Data-Driven Insights on Venture Capital with Steve Kim

49:06
 
اشتراک گذاری
 

Manage episode 342353406 series 2832826
توسط Village Global توسط Player FM و جامعه ما پیدا شده است - کپی رایت توسط ناشر، و نه متعلق به Player FM، و صدا به طور مستقیم از سرور های آنها پخش می شود.برای پیگیری به روز رسانی در Player FM دکمه اشتراک را بزنید، و یا فید URL را به دیگر برنامه های پادکست بچسبانید.
Steve Kim (https://www.linkedin.com/in/stevenrkim/), Partner and head of Investment Strategy at Verdis, a 9-generation single family office, joins Olga Serhiyevich (@olgaserhi), Head of Investor Relations at Village Global, on this episode.
Takeaways:
Early stage venture is a power law asset class where the returns of the asset class are driven by outliers. The best way to increase probability of getting asset class rate of return is by increasing variance in the portfolio through diversification. Pattern matching tends to reduce variance and contrary to industry’s beliefs, is undesirable from the systematic approach perspective.
There is no limit to diversification beyond practical limitations of being able to see and invest in all the relevant deals for GPs. 20-30 portfolio companies is a typical level of diversification in other asset classes including growth equity and buyout where returns are normally distributed. In early stage venture (pre-seed to Series A) this level of diversification is less likely to produce industry average returns on a consistent basis.
The average rate of unicorn production is 1-2% in the industry but it varies across sectors, vintage years and geographies. So, Verdis chooses to maximize diversity across the number of companies, sectors and vintages because there is no clear indication in data that subsets of those are more likely to produce outliers but invest with a bias towards key geographies due to higher concentration of unicorns there.
Most of the outliers in the US of the last decade came from two geographies - California and New York. The magnitude of these outliers was also significantly greater than unicorn companies built elsewhere. For example, on average it takes 4 non-California outliers to equal the magnitude of outcome of a California unicorn.
Startup exit data from other geographies looks a lot more normally distributed which calls for a different approach.
Data-driven investment strategy’s main drawback is the backward-looking nature of the approach. But it’s useful in that it provides a systematic approach to guide portfolio construction.
If managers believe that the part of the VC asset class they focus on follows power law distribution, then they would want to have the most diversified portfolio as possible with a lot more than traditional 10-20 companies. In the power law world, losses don’t matter.
One of the key insights from investing in venture for almost two decades is that most managers are going to stage-drift. Allocating to emerging managers who often focus on early stage due to smaller fund sizes and comfort with first check investing is one way for LPs to hedge against stage-drift.
In Verdis’s view, low reserves and quick capital deployment cycle is advantageous to LPs focused on multiples not IRRs.
Thanks for listening — if you like what you hear, please review us on your favorite podcast platform.
Check us out on the web at www.villageglobal.vc or get in touch with us on Twitter @villageglobal.
Want to get updates from us? Subscribe to get a peek inside the Village. We’ll send you reading recommendations, exclusive event invites, and commentary on the latest happenings in Silicon Valley. www.villageglobal.vc/signup

641 قسمت