in several years feels unlikely.
Could be driven by sporadic reporting by FoFs (where only the best share their data), would also need to look at the relative sizes of the universes. Perhaps if there are only 50 FoFs, you should actually just sample 50 VCs randomly, not use the huge universe which will have the effect of moderating the quartile bands.
For the heck of it, I went and downloaded the same data from PB, and I got wildly divergent numbers (happy to share my raw data)
- Direct VCs had higher top quartile figures in 10 out of 15 Vintage years (expected result)
- Direct VCs had higher median TVPI figures in 7 out of 15 vintage years (a bit lower than expected but right around 50%)
- Direct VCs had higher bottom quartile TVPI figures in 3 out of 15 vintage years (right in line with expectations since FoFs should have a higher floor)
Also, my top quartile numbers are much lower than yours (typically in the high 1.8x TVPI in most years, breaching 2.5x in exceptional years), it seems you may have chosen "Top Decile" accidentally.
There is no way to diversify away risk and consistently generate higher raw returns, otherwise all assets on earth would flow into FoFs because they have found shang-gri-la. Sharpe ratio remains undefeated at the asset class level.
I agree the data is eye-opening, and it's quite possible you are looking at a different data set, as this was a custom project that I worked with PB on. Of course any single venture fund has the possibility of outperforming a FoF, but the VC asset class as a whole has such high return dispersion with pretty lousy median returns and very bad lower quartile returns. FoFs simply tend to have better access (because it's all they do) and do a better job of finding top quartile managers and avoiding bottom quartile managers.
You're also correct that the universe of VC FoFs is much smaller than the total VC universe. But this only reinforces my point. If the universe of FoFs is much smaller, and, on average, performs better than the significantly larger broader VC market that has more dispersion, and, on average, worse returns, doesn't that make VC FoFs even more attractive? (probability of outperforming is higher)