Venture capitalists tout themselves as frontier technology investors, but most of us are using the same infrastructure tools we’ve used for the past 20+ years — Excel and recent college grads searching Google.
We’ve seen some modest progress in people upgrading from Excel to Google Sheets, along with the use of CRM and cloud-based storage services, but according to Sebastian Soler, who oversees data science at Lux Capital, less than 5% of American VCs have a full-time team member who’s focused on technology.
“While the arguments for adopting the latest technology are now too compelling to ignore, finding the required budget for specialized tools can often prove to be a major challenge, especially for smaller managers,” said Tim Friedman, founder of PEStack. “Comprehensive market data can cost upwards of $25k for a leading service, portfolio monitoring can be double that, add in front office tools and you’re quickly into six-figure sums. My advice is: there are now more products than ever which focus on quick implementation and offer a lot of functionality at a fraction of the cost of some of the larger legacy providers.
TotemVC* is one example of a high-quality solution that offers a powerful platform with a transparent, affordable monthly rate. One piece of advice would be to use a service like [PEStack’s] free Vendor Profiles platform to identify viable providers and build up a shortlist. We also track sample clients so that our users can see what their peers are using. I would always advise managers to talk to other professionals to get the real inside scoop on which products work well, how painful the implementation was, and how good the ongoing support is.”
Jonathan Balkin, founder of Lionpoint Group, observed that the highest-impact technology initiative for a new PE/VC fund is typically to configure and enforce usage of a CRM system. The next most impactful initiative is usually to create an easy-to-use LP portal.
Johann Kratzer of Blue Future Partners, a VC fund of funds, wrote that “the majority of the hundreds of funds we’ve diligenced rely predominantly on their relationships to source deals. However, an increasing share of VCs in our database are trying to build a competitive advantage by crawling large amounts of publicly available data and building analytical functions to flag companies with accelerating traction to them. However, it’s still quite unclear, how (and if) the use of technology correlates with financial returns for VC firms.”
Historically, investing was a manual, artisan process. An investor had few hard metrics other than the actual financials and little technology to make the process scaleable.
Over the past few decades, however, better metrics became available, allowing investors to take a more analytical, data-driven approach. The extreme example of this are algorithmic investors in the public markets who design algorithms that trade on the designer’s behalf, as opposed to making trading decisions directly. High-frequency trading, algorithmic by nature, is estimated to account for at least 50% of trading volume in U.S. equity markets.
Quantitative, technology-enabled investing in private companies makes sense but is structurally very difficult and will become a more common strategy at a much slower rate. The private markets are more opaque; they offer less of the hard data critical to a true quant approach. In venture capital in particular, early-stage companies are often operating in frontier industries where the rules are unpredictable and conventional analytic frameworks may be misleading.
Even for later-stage companies with predictable financials, the lack of liquidity, audited financials and standardized metrics creates real challenges to scaling quantitative investing. However, there is a lot of room to use technology to make the actual investing process more efficient.
Why is it now more feasible to use technology in the VC investing process?
“Structured, accurate and accessible data never really existed before for the private markets, at scale,” says Sebastian Soler. “Advances in machine learning, specifically natural language processing, have made generating these baseline, aggregate datasets possible, at scale, with high accuracy. Sources like Crunchbase, AngelList, and SeedInvest even give this data away for free or very low cost. The problem that faces startup investors now is how to mine this new data layer efficiently to increase returns.”
As an example of what’s possible, CircleUp has built “Helio,” a platform that proactively finds, classifies and evaluates consumer packaged goods companies — a “knowledge graph for CPG.” It specifically focuses on CPG because CircleUp believes that’s the sector in private markets where analytic techniques are most applicable in origination. Today, Helio tracks more than 1 million companies across more than 200 sources, comprised of a combination of private, public and partner data.
CircleUp has built three investing strategies on top of the Helio platform: CircleUp Growth Partners (a $125m venture fund), CircleUp’s credit fund ($200m AUM) and their upcoming “Systematic VC” strategy, which relies entirely on Helio to quantitatively select the portfolio of companies.
Stereotypically, venture capitalists are viewed as less numerically focused than private equity investors. With regard to analyzing a given company’s financial model, that is a reasonable stereotype, given that VCs do not typically use financial leverage and financial forecasts of early-stage companies have a very high uncertainty rate. That said, I’d suggest the VC industry is more advanced than private equity in using technology in many other steps in the overall investing process, given the technology savvy of VCs, the transparency of many companies in the technology industry and the relative receptivity of our limited partners to trying new ideas.
In private equity, the investors most proactively trying to use quantitative techniques are primarily sector specialists. According to one study, sector specialists regularly outperform generalist funds, returning an aggregate 2.2x multiple on Invested Capital versus 1.9x MOIC. Sector specialists have always subconsciously correlated certain factors as drivers for future growth, but this knowledge is now being distilled into quantitative analysis.
“Sector funds, because of their ability to mine sufficiently targeted data, are working backwards to identify key correlations between different KPIs — including more unusual ones like positive PR releases to more basic fundamental analysis — that drive value at exit,” said Mariya Osadchaya-Isa, who worked previously at a sector-focused private equity fund.
“This information is used to automate screening during origination. Automated screening ensures a higher quality deal pipeline that isn’t as prone to performance drags like mission creep and personal bias. Sector funds lend themselves well to doing this because it’s easier to quantify the characteristics which drive value on a sector basis, e.g., if you focus on SaaS companies you can identify annual change of >10% in recurring revenues as an indicator of future stability of revenues. This kind of ability to build a multivariate analysis suitable to be deployed on a company by company basis is not suited for generalist funds, unless they have sector strategies nestled underneath a generalist umbrella.”
I have embedded below the presentation from a talk I gave at the “Alpha Innovation Required” summit on “Who is Building the Renaissance Technologies of Private Markets?” (A video is available here.)
*I’m an investor in this company, via ff Venture Capital, HOF Capital and/or an affiliate thereof.
Special thanks for their comments to Blue Future Partners, Karim Fattal, Steven Greenberg; Clay Hunt, Andrew Kangpan; David Levine; Mariya Osadchaya-Isa; Sebastian Soler; Andrew Sudol; Jim Tousignant, CEO, FinTech Studios; Franklin Tsung; various anonymous friends; and to research firm AskWonder, in which I’m an investor.