Why Quant Funds Are The Future of VC
3 Reasons Why Quant Funds Will Replace Gut Investing in Venture Capital
3 years ago we realized we had a problem. We (Connetic Ventures) were a small, underfunded VC with little brand awareness and an office in Covington, Kentucky. Greater Cincinnati is a great place to live but startups aren’t launching here daily, so our opportunity to find a sufficient number of investments was not ideal.
Here’s what we didn’t have: travel budget, large check size to attract companies, expansive VC network
Here’s what we did have: a team with a background in data and machine learning and confidence that data-driven investing was the future
So…. in 2017 we began our journey in trying to create one of Venture Capital’s first true quant funds. After nearly 3 years in, we are just now starting to reap some benefits. We have had nearly 2,000 startups apply for funding and have analyzed more than 2M data points ranging from founder personalities to startup cap tables.
Through this process, we’ve learned a lot. We know most of the information we have is directional at this time, but it will continue to get better over time.
So, I wanted to share 3 interesting things we have learned that make us believe quant funds are the future of VC
#1: Quant investing can fix the glaring diversity and gender gap in VC
There have been a series of events in 2020 that have amplified the constant discrimination in venture capital funding.
One of the most encouraging and positive benefits that we have seen from automating our investment process is that it appears to have removed race, gender, and age from the decision-making process.
Since the launch of Wendal (our automated due diligence platform) 1.5 years ago, here is the data (sample size is 1,954 companies)
- 48% of applicants have a female or minority CEO
- 49% of Wendal’s investment recommendation had a female or minority CEO
- 51% of investments Connetic Ventures has made had a female or minority CEO (8x higher than VC average)
We couldn’t believe the numbers at first, they seemed to be too perfect. However, we have checked and double-check, it appears that removing the traditional “PITCH” from the process has removed many of the inherent biases that creep up.
Over time, we could see some potential biases creep back into the model but this is an incredibly encouraging start. We hope other funds start to deploy more data-focused screening parameters.
#2: Financials / Key Metrics are key indicators of future success
We have broken down cap tables, valuations, option pools, revenue growth, CAC, LTC, and a number of other quantitative variables.
A number of these are strongly correlated to future company success, while others aren’t. We weren’t entirely sure how this was going to net-out, especially at the pre-seed and seed stages that we invest in.
This was very good to validate though so we can continue to refine and pull additional data into our model. Nearly every day it seems that new data providers are popping up in the private markets so as the actual data continues to evolve and get better, these will be a very significant component to a large number of fund’s evaluation models.
#3: Data-driven investing forces continuous innovation
Finally, when you have a scalable and repeatable process, you start to identify areas of your model or process that need improving. Something we realized early on was that we were missing an incredibly important part of evaluating early-stage companies: THE TEAM
We had a sophisticated machine learning platform and process that was working pretty well but we had no way to automatically screen and evaluate the team.
To address this we spent nearly a year researching behavioral assessments and ending up building our own. We created a behavioral assessment called startupDNA to analyze role fit and overall team composition.
Here is a snapshot of the various profiles and key axis that drive these profiles:
Through extensive backtesting of more than 600 outcomes, we have found that teams do matter and people are hardwired to excel in certain roles.
Out of 22 potential startupDNA profiles, there are a handful of CEO profiles that have >70% chance of returning capital. The VC average is 35–40%
StartupDNA is now a large portion of our algorithm and is something we would have never created if we didn’t pursue a data-driven investment process.
If any of this sounds interesting or you are a startup looking for funding, please drop me a line at hjelm@connetic.ventures
ALSO…. Thank you to the many people that have supported us while everyone else said no. And a big thank you to every founder that has entertained us and allowed us to get smarter by going through our process, regardless of outcome. We’ve missed a lot of great deals but I expect that number to get smaller over time.
