[Reposted from Prof. Noam Wasserman's blog here.]
When do co-founders split the equity equally, and does an equal split affect their venture's later valuation?
We are currently conducting our annual CompStudy survey. (I'll soon be posting details about how to participate.) Over the last two years, I have enhanced the Founding Team section of the survey, so we can get a deeper view of various founding issues.
The first issue I am revisiting is how founders split equity among themselves, an issue that I first surfaced in a series of posts over the last few years (e.g., Splitting the Pie: Founding Team Equity Splits, Equity-Split Results, Part 1: When Do Teams Split Equally?, Equity-Split Results, Part 2: Implications for Team Stability, The Idea Premium: How Much (Equity) is Your Idea Worth?, and New: "A Note on the Legal and Tax Implications of Founders' Equity Splits"). For these new analyses, I have been collaborating with Prof. Thomas Hellman, and our initial paper has just been accepted for the annual summer-time conferences of both the Academy of Management and the National Bureau of Economic Research (NBER).
In the paper, we take a detailed look at the drivers and the financing consequences of founder equity splits. Below I outline the core results in each of these two areas, and would love to get your input about them (especially about the financing consequences).
Note: For these analyses, we combined the data from the 2008 and 2009 CompStudy surveys, including both Technology and Life Sciences ventures. Across the two years, we received complete survey submissions from a total of 576 multi-founder teams. Dropping 65 repeat respondents in 2009 gave us a full dataset of 511 ventures, which included 1,476 total founders.
Part 1: Drivers of the Initial Split
In the initial models, we examine whether the founders' backgrounds affected whether they split equally (the "1/N" model) or unequally. The core hypothesis is that negotiating an equity split is "costly" (both directly and indirectly), so teams that are relatively homogeneous (e.g., they are "close enough" regarding their years of work experience, prior founding experience or lack thereof, and other factors) should be more likely to avoid the negotiation and just split equally. In contrast, more heterogeneous teams should be more likely to engage in the negotiation, in order to arrive at an unequal split that tries to match equity stakes to expected contributions.
The analyses support that hypothesis, and also suggest that less experienced teams are more likely to split equally. More specifically, the following teams were more likely to split equally:
On the flip side, the following teams were more likely to negotiate an unequal split:
Within the teams that negotiated an unequal split, the following founders received significantly bigger equity stakes:
Any thoughts on these results?
Note: When we controlled for the differences listed above (i.e., prior founding experience, idea person, and seed capital), each founder's overall years of prior work experience did not affect that founder's equity stake.
Part 2: Financing Effects
This part of the analysis is a little more tentative, but has interesting initial results. For this, we looked at the determinants of (the natural logarithm of) the venture's pre-money valuation during its first institutional round of financing. Of our full sample, 298 ventures had raised a round within 3 years of founding, and were included in this analysis.
The question was whether the founders' equity split (equal vs. unequal) was linked to their pre-money valuations. In these models, we controlled for the following factors:
(We also included dummy variables for the venture's location, industry, and the year in which the round was raised.)
The variables that were significantly linked to a lower valuation were as follows:
The variables that were significantly linked to a higher valuation were as follows:
We next focused on the "equal equity split" result. We separated it into "quick-equal splits" (the proverbial "Quick Handshake," where the founding team spent a day or less negotiating the equal equity split) and "slow-equal splits" (the teams who spent more than a day negotiating, but still ended up splitting equally), and reran the models. This showed that the negative impact of equal splits is almost entirely attributable to the quick-equal splitters (which had a significant and negative impact on valuation) and not to the slow-equal splitters (not significant).
We do not believe that there is a direct effect of the quick-equal split on valuation -- i.e., that it is the split itself that causes a lower valuation. (Please comment if you disagree!) Instead, we believe that the quick-equal split may indicate something deeper (more indirect) about the founding team: a preference to avoid tough discussions, an inability to negotiate effectively, or some other negative trait regarding individual characteristics or team dynamics.
*** Any thoughts on this linkage between quick-equal splits and firm valuation? ***
Next Steps: At the venture level, we're trying to quantify how much the valuation is affected. At the founder level, we're also trying to estimate how much money the strongest equal-splitting founder leaves on the table (compared to if that founder had gotten the full unequal stake s/he deserved). Looking forward to posting those results...