Starting A Venture Gapital Fund

Yesterday, I had a fun meeting with Sloan Entrepreneurs for International Development. My host, Jenny Kwan, asked me to come up with a topic to brainstorm with this group. So we spent an hour discussing what it would take to start a venture fund for enterprises in the BOP. What would be the same? what would be different?

We currently have a bifurcated market. On the one hand, traditional venture capital funds (VCs) look for deals that have the potential to return 30%+ on multi-million dollar investments. On the other hand, foundations and groups like Ashoka "invest" in social entrepreneurs without an expectation of even a return of principal. Kiva provides a return of principal, as do the "social businesses" advocated by Muhammad Yunus. So the question was whether a fund would work to fill this "gap" for BOP businesses that can generate a positive return, but would not attract VC due to a number of factors (smaller deal size, unfamiliar markets, lower ROI).

Here are a few ideas we kicked around as "venture gapitalists":

Fund size: I poked at the idea of "go big or go home" with a $100 million+ fund, but the general sense of the group was that this would need to be a relatively small fund ($15-25 million).

Deal flow: Fund size was driven by concern about deal flow in this sector. There was a good discussion about whether the fund will pick sectors (such as Acumen Fund) or instead invest in emerging businesses (such as those assisted by Endeavor). One idea that emerged was that it might be possible to carve up the General Partners' carried interest to those that helped with deal flow. We also discussed that others were doing due diligence on BOP firms for awards, etc, and it might be possible to take advantage of this starting point to help with deal flow. Several people also were concerned about investing in social entrepreneurs- they wanted to invest in profit maximizing entrepreneurs who would be pursuing liquidity for investors.

Team: Unanimous sense that you would need to build a team with local connections in the countries and sectors. Some concern that there were not many "serial social entrepreneurs" that could be recruited. Interesting to see the MBA perspective that few would be willing to start with fund if they felt it would require long term sacrifice on compensation. Most seemed to think that, properly structured, the fund could be competitive on compensation.

Cost structure: Many were concerned with the additional cost of a global network. While salaries are lower in other parts of the world, opening offices, reviewing deals, sitting on portfolio company boards all raised concerns.

General Partner fees: With cost concerns, some students felt a higher management fee would be required. 3-5% was proposed, but the "investors" were skeptical.

Syndication: The lack of other funds was a concern, as there are few others (yet) to share risk, due diligence, etc.

Investment strategy: some concern about investing in subsequent rounds (we noted that funds that are too small get squeezed as winners emerge in a portfolio if they can't invest in subsequent rounds). The sense of the group was that early rounds would be several hundred thousand (NOTE: this 1/10 conventional VC) and perhaps a bit over $1 million total capital. So the portfolio would be 15-30 companies.

Instruments: we didn't get into this in the main discussion, but after class we had an interesting chat about "additional terms" in, for instance, a convertible debt instrument. What if, for instance, the interest rate (or conversion discount) varied based on achieving social/environmental goals. The idea was that some investors might be willing to trade down on valuation if social returns were ACTUALLY being generated (as opposed to merely promissed) by the entrepreneur.

In short, we didn't figure it out. My sense is a number of people have been thinking about this, but have yet to come up with a business model that "works" enough to attract the human capital to run it, the financial capital to fund it, and an efficient deal flow screening mechanism. Just a matter of time, I think, before a few more "testable" models will emerge.